We have a lot of power supply behind-the-scenes videos over on Patreon, including one we just published this morning! Check them out here and support us at the same time: patreon.com/gamersnexus If you want to see our PSU efficiency testing in action, check out our Aresgame AGS850 review: ruclips.net/video/EmeFRtoiIrQ/видео.html We did a PSU explosion speedrun here: ruclips.net/video/7JmPUr-BeEM/видео.html Here's a video on PSU & GPU power behavior form previously, featuring electrical engineer TiN (formerly EVGA): ruclips.net/video/V2dqHqlyWio/видео.html Finally, if you want to see what it looks like when a PSU safely trips OPP, we did a speedrun of that as well: ruclips.net/video/2SleaZ68ZO0/видео.html
@@ianmoone8244 True, and I wouldn't disagree. Though I'd be inclined to avoid any company willing to do something as made up and deceptive as "90 Plus". It's like one job applicant with a college degree - showing they spent the time and money at least to get a diploma. Then another job applicant with a "Harverd University" certificate they hope you don't scrutinize too closely.
@@bitelaserkhalif should have subscribed to the ibxtoycat channel, since 99% of house fires happen to those who are not subscribed (wonder if anyone will get the reference)
yup after this video my average trust for 80+ went down the shitter If its this simple to bypass a psu's actual certification then it might as well not have it. Not to say that I wont factor it in but my likelihood of recommending a product with that factor in mind is none. user reviews is where its at
Yes. Not been a fan of the prebuilt series but I get there is maybe an audience for it. I don’t really watch content to get dunked on so no loss on my part but if gn is looking for feedback, keep this stuff up.
igorsLAB actually did a video on 80 Plus certificates last year, but it's good to see Gamers Nexus bringing attention to this to a more general audience
This is something so many of my friends don't get. 80+ is not a mark of quality, it is a rough guide on efficiency. Thank you Steve. This is some great PSA work.
Efficiency is often related to the quality of internal components though. You are highly unlikely to find a titanium certified no brand name PSU. It’s not an exact science, but there is a link.
Well, they have stock they reused and retest for all sorts of reasons from verifying baselines to creating new test benches. It's important to know when stock's defective, so they don't do something like creating a test bench with bad components. And, unlike you and I, they can't throw out bad or defective items on the off-chance they need to provide proof of their claims. So, while it seems silly for those stickers, the stickers do serve a good purpose.
I think it's important to remember where 80 Plus came from. Before the standard was established it wasn't uncommon that PSU efficiency was in the 60% range. That is if you burned 200W at the wall then 80W was immediately converted to heat in the PSU while the computer itself got 120W to feed on. For a PSU 70% efficiency was considered very good, but efficiency was not something that you ever heard a word about when talking PSU's. This might not sound that horrible, but multiply that by a few hundred computers with these 60% efficiency PSU's in a business and it adds up. At the same time there were a lot of talk about how we had to start saving power. There were a lot of talk about how we should turn off the light in a room when we left it, turn off TV's and such instead of having them in standby and in some countries daylight saving time was introduced in order to make better use of the daylight during the summer to cut down on the need for electrical lighting. With all of this going on, and computers needing more power making more efficient PSU's was quite logical. When 80 Plus was first introduced there were very few PSU's that could pass the certification. A few years ago the 80 Plus gold PSU's was as common as hens teeth. That is they were hard to make and only a handful of PSU's were able to achieve this efficiency. Back then a design that was able to pass this was very likely to be well thought through and very well designed and built, and the prices reflected that. The same was the case with the Platinum grade PSU's when those first became available. These were the absolute top of the line PSU's and only used the best components. As more and more PSU manufacturers managed to create platforms meeting the Gold or even Platinum standard the law of price and availability meant that the competition grew and price became more important. Today making a Gold compliant PSU is no longer some kind of black art, but making them cheap is. The result is that there are a lot of certified gold compliant PSU's being sold that are built to a price point that's way to low to allow for the nicer things like stable voltage control and quality long life components. And that's why 80 Plus is no longer of any use when you try to find out the quality of a PSU, other than what efficiency it has, and then that's only valid if the PSU's sold are built to the same standard as the PSU used for certification.
There are other youtubers who test power supplies, but yeah it's a hassle that even with that crap sticker we still have to search tests from a known community person to know if it's a good buy since the companies pay the standard for nothing as it's all just a big BS for other companies to rip off buyers and the company making those shitty standards to get rich without providing any kind of service at all to customers other than snake oil...
@@guily6669 I fully agree. and also, who stops a random company faking this sticker? it's not like GN could or even want to afford a literal army of lawyers.
I love how GN makes a reference to the Gigabyte power supply nearly, if not every video they make. Gigabyte deserves every bit of flac, and more for the crap they pulled.
I remember in late 2000s when I was active in a tech forum and recommending PC parts to look for PFC and 80 Plus as minimum elements to consider for a quality PSU, but only as a first filter to remove quick junk. Nowadays it's much harder to find quality PSUs as there is way more brands than before and multiple OEMs are used by the same brand even in the same product line.
It's much easier today to find quality PSUs because our standards for judging them have risen so much, that today a barely more than average PSU is a near equivalent to a high end PSU from 20 years ago, except when the PSU has chinese capacitors and a sleeve bearing fan, the two most common components to fail as long as the PSU is operated within a margin of its rating, dust cleaned out once in a blue moon, and doesn't seen any power line surges. Remember, testing PSU against their upper limits or even 70% of it keeps the manufacturers more honest, but you don't have to limit yourself to just buying barely enough PSU unless you have some monster mining or gaming rig where every watt counts. If you pick a less than ideal PSU but it powers the system fine for a decade or longer, what does it matter if someone else likes a different PSU more for philosophical reasons? Look to OEMs who have a lot of experience providing enough PSU for the system (as shipped) without expensive excess, but of course OEMs sometimes have quality control problems too.
So one of the things I would had liked to see is more discussion over where PSU's were before and after the 80+ certifications. When it was introduced there was a massive issue of PSU's being overly inflated ratings for what they were. It was not uncommon to have PSU's 20-30% below their advertised wattage. There was also tons of HUGE QC issues with these PSU's and scant reviews into them in any details so the risk was very real to buy a PSU that would literally catch fire or die. In the early 2000's I had replaced many burnt out PSU's. A lot of times they were advertised more on their cool wiz-bang (windows lights fancy cables) features vs actual quality. PC Power and Cooling was a premier supplier of PSU's because they used to actually include a sheet with their test results of your PSU which was pretty much the only mark of quality we had back then. When HardOCP started testing the PSU's we started seeing tons of popular PSU's that would just fail all the time. It was really bad. Getting a 600w PSU? You can safely assume it would only ever be able to hit 420w. I would like to see an expansion of the certifications maybe the UL certifications in terms of PSU Safety coupled with 80+ for efficacy. I don't know if there is a way to have a industry standard of "This is a high-quality unit' without just relying on reviews.
I doubt the GN team has enough people, time nor interest in doing so though. And considering that certifications without a regular recertification process are kinda pointless consumers are better of buying goods tested by knowledgeable sources, preferably multiple different ones.
@@barcodeguy7627 GN, or rather Steve, said thats his endgame with the GN brand. But he said GN would not be certifying PC stuff as long as the GN brand is also doing review, because that's a conflict of interest.
@Samurai Shampoo Imagine simping so hard for a company that you still defend them even after they certified "gold" a power supply that literally explodes yikes.
I'm really glad you guys as a bigger outlet tackle this stuff. It gets kind of tiresome trying to explain to people that the efficiency rating is no indicator for the quality of the PSU.
Yes, they are confusing and there's not a lot of great documentation on them for entry-level and intermediate knowledge levels. We're trying to help with that!
@@SpaceLion949 You shouldn't thrust those forums too much. There is always someone trying to push some agenda at some point and you can never really tell where. Might be just my personal experience though.
@@_Teorias_ I have read the methology, as I said, it's good to get a decent general idea. Ideally you would want to read specialized reviews for the specific unit, but not everyone is capable of understanding those. Electricity is a topic that requires a lot of study. I don't really know what you mean buy pushing an agenda. I would 100% prefer to refer to some kind of guide like that before making a PSU purchase instead of not knowing anything at all about the product I'm interested in. For people that don't have knowledge in the topic, I'd say is a good advice.
ngl, Gamers Nexus content is keeping my interest on PC hardware alive. I'll save this video whenever someone asks me if a PSU is good/better based on 80plus alone. Back to you Steve.
TL;DR - The 80 Plus certification program can be a good indicator to consider when purchasing a power supply. But it should not be your only point of consideration.
Look at the ATX standard a PSU claims to follow. The ATX standard focuses on physical size, cable connectors, voltages, ripple etc that relates to the PSU being *usable* and *replaceable* in a system. So a motherboard vendor must design based in the ATX standard since that's the standard that dictates what is handed over to the power connectors on the motherboard.
What's funny to me is Coolermaster flexing about standards. I had one coolermaster psu about 5 years ago. In returned it because 12V line dropped to 11.4V in load.
@@Herbertti3 yeah, that's pushing it, I recall 'good' XT power-supplies being in the 12.25V to 11.75V range eons before ATX12V lol. 11.4V is oof. that was probably on the edge and just skimmed past QA at that voltage.
Can just say my 12V drops to 11.74V and have ben like that for 12 years soon can just say so long it stays stable around 12V its fine for me when I oc my GPUs 11.4V that's high drop from 12V and every 0.1V does a bigg difference for my oc
Rarely hear it mention on this type of videos.... But inefficient comes from heat and magnets field generated on the PSU and there is also noise and unstable voltage when power goes up and down due to the dynamic way a PC draw power (turbo boost + power off/iddle on components)
There was a website back last time I bought a power supply (in 2011) that did EXTENSIVE testing on components including power supplies, but I can't remember the name. They verified the efficiency sicker, did ripple and "clean" power measurements, talked about the rails (having many 12v rails was all the rage back then as it was "more stable"), and even a tear-down looking for all the safety and quality components installed and overall construction. I wish I could remember the name of the website, because I can't find it again. Its possible it's gone now.
@@cameronwebster6866 I really don't know. I thought it was "hardware secrets" or something very very close to that, but googling that didn't get me there when I tried to find it again. I was fairly sure "hardware" appeared in the name somewhere, but I no longer have any idea at all.
@@GamersNexus want a thermal paste for laptop with which paste should I go with kryonaut, noctua nt nh1 ,or artic mx4,or kingpin kpx or gelid gc extreme which one will last 8 -10 months without drying or significant rise in temp
80+ Gold might as well be a legacy standard by now. Realistically, we need to revamp based on the Titanium specifications and test down to standby efficiency. Maybe they should add in those extra protection features as well fir certification.
Protection really should be covered by an improved ATX specification - the standard that defines voltages, regulatiin, ripple etc. Another thing - Titanium is hard to reach and it's mainly big PSU that can manage this certification. It's a bit like the star rating for car impact testing. Small and large cars can get the same rating - but they need to keep a lower bar for 5 stars from a small car, because it can't have as big zones to take up energy etc. PSU technology is constantly moving forward, but having mass market "normal" (commodity) PSU manage Titanium is still out of our abilities. We have reached a level where Gold is reasonably easy to fulfill, but it's still making an impact on the total BOM cost.
Those low load/idle scenarios are why it's not the best idea to buy a 1000W unit for basically anyone. Even with a 5800X and RX 6700 XT you're barely over 400W under full load, but are under 80W in idle. With a 1000W unit idle load goes way below 10% and that is where efficiency goes down quickly. For something like that a good 550W unit will be a way better choice with around 15% idle and 70% load, moving pretty much within the 20% and 80% test points, where PSUs are quite efficient.
Gonna just ask you straight-up, Steve: is there any specific PSU series from any brand that you deem consistently reliable? Or are all brands incapable of such a feat nowadays, haha.
@@RetropUk Oh yeah, I did my research for specific units and settled on a specific Seasonic unit when I built this machine last year, as it offered nice quality parts, has good power handling and no weird cutoff issues with strong/rapid load transients (kind of a must-have if you want to run an overclocked Ryzen 3000 series CPU alongside a Powerplay-overcurrented Vega, the current peaks are not to be underestimated!) I was aiming to get a reply the general public might have a good use for, as sort of a sidenote to this video. :)
Most brands have low-end and high-end series products. They may make excellent - but expensive - products. But also wants some of the market shares from the mass market. Just look at how many different series of PSU Corsair has.
Reminds me of the Corsair VS450, around the cheapest from a respectable brand, but you get what you pay for, it's their most lowend model, so expect some rather mediocre quality. On the other hand, there are some decent LC-Power units that are better than the name suggests. They aren't super high quality units for the enthusiast overclocker, but okay enough to not explode if you put a bit load onto it. In the end if all comes down to the internals, and those can be quite different between different series of the same brand and even different models within a series (Looking at the recently famous Gigabyte ones, where only the 750 & 850W ones were crap, but the rest are fine) Basically, if you see a unit and nobody among your friends/family/colleagues has ever heard of the company stay away, but if it's a well known brand, take a look for reviews for that particular model and revision.
I had to settle for Obtainium Brown 25- because of shortages. It runs on coal and dead pandas. The smog helps to hide the pixilation on screen though, so that's nice.
I was concerned with 10%< loads consumption and was amazed how anything below titanium doesn't have a minimum percentage efficiency. I went with a seasonic titanium one year and half ago when it costed like 150€ new for a 750w.
honestly if i ever win the lottery im throwing money at gamers nexus to go out find a huge team of legit people and to test many many products out there. just to find out what is good or not
I must be old. I remember when they introduced the 80 plus standard and it was clear at the time that it was a measure of power efficiency only. It is strange that the perception of what it means has warped over time.
I'm so thankful for people like Steve who call out big companies and their bullshit. And who has a def impact on sales of shitty products and force manufacturers to think twice, or even try just a little bit harder. And then he lets us know which companies are just lazy and try to strong-arm him instead..
Yeah, one of my ex's had a Geo Metro. It survived a few cross country trips (Back and Fourth between Cincinnati, OH and Phoenix, AZ) It even handled the Apache Trail without falling off the cliff that ended so many 4WD vehicles. Not my first pick in cars, but that one did hold up on it's own.
I work in a Quality Control lab at an OSB plant. We have a grade stamp printed on all of our boards that it meets a certain standard/span rating, it is required by the government. The company that issues that stamp will come at random times to audit our lab, and test random boards in the warehouse himself. This same thing should be done for PSUs
Well known and good are two *different* metrics. And being well known from one niche doesn't mean a company has skills in other niches. So the "well known" may originate from motherboard or graphics cards products and the PSU "division" might be a CEO + board of directors that gets an idea that maybe there is money in PSU too. And then buys a design from a company you have never heard about.
@@OdinAlgeron But it's the brand on the label and box that a buyer will see. And Corsair is a good example of how just a brand name alone isn't enough.
It doesn't help that manufacturers and stores use the rating as a marketing tactic and push it as a quality standard, and that most people don't have a background in electronics. Not even for people that don't build systems often, because a lot of people that do also don't know how to properly evaluate an SMPS because...either they just aren't interested in electronics, think it's too hard, or don't even know that they're being manipulated. You're doing a noble service for the community, good sir. It ain't their fault, you're one of the few people that have even tried to explain that it's not a standard of quality. It's typical marketing wankery that lies and manipulated people into shit. Most people would have no idea what harmonic distortion or power factor even is, and there's nothing wrong with that. There is something wrong with manufacturers and retail outlets using that to fucking lie to people and push the idea that 80 Plus is the standard of quality they should be looking at to judge an SMPS. ...marketing seems to be responsible for a lot of this dirty tactic bullshit, huh
Hey Steve, just curious as to what your thoughts are on Cybenetics' rating/certification efficiency levels and if they would be better than the 80 Plus system.
In some ways, "80 Plus Gold" or "80 Plus Platinum" used to mean somewhat more for overall power supply quality than it does these days. It meant more not due to any testing for quality directly, but simply because it used to be the case you needed quality components and engineering to reach those efficiency levels. In recent times however, advances in the technology have allowed cheaper components and lazier engineering to reach those efficiency levels, Of course, even with good components/engineering you can still have systematic assembly/QC failings, but having quality components/engineering is the first step.
Yes, back in the day you couldn't just whip out a dogshit 80plus gold from a backalley in China. Last PSU I bough was a titanium one after extensive research, and I was baffled by the variation displayed by the units and yes all of it were in normal range but it's still interesting to to see. There is still a chance that a worse product exhibits behaviour that lies outside "normal" limits
I don't know how good of a change it would be, but it would be nice to see when exactly the PSU was 80+ certified. For example, seeing 80+ Gold July 2021 would show immediately that it's a newer unit and it was certified or re-certified recently, which helps sort out some of the BS.
I have a pretty old 550 watt seasonic power supply (around 12 years old now) I got from a used PC and used it again for my Dell optiplex "gaming" PC. Still seems to work pretty good but I'd imagine the capacitors are dying out.
Everyone should, if at all possible, plug the computer Power into a UPS (uninterruptible power supply) which ALSO HAS line ‘conditioning’ and ‘ brown-out protection. This will go a LONG WAY to protecting your rig ... and the power supply ... from damage from ‘dirty power’ and power line fluctuations which depending on location can be horrendous.
Long time viewer just writing to say I'm really glad that there's been time to do all these power supply videos, can't wait to see reviews and/or round ups to know where PSUs compare
I still have a Corsair HX650 from 12 years ago that still somehow works. It's kept my computer on when the lights flicker during a thunderstorm... pretty decent. I wonder if they still make the same quality power supplies
Im quite sure they do. I have a VS550 from 2012 (that have been living in a heavy smoker home) got the psu for free as spare parts. Yes, the fan sounds like an idleing fire engine but the PSU still delivers. My personal CX750 from 2013, still working, and i just replaced it with a new model of CX750F with leds.
The thing I love most about the power efficiency graphs is how clearly it shows how little you gain by being in a PSUs "Sweet spot" and puts to rest the old wives tale of having too high wattage of a power supply.
Yes and no; modern systems idle at extremely low power, and only Titanium actually lists any efficiency below 20%. While modern supplies have pretty flat curves from 15-30% up to near peak load, efficiency still can fall off a cliff below 10% or so. If you have an 800W power supply and a system that idles below 80W (a fairly generous idle number these days) then you're down in the area where efficiency can be 50-60% or lower even on fairly good supplies. You certainly don't need to go crazy optimizing your supply for the exact expected load, but getting a PSU that's rated for double or triple your expected load will still leave you a lot lower on the curve at idle than you'd like, since modern components are so much better about reducing idle power than they used to be.
@@siberx4 just to add some numbers: 80+ gold, lowest tested load (20%) total - tested 400w - 80w 640w - 128w 800w - 160w 1200w - 240w And that's generous not only for idle in most consumer pcs, it's generous for normal office work and web browsing as well. Gaming will be usually above, though in a low spec system you could still use less power than some of these bigger PSUs are tested for. Consoles usually are pretty power efficient I believe, but ps5/series X both use around 210w peak iirc - just to have another comparison.
@@siberx4 Though to be fair low efficiency with very low loads isn't that bad given it's low power in the first place, so while in relative terms it's not great, in absolute terms is not a big deal in regards to extra heat output and wasted power consumption.
@@kazedcat True enough, but computers tend to spend a lot more of their time at or near idle than they do at full load so in terms of total energy, idle has a more significant impact than its low power consumption would otherwise imply. If you run your system with "regular" tasks like web browsing 8 hours a day at say 50W and run it full load 1 hour a day at 400W, you'd actually be split 50/50 for your total kWhr consumed between idle and full load. If your idle is very inefficient, it would still make up the majority of your overall power consumption in that case.
There is a reason that for a very long time, product lists made by comsumer groups have existed and continues to exists. There are a couple PSU and SSD lists out there and they mostly deal with the products quality as opposed to its collection of badges, awards or its marketing.
They should test sample provided by manufacturer before mass production and then one they buy themselves from random retailer after launch. If the retail sample does not match manufacturer sample CPSC should issue recall and pass any and all costs on the manufacturer. That would quickly solve this issue.
When building a PC these days I just don’t think I can trust a single component especially when there are zero repercussions for intentionally selling faulty components and pulling the old bait and switch methods that have been reported on.
I once had a power supply short out and completely die on me before. Granted it was because I was using an underpowered PSU after upgrading the computer I was using, but still. It ran Division 2 for around 5 minutes before I saw a flash and everything powered off with a smokey smell emitting from the power supply. Only know it was a short because I opened it up and nothing looked damaged.
What was the outcome of this? Did your computer still work afterwards? Sometimes a PSU failure takes the the whole system with it. Or did you just replace the power supply and everything still worked.
@@MrR2185 I had to replace that power supply, it was 100% dead. I had concerns initially that it fried everything else (though it would have still been under warranty for the other parts), but after receiving the replacement which had higher wattage, everything worked fine, so I got really lucky.
In the end there are less manufacturers than brands. Seasonic is one of them who does both, build them for other companies and sell them under their own name. FSP is another one.
G'day Steve, Company Marketing take every opportunity to Twist the meanings of Buzz Words & Phrases to make their Product seem the best, but I'm so impressed by the work GN does to help people understand even basics of Tech Specs & Quality Standards & how they Truly Apply in the PC & Digital Gaming space.
I think the most important part about this piece is the detailed explanation of the certification process, because a lot of certification processes across all sorts of industries share a lot of similarities to this. A person could make an entire career out of simply doing reviews on certification systems themselves. Somebody should.
And the most wold about it is, 80PLUS is based on 120V but PSUs are more efficient on a 230V grid. So something that doesn't get an 80PLUS label can still reach >80% in almost everywhere outside of North America as everyone noton North and Central America or Japan runs on 220-240V
I learned long ago anything that delivers power, where money can be made by obscuring and obfuscating the specifications to fool people who don’t know any better, even well known brands are going to abuse that for profit. It’s been going on with PSU’s for as long as I’ve been building, 20 years. And it’s absolutely rampant with car audio amplifiers. Even brands like Sony straight up lie about their amps power output and hide the RMS rating in tiny text somewhere on the box if they show it at all. 800 watts “peak power”, 150 watts RMS. That’s a weak amp that sounds like a beast and it’s probably over priced.
I've been running Seasonics for over 10 years in my most important rigs. They are not cheap; however, the rigs keep running, no weird faults, no breaking down and they are pretty efficient.
The reason why 80 plus always made sense to me is that: if the power supply had the 80 plus platinum or titanium mark on it then it have to be made with high quality components even golden ones usually comes in a higher quality. It is not logical for the manufacturers to make such an expensive power efficient unit with poor quality components. Poor quality components also can't even claim high efficiencies BUT, you came and just proved that is no always the case and that is cool. Thank you steve
I view this as the same as with restaurants. Food may be good, but if the kitchen isn't clean, I don't wanna eat there. Or if the stuff is made with roadkill. Something like that. With that in mind, there's a rating system our government uses for all establishments serving food or drinks. And that system uses 4 different emojis, with frown, neutral, slight smile and smile. And then the final big grade is equal to the worst grade they receive. This works for restaurants. For technical stuff like this, think we can stick with the Bronze, Silver, Gold, Platinum, Titanium system, but just expand it to measure all of these things we want to keep track of and then worst grade is the certification. And you're not allowed to switch any components whatsoever - like with ESD certifications. And as for how you're allowed to market the component, you'd be allowed to print out the certification details and the rating received, but on the item itself you couldn't have any rating other than the overall rating.
I have a 1200 Watt Vulcan PSU 80 Plus Bronze, runs for over 10 years now. It powered 8 different machines, almost all SLI/Crossfire high-end setups and it is still running. I do clean it every time I build new stuff around it though. I guess it's pretty good :) It is massive and has incredibly thick components in it. I don't even know if the brand still exists lol, back then I never ever heard from it. I bought it directly from a distributor who could not get rid of them, so I got it for half the price, around 100 euro's. Best buy ever :)
"80+" really says very little about efficiency too. In particular, a higher rating (silver, gold, whatever precious metal struck their fancy) doesn't mean the power supply will be more efficient for a particular workload - usually they're just the same power supply design rated for more power, so the overheads only get proportionally smaller when compared to the power rating. They still consume an equal amount of power themselves. That aside, of course it's like most brands not saying anything about those who don't use the logo. I wasn't aware they specified testing at 60Hz specifically. That's neat in that it means those aren't real world conditions for any power supply in Europe.
I personally put a LOT of Faith in EVGA PSUs. For some, it's Corsair or Silverstone. Could you do a piece Validating/invalidating the Avid PC builder's sometimes "Blind" faith in these popular and reputable brands? Are our preconceived self-assurances warranted? I am pretty sure that most of these companies don't own PSU factories of their own. They are more like design houses. SO which ones use "Good" factories and components consistently? These are the type of questions I would like to see addressed. A new standard is certainly in order NOW. Great informative Vid!!!
You are at a point in your career where you can honestly change the industry. Your voice is being heard. It would be nice if there were another certification that covers everything that 80plus doesn't. You should make a video covering all the psu's you have tested and rank them most recommended to least recommended without including costing as a factor of recommendation.
I would love for you to test a mining power supply (Octominer shown, or MiningCave) Both rate their units as 90+ Gold but I think that's only if you run it on ~240v, but if you're pulling well over 1KW at the wall sustained with one system(6+ GPUs), I'd think the unit has to be at least decent not to explode on first use
Honestly I don't blame them for sticking 90+ Gold stickers on these things, the whole 80+ thing is clearly a scam designed to make companies give theses guys money for basically nothing...you can send them any unit you like for testing, if by some miracle one of your retail units is found wanting they are "encouraged" to send another for testing...at their own cost...and anyone who challenges them gets stuck with the bill when they send the higher quality prototype back for testing (probably the same one they sent in the first place). They don't even have the decency to post it back to them for free after they just paid them thousands for 20 minutes work! There's zero benefit for any of this to the consumer, just another dirty marketing scam.
80+ is a mark of quality in a very limited extent, because there is a limit to how many corners you can cut specially when it comes to the actual design of the PSU while still achieving anything close to the rating (for example you can't label a 400W PSU as 750W PSU, because there is no way it would even reach the power level required nevermind efficiency). The PSUs that don't meet 80+ standard are so much worse that they make that Gigabyte unit that you exploded look like a luxury product. One of these days I hope you test one of those, I don't think you've exploded an early 2000s pre-Pentium 4 PSU design sold in 2021 yet. They do still sell those, though it took me a while to find one on Newegg. Seems like they are much less common today than they were years ago. Speaking of which, I am submitting a request for a playlist of power supply reviews. Right now there is no playlist of that.
With the Cybenetic's testing I personally see a potential problem, they do 5VSB which is arguably the most inefficient part of a power supply. However are there not plans to phase out 5 volts and move that work to the motherboard, therefore making that testing moot.
The two most important components in any PC build for me is #1 The Case, and #2 The Power Supply. Those are the biggest causes of unstable, flaky builds where components get too hot because of crap airflow or a power supply with too much ripple and voltage creep under load and higher temperature. Bad PSUs can cause lots of odd problems that are very hard to diagnose..
As someone who has been building my own systems as well as part time for others, for decades I never depended much on the 80+ ratings, because I had used so many PSU over the years that were prior to the rating system, and I still have many of those PSU from systems I've retired, as well as repairing some. The main things I learned are to get away from idealism, for example lots of people like to think a group regulated PSU is bad, but it really doesn't matter if the rails match the load of the system, none droop excessively and/or kick in voltage protection. Another thing is that it doesn't fail, and I know that seems simple enough but I would rather a PSU not have a lot of money poured into higher efficiency before spend on highest quality capacitors and DUAL BALL BEARING fan. Don't give me some trumped up, specially named sleeve bearing fan and tell me that's okay for a horizontally mounted application because it really isn't if I want the fan to last the otherwise viable lifespan of the PSU. Dual ball bearing fan makes a little bearing whine but really not that much once PSU moved to 12cm or larger fans that don't need high RPM to get the job done, unless of course taking about a high wattage PSU with a large load. That is not the average system out there so we'd be talking exceptions rather than the norm. Also, use a good surge protector. Transistors don't fail from overheating if you have enough margin in the design but are still susceptible to mains line spikes. It makes me wonder if the next evolution of PSU is going to be some built in surge protection. Even a mere couple MOVs would beat nothing and be little expense and real estate taken.
A question: My friend is an electrical engineer. He once (~ 2000-ish) told me a simple rule of thumb: If you have two PSUs (same specs, like 650W - in the "old" days you didn't know anything else), the one that is heavier is probably better. I adopted this rule, and never had any problems up to 2010. In 2010 was the "crossover" point when I started buying everything online. So in the last 10 years, I first read a lot of different reviews before buying. But I'm curious. If you have two PSUs, with the same specs, was it true that the one that's heavier was probably better?
It's still usable. The main thing that makes the weight difference is the size of main transformator (trafo). Trash PSU always uses cheapo-small-light weight trafo.
I did get 80plus titanium psu for my computer. Because of the 10% efficiency requirement when computer is idle. I did choose the power rating by theoretical maximum power of what the computer would draw with the components of my choosing. I wanted that max power draw to be in the ballpark of about 60% load of the psu. Maybe I did end up choosing too powerful psu, but at least it is almost always between 90-94% efficiency.
Paying extra for a PSU that runs hotter due to marketing decisions is misguided. That's not overkill at all, rather a crippled unit that would be so much better with a fan strapped on anyway even if running at such low RPM that it's inaudible. Granted, such designs depend on unequal case pressurization so it just moves the burden of airflow to a different fan, more likely one mounted on a case wall so more noise escapes than would from a PSU fan itself which is kind of ironic. I suppose that doesn't make it "bad" per se but definitely a less universal fit if trying for low noise yet still need case fans to move air.
@@stinkycheese804 Not really. I tried a ton of PSUs over the years anything from cheap to expensive and "high quality" and the only PSU with a fan i would consider silent is the Corsair SF750 and only because the fans basicly never spin up even under load anyway and has no whine either. Any other fan PSU either had horrible fans, horrible fan curves or whine when under load. I using the Seasonic Prime Fanless Platinum 500 and Titanium 700 and both of them never get hot even under complete load, not even warm to the touch, they basicly stay at room temp. And both of them are 100% silent. I know a lot of people don't really notice it that much and don't even care if their fans run at like 3000 RPM, but i basicly built my two PC around complete silence and no other PSU aside from these and the SF750 did the job.
We have a lot of power supply behind-the-scenes videos over on Patreon, including one we just published this morning! Check them out here and support us at the same time: patreon.com/gamersnexus
If you want to see our PSU efficiency testing in action, check out our Aresgame AGS850 review: ruclips.net/video/EmeFRtoiIrQ/видео.html
We did a PSU explosion speedrun here: ruclips.net/video/7JmPUr-BeEM/видео.html
Here's a video on PSU & GPU power behavior form previously, featuring electrical engineer TiN (formerly EVGA): ruclips.net/video/V2dqHqlyWio/видео.html
Finally, if you want to see what it looks like when a PSU safely trips OPP, we did a speedrun of that as well: ruclips.net/video/2SleaZ68ZO0/видео.html
Hello GN, I really like your mouse pad but could you make a glow in the dark option because that would be awesome.
4:39 Noooooo! It’s an LTT Segway! Arrgghhhhh! 💥💥🤯🤯
your videos lack cats recently.
Can you test more psus the way you test cpu coolers to see which ones are actually good?
actually the beer foam is just some aerated liquid, it has a density between 1/4 and 1/7 of the actual liquid. So this analogy doesn't work that well.
You need to create a GN certification program.
Tiers:
"It doesn't explode"
"Better than Dell"
"It's fine"
A higher rating would be "We like it"
You forgot the other tier, fireworks
Is it sad that id probably hold all future psu's to this standard?
"buy it right now" -the next decent ~~$250 GPU that shows up in the market if such a thing were to ever happen again
"It doesn't offend us"
"90 Plus is how many firefighters will show up to fight the fire in your house if you use that." this is pure gold
you mean pure 80+ gold?
@@Cinetyk "90 Plus" was referring to the reddit screenshot at 9:09
@@rhekman he’s making a psu pun off the original comment. Instead of saying “this is pure gold” he’s being funny saying “this is pure 80+ gold”
Both are the same thing in the end, marketing to sell!
@@ianmoone8244 True, and I wouldn't disagree. Though I'd be inclined to avoid any company willing to do something as made up and deceptive as "90 Plus".
It's like one job applicant with a college degree - showing they spent the time and money at least to get a diploma. Then another job applicant with a "Harverd University" certificate they hope you don't scrutinize too closely.
i used to think it meant it had an 80% chance of not blowing up
Those are good chances!
@@GamersNexus That would be an excellent chance for the gigabyte psu.
Don’t know...but now i’m worrying about my sfx cooler masster v850...
Exactly lol, my thoughts too
Me when i first got into pc building
I lost it at the picture of the "Octominer" PSU with a "90 PLUS Gold" rating.
90+ explosions and house fires
@@bitelaserkhalif should have subscribed to the ibxtoycat channel, since 99% of house fires happen to those who are not subscribed
(wonder if anyone will get the reference)
@@chlorobyte_projects I know the reference. I'll leave this here: 500.
can't wait when they start slaping 100 PLUS 😂😂
@@Romanitto 110 plus when? 110% efficiency, don't ask how that works!
These are the videos the community needs
fo sho
yup after this video my average trust for 80+ went down the shitter If its this simple to bypass a psu's actual certification then it might as well not have it. Not to say that I wont factor it in but my likelihood of recommending a product with that factor in mind is none. user reviews is where its at
Yes. Not been a fan of the prebuilt series but I get there is maybe an audience for it. I don’t really watch content to get dunked on so no loss on my part but if gn is looking for feedback, keep this stuff up.
igorsLAB actually did a video on 80 Plus certificates last year, but it's good to see Gamers Nexus bringing attention to this to a more general audience
Too right mate
That reddit comment about 90+ being the number of firefighters who will show up to your house if you use that power supply is hilarious!
This is something so many of my friends don't get. 80+ is not a mark of quality, it is a rough guide on efficiency. Thank you Steve. This is some great PSA work.
Thanks for saying what the video said. 👍
well i used to believe what your friends think, and alot of people selling these still think that
Show me a crappy Platinum or Titanium rated PSU, and I'll give you $1000.
Im really happy to see this; i voted for this in the poll
Efficiency is often related to the quality of internal components though. You are highly unlikely to find a titanium certified no brand name PSU. It’s not an exact science, but there is a link.
15 mintute intervals, Gigabyte would classify that as Extended testing and edge case usage
''only servers run for that long'' -someone at Gigabyte, probably...
@@theoakun3635servers run 15+ days so it makes sense
I did that experiment a few times last night. I think it definitely helped me understand power factor
It’s hilarious how you had “do not use” stickers on the Gigabyte power supply boxes!
Well, they have stock they reused and retest for all sorts of reasons from verifying baselines to creating new test benches. It's important to know when stock's defective, so they don't do something like creating a test bench with bad components. And, unlike you and I, they can't throw out bad or defective items on the off-chance they need to provide proof of their claims.
So, while it seems silly for those stickers, the stickers do serve a good purpose.
@@jackielinde7568
"We had a testbench with a 3090 and
I think it's important to remember where 80 Plus came from. Before the standard was established it wasn't uncommon that PSU efficiency was in the 60% range. That is if you burned 200W at the wall then 80W was immediately converted to heat in the PSU while the computer itself got 120W to feed on. For a PSU 70% efficiency was considered very good, but efficiency was not something that you ever heard a word about when talking PSU's. This might not sound that horrible, but multiply that by a few hundred computers with these 60% efficiency PSU's in a business and it adds up. At the same time there were a lot of talk about how we had to start saving power. There were a lot of talk about how we should turn off the light in a room when we left it, turn off TV's and such instead of having them in standby and in some countries daylight saving time was introduced in order to make better use of the daylight during the summer to cut down on the need for electrical lighting. With all of this going on, and computers needing more power making more efficient PSU's was quite logical. When 80 Plus was first introduced there were very few PSU's that could pass the certification.
A few years ago the 80 Plus gold PSU's was as common as hens teeth. That is they were hard to make and only a handful of PSU's were able to achieve this efficiency. Back then a design that was able to pass this was very likely to be well thought through and very well designed and built, and the prices reflected that. The same was the case with the Platinum grade PSU's when those first became available. These were the absolute top of the line PSU's and only used the best components.
As more and more PSU manufacturers managed to create platforms meeting the Gold or even Platinum standard the law of price and availability meant that the competition grew and price became more important. Today making a Gold compliant PSU is no longer some kind of black art, but making them cheap is. The result is that there are a lot of certified gold compliant PSU's being sold that are built to a price point that's way to low to allow for the nicer things like stable voltage control and quality long life components.
And that's why 80 Plus is no longer of any use when you try to find out the quality of a PSU, other than what efficiency it has, and then that's only valid if the PSU's sold are built to the same standard as the PSU used for certification.
This is a high-quality comment that deserves many more upvotes.
Are you ok dude?
We need GN+ stickers on power supplies. Gamers Nexus certified
If only... Sadly there's too many PSUs
@@Empireo_sebastian They are effectively making their own lab in the new building. Perhaps a testing group of high ethics will be part of their future
There are other youtubers who test power supplies, but yeah it's a hassle that even with that crap sticker we still have to search tests from a known community person to know if it's a good buy since the companies pay the standard for nothing as it's all just a big BS for other companies to rip off buyers and the company making those shitty standards to get rich without providing any kind of service at all to customers other than snake oil...
@@guily6669 I've been looking for exactly those reviewers. Could you drop some names?
@@guily6669 I fully agree. and also, who stops a random company faking this sticker? it's not like GN could or even want to afford a literal army of lawyers.
"The beer is the real power!" ................. AMEN!
now we know alot more about steve, talking about beer, selling beer glasses, beer mats
@@arch1107 You review enough pre-built PCs, you'll be driven to drinking as well.
@@rhekman lol true
oktobeer confirmed tommorow
@@arch1107 the belly is a giveaway
The wall of dead Gigabyte power supplies--wow. GN is ruthless.
I love how GN makes a reference to the Gigabyte power supply nearly, if not every video they make. Gigabyte deserves every bit of flac, and more for the crap they pulled.
Good old beer analogy for apparent power, I knew it was coming when you mentioned power factor lol
I remember in late 2000s when I was active in a tech forum and recommending PC parts to look for PFC and 80 Plus as minimum elements to consider for a quality PSU, but only as a first filter to remove quick junk.
Nowadays it's much harder to find quality PSUs as there is way more brands than before and multiple OEMs are used by the same brand even in the same product line.
It's much easier today to find quality PSUs because our standards for judging them have risen so much, that today a barely more than average PSU is a near equivalent to a high end PSU from 20 years ago, except when the PSU has chinese capacitors and a sleeve bearing fan, the two most common components to fail as long as the PSU is operated within a margin of its rating, dust cleaned out once in a blue moon, and doesn't seen any power line surges.
Remember, testing PSU against their upper limits or even 70% of it keeps the manufacturers more honest, but you don't have to limit yourself to just buying barely enough PSU unless you have some monster mining or gaming rig where every watt counts.
If you pick a less than ideal PSU but it powers the system fine for a decade or longer, what does it matter if someone else likes a different PSU more for philosophical reasons? Look to OEMs who have a lot of experience providing enough PSU for the system (as shipped) without expensive excess, but of course OEMs sometimes have quality control problems too.
As it stands today: filter by manufacturer: Seasonic. Doesn't guarantee they won't make shit in future though.
So one of the things I would had liked to see is more discussion over where PSU's were before and after the 80+ certifications. When it was introduced there was a massive issue of PSU's being overly inflated ratings for what they were. It was not uncommon to have PSU's 20-30% below their advertised wattage. There was also tons of HUGE QC issues with these PSU's and scant reviews into them in any details so the risk was very real to buy a PSU that would literally catch fire or die. In the early 2000's I had replaced many burnt out PSU's. A lot of times they were advertised more on their cool wiz-bang (windows lights fancy cables) features vs actual quality. PC Power and Cooling was a premier supplier of PSU's because they used to actually include a sheet with their test results of your PSU which was pretty much the only mark of quality we had back then. When HardOCP started testing the PSU's we started seeing tons of popular PSU's that would just fail all the time. It was really bad. Getting a 600w PSU? You can safely assume it would only ever be able to hit 420w.
I would like to see an expansion of the certifications maybe the UL certifications in terms of PSU Safety coupled with 80+ for efficacy. I don't know if there is a way to have a industry standard of "This is a high-quality unit' without just relying on reviews.
Wtf are you talking about?
We need a new rating for PSUs that defines the overall quality and chances of explosions. The Burke Rating.
Won't happen. These companies thrive on marketing tactics to mislead their customers.
ltt psu tier list for now.
you mean the beve sturk rating?
The Beve Sturke rating scale
@@eveangreen1871 wow sorry I didn’t see you beat me to it
I love how every episode now has a joke at Gigabyte’s expense.
@Samurai Shampoo I don't know... I find the joke about Gigabyte's power supply funny every time.
Yeah I'm pretty sure the exploding power supply clip has made it into EVERY video released since the occurrence :D
@@sultanofsick could you say that the topic has BLOWN UP since the occurrence? :D
@@LaughingSkull451 haha
Man, GN is on a roll lately! Loving the content.
Gamers nexus really could become a certification company, the whole industry knows that these guys know what they are talking about.
I doubt the GN team has enough people, time nor interest in doing so though.
And considering that certifications without a regular recertification process are kinda pointless consumers are better of buying goods tested by knowledgeable sources, preferably multiple different ones.
@@barcodeguy7627
GN, or rather Steve, said thats his endgame with the GN brand.
But he said GN would not be certifying PC stuff as long as the GN brand is also doing review, because that's a conflict of interest.
@Samurai Shampoo Imagine simping so hard for a company that you still defend them even after they certified "gold" a power supply that literally explodes yikes.
I'm really glad you guys as a bigger outlet tackle this stuff. It gets kind of tiresome trying to explain to people that the efficiency rating is no indicator for the quality of the PSU.
Learned this 2 years into my first build in 2014 (ish?)
Thank god there's some in depth PSU review articles online now
Needed this PSUs are confusing 🥺
Yes, they are confusing and there's not a lot of great documentation on them for entry-level and intermediate knowledge levels. We're trying to help with that!
@@GamersNexus keep calling out the cheaters and the bad manufacturers Steve.
@@GamersNexus There's a PSU Tier list on the LTT forum, I usually use it to get a decent general idea of the quality of the PSU I'm buying.
@@SpaceLion949 You shouldn't thrust those forums too much. There is always someone trying to push some agenda at some point and you can never really tell where. Might be just my personal experience though.
@@_Teorias_ I have read the methology, as I said, it's good to get a decent general idea. Ideally you would want to read specialized reviews for the specific unit, but not everyone is capable of understanding those. Electricity is a topic that requires a lot of study.
I don't really know what you mean buy pushing an agenda. I would 100% prefer to refer to some kind of guide like that before making a PSU purchase instead of not knowing anything at all about the product I'm interested in. For people that don't have knowledge in the topic, I'd say is a good advice.
ngl, Gamers Nexus content is keeping my interest on PC hardware alive. I'll save this video whenever someone asks me if a PSU is good/better based on 80plus alone.
Back to you Steve.
That's assuming they don't plug their ears and say they can't hear you.
TL;DR - The 80 Plus certification program can be a good indicator to consider when purchasing a power supply. But it should not be your only point of consideration.
80+Gold, who cares if the 12v rail drops to 7V at 40% load, so long as the efficacy is there, lol.
Look at the ATX standard a PSU claims to follow. The ATX standard focuses on physical size, cable connectors, voltages, ripple etc that relates to the PSU being *usable* and *replaceable* in a system.
So a motherboard vendor must design based in the ATX standard since that's the standard that dictates what is handed over to the power connectors on the motherboard.
What's funny to me is Coolermaster flexing about standards. I had one coolermaster psu about 5 years ago. In returned it because 12V line dropped to 11.4V in load.
@@Herbertti3 5% drop is juuuuust on the limit for +/- 5%.
@@Herbertti3 yeah, that's pushing it, I recall 'good' XT power-supplies being in the 12.25V to 11.75V range eons before ATX12V lol. 11.4V is oof. that was probably on the edge and just skimmed past QA at that voltage.
Can just say my 12V drops to 11.74V and have ben like that for 12 years soon can just say so long it stays stable around 12V its fine for me when I oc my GPUs 11.4V that's high drop from 12V and every 0.1V does a bigg difference for my oc
"90 plus is how many firefighters will show up at your house if you use this PSU" is a good one lol.
Rarely hear it mention on this type of videos....
But inefficient comes from heat and magnets field generated on the PSU and there is also noise and unstable voltage when power goes up and down due to the dynamic way a PC draw power (turbo boost + power off/iddle on components)
“the power we’re going to need if we review another power supply that lies about its efficiency “ 😆
I've been telling people for years that while an '80+ Gold' will "typically" indicate better quality components are used, it's not guaranteed.
Sounds like a market gap for fully tested and rated power supplies.
There was a website back last time I bought a power supply (in 2011) that did EXTENSIVE testing on components including power supplies, but I can't remember the name. They verified the efficiency sicker, did ripple and "clean" power measurements, talked about the rails (having many 12v rails was all the rage back then as it was "more stable"), and even a tear-down looking for all the safety and quality components installed and overall construction. I wish I could remember the name of the website, because I can't find it again. Its possible it's gone now.
@@sultanofsick Johnny guru?
@@cameronwebster6866 I really don't know. I thought it was "hardware secrets" or something very very close to that, but googling that didn't get me there when I tried to find it again. I was fairly sure "hardware" appeared in the name somewhere, but I no longer have any idea at all.
This is the kind of content that sets GN apart from the rest of tech RUclips. Awesome stuff.
only tech youtubers? the only other close is johnny guru and that is sort of gone now
This is the exactly the kind of content that I love from Gamers Nexus and why the community needs him.
EDIT: for them! I forget there is a whole team.
Needs them! It's a whole team behind these videos!
@@GamersNexus my bad, fixed the comment, super cool yall replied though. Love the content.
@@GamersNexus want a thermal paste for laptop with which paste should I go with kryonaut, noctua nt nh1 ,or artic mx4,or kingpin kpx or gelid gc extreme which one will last 8 -10 months without drying or significant rise in temp
@@aringiri1946 Toothpaste and honey
Hardware manufacturers sweating profusely whenever GN posts a new video...
80+ Gold might as well be a legacy standard by now. Realistically, we need to revamp based on the Titanium specifications and test down to standby efficiency. Maybe they should add in those extra protection features as well fir certification.
Protection really should be covered by an improved ATX specification - the standard that defines voltages, regulatiin, ripple etc.
Another thing - Titanium is hard to reach and it's mainly big PSU that can manage this certification. It's a bit like the star rating for car impact testing. Small and large cars can get the same rating - but they need to keep a lower bar for 5 stars from a small car, because it can't have as big zones to take up energy etc.
PSU technology is constantly moving forward, but having mass market "normal" (commodity) PSU manage Titanium is still out of our abilities. We have reached a level where Gold is reasonably easy to fulfill, but it's still making an impact on the total BOM cost.
Those low load/idle scenarios are why it's not the best idea to buy a 1000W unit for basically anyone.
Even with a 5800X and RX 6700 XT you're barely over 400W under full load, but are under 80W in idle. With a 1000W unit idle load goes way below 10% and that is where efficiency goes down quickly. For something like that a good 550W unit will be a way better choice with around 15% idle and 70% load, moving pretty much within the 20% and 80% test points, where PSUs are quite efficient.
Gonna just ask you straight-up, Steve: is there any specific PSU series from any brand that you deem consistently reliable? Or are all brands incapable of such a feat nowadays, haha.
Not Steve and I’d still do my home work but superflower/ evga and seasonic are usually solid.
@@RetropUk Oh yeah, I did my research for specific units and settled on a specific Seasonic unit when I built this machine last year, as it offered nice quality parts, has good power handling and no weird cutoff issues with strong/rapid load transients (kind of a must-have if you want to run an overclocked Ryzen 3000 series CPU alongside a Powerplay-overcurrented Vega, the current peaks are not to be underestimated!)
I was aiming to get a reply the general public might have a good use for, as sort of a sidenote to this video. :)
Most brands have low-end and high-end series products. They may make excellent - but expensive - products. But also wants some of the market shares from the mass market. Just look at how many different series of PSU Corsair has.
Most of Seasonic Power Supplies are great. Except the S12 series, those are OEM and not made by Seasonic themselves, but are probably also fine.
Reminds me of the Corsair VS450, around the cheapest from a respectable brand, but you get what you pay for, it's their most lowend model, so expect some rather mediocre quality.
On the other hand, there are some decent LC-Power units that are better than the name suggests. They aren't super high quality units for the enthusiast overclocker, but okay enough to not explode if you put a bit load onto it.
In the end if all comes down to the internals, and those can be quite different between different series of the same brand and even different models within a series (Looking at the recently famous Gigabyte ones, where only the 750 & 850W ones were crap, but the rest are fine)
Basically, if you see a unit and nobody among your friends/family/colleagues has ever heard of the company stay away, but if it's a well known brand, take a look for reviews for that particular model and revision.
This is why I only buy PSUs with the Unobtainium 1000 ratings.
i see your poor, my psu has a unobtainium 1000+1 rating
Wait, is this higher tier than unobtainium Diamond super duper excellent? 🤔
I had to settle for Obtainium Brown 25- because of shortages.
It runs on coal and dead pandas.
The smog helps to hide the pixilation on screen though, so that's nice.
@@KillaBitz Yours does AA without taking up GPU resources?! Neat! 🥳
@@KillaBitz saves on the heating bill am i right, when i was in college i had a similar psu and modded it into an oilburner
I was concerned with 10%< loads consumption and was amazed how anything below titanium doesn't have a minimum percentage efficiency. I went with a seasonic titanium one year and half ago when it costed like 150€ new for a 750w.
GN should create their own standard for power supply quality
honestly if i ever win the lottery im throwing money at gamers nexus to go out find a huge team of legit people and to test many many products out there. just to find out what is good or not
Been waiting on this. Informative and proves the industry needs an new standard of rating parts.
I must be old. I remember when they introduced the 80 plus standard and it was clear at the time that it was a measure of power efficiency only. It is strange that the perception of what it means has warped over time.
I'm so thankful for people like Steve who call out big companies and their bullshit. And who has a def impact on sales of shitty products and force manufacturers to think twice, or even try just a little bit harder. And then he lets us know which companies are just lazy and try to strong-arm him instead..
The million-dollar question is: what are the recommended sources?
My PSU is a Corsair 1200w 80+ Platinum... It's been running close to 10 years without any issues. 👍
The Geo Metro would like a word with you about being absolutely crap quality and crap safety, but still getting great mileage
Yeah, one of my ex's had a Geo Metro. It survived a few cross country trips (Back and Fourth between Cincinnati, OH and Phoenix, AZ) It even handled the Apache Trail without falling off the cliff that ended so many 4WD vehicles. Not my first pick in cars, but that one did hold up on it's own.
I work in a Quality Control lab at an OSB plant. We have a grade stamp printed on all of our boards that it meets a certain standard/span rating, it is required by the government. The company that issues that stamp will come at random times to audit our lab, and test random boards in the warehouse himself. This same thing should be done for PSUs
Usually just go with something well known like EVGA or Seasonic or Corsair or BeQuiet or something
@Stolas for exploding nowadays 😆
Until they start pulling a Western Digital, MSI, and/or Gigabyte.
Well known and good are two *different* metrics. And being well known from one niche doesn't mean a company has skills in other niches. So the "well known" may originate from motherboard or graphics cards products and the PSU "division" might be a CEO + board of directors that gets an idea that maybe there is money in PSU too. And then buys a design from a company you have never heard about.
corsair is not a manufacturer like seasonic
corsair buys units and rebrands them, some are from good builders some are chinese crap
@@OdinAlgeron But it's the brand on the label and box that a buyer will see. And Corsair is a good example of how just a brand name alone isn't enough.
It doesn't help that manufacturers and stores use the rating as a marketing tactic and push it as a quality standard, and that most people don't have a background in electronics. Not even for people that don't build systems often, because a lot of people that do also don't know how to properly evaluate an SMPS because...either they just aren't interested in electronics, think it's too hard, or don't even know that they're being manipulated. You're doing a noble service for the community, good sir.
It ain't their fault, you're one of the few people that have even tried to explain that it's not a standard of quality. It's typical marketing wankery that lies and manipulated people into shit.
Most people would have no idea what harmonic distortion or power factor even is, and there's nothing wrong with that. There is something wrong with manufacturers and retail outlets using that to fucking lie to people and push the idea that 80 Plus is the standard of quality they should be looking at to judge an SMPS.
...marketing seems to be responsible for a lot of this dirty tactic bullshit, huh
Hey Steve, just curious as to what your thoughts are on Cybenetics' rating/certification efficiency levels and if they would be better than the 80 Plus system.
Talked about it a little bit at the end!
Yep. My 3 year old Seasonic Platinum power supply died and leaked all over the place & even leaked onto the rug & destroyed that.
Holy crap. What's the model? Did you over use it ?
In some ways, "80 Plus Gold" or "80 Plus Platinum" used to mean somewhat more for overall power supply quality than it does these days.
It meant more not due to any testing for quality directly, but simply because it used to be the case you needed quality components and engineering to reach those efficiency levels. In recent times however, advances in the technology have allowed cheaper components and lazier engineering to reach those efficiency levels, Of course, even with good components/engineering you can still have systematic assembly/QC failings, but having quality components/engineering is the first step.
Yes, back in the day you couldn't just whip out a dogshit 80plus gold from a backalley in China. Last PSU I bough was a titanium one after extensive research, and I was baffled by the variation displayed by the units and yes all of it were in normal range but it's still interesting to to see. There is still a chance that a worse product exhibits behaviour that lies outside "normal" limits
I don't know how good of a change it would be, but it would be nice to see when exactly the PSU was 80+ certified. For example, seeing 80+ Gold July 2021 would show immediately that it's a newer unit and it was certified or re-certified recently, which helps sort out some of the BS.
When buying a PSU, it's just a question of which one from Seasonic to choose 😅
I have a pretty old 550 watt seasonic power supply (around 12 years old now) I got from a used PC and used it again for my Dell optiplex "gaming" PC. Still seems to work pretty good but I'd imagine the capacitors are dying out.
Everyone should, if at all possible, plug the computer Power into a UPS (uninterruptible power supply) which ALSO HAS line ‘conditioning’ and ‘ brown-out protection. This will go a LONG WAY to protecting your rig ... and the power supply ... from damage from ‘dirty power’ and power line fluctuations which depending on location can be horrendous.
I feel betrayed like how Luke's father was Darth Vader
why would u spoil bro i habent watched the star wars
@@L39T Could be worst .... at least he didn't tell you that leila is his sister!
@@titiwin25 yeah you're right. if he told me that i'd commit a war crime!
Spoilers WTF >:C
I'm crying how could you spoil
Long time viewer just writing to say I'm really glad that there's been time to do all these power supply videos, can't wait to see reviews and/or round ups to know where PSUs compare
Only thing I will trust is GN certified :)
You guys are really good at what you do, so I don't think that will will be a far fetch idea.
I still have a Corsair HX650 from 12 years ago that still somehow works. It's kept my computer on when the lights flicker during a thunderstorm... pretty decent. I wonder if they still make the same quality power supplies
Im quite sure they do. I have a VS550 from 2012 (that have been living in a heavy smoker home) got the psu for free as spare parts. Yes, the fan sounds like an idleing fire engine but the PSU still delivers.
My personal CX750 from 2013, still working, and i just replaced it with a new model of CX750F with leds.
The thing I love most about the power efficiency graphs is how clearly it shows how little you gain by being in a PSUs "Sweet spot" and puts to rest the old wives tale of having too high wattage of a power supply.
Yes and no; modern systems idle at extremely low power, and only Titanium actually lists any efficiency below 20%. While modern supplies have pretty flat curves from 15-30% up to near peak load, efficiency still can fall off a cliff below 10% or so. If you have an 800W power supply and a system that idles below 80W (a fairly generous idle number these days) then you're down in the area where efficiency can be 50-60% or lower even on fairly good supplies.
You certainly don't need to go crazy optimizing your supply for the exact expected load, but getting a PSU that's rated for double or triple your expected load will still leave you a lot lower on the curve at idle than you'd like, since modern components are so much better about reducing idle power than they used to be.
@@siberx4 just to add some numbers: 80+ gold, lowest tested load (20%)
total - tested
400w - 80w
640w - 128w
800w - 160w
1200w - 240w
And that's generous not only for idle in most consumer pcs, it's generous for normal office work and web browsing as well.
Gaming will be usually above, though in a low spec system you could still use less power than some of these bigger PSUs are tested for.
Consoles usually are pretty power efficient I believe, but ps5/series X both use around 210w peak iirc - just to have another comparison.
@@siberx4 Though to be fair low efficiency with very low loads isn't that bad given it's low power in the first place, so while in relative terms it's not great, in absolute terms is not a big deal in regards to extra heat output and wasted power consumption.
80% efficiency at 90% load waste a lot more energy than a 20% efficiency at 10% load
@@kazedcat True enough, but computers tend to spend a lot more of their time at or near idle than they do at full load so in terms of total energy, idle has a more significant impact than its low power consumption would otherwise imply. If you run your system with "regular" tasks like web browsing 8 hours a day at say 50W and run it full load 1 hour a day at 400W, you'd actually be split 50/50 for your total kWhr consumed between idle and full load. If your idle is very inefficient, it would still make up the majority of your overall power consumption in that case.
13:00 So during warm-up and load transitions, there could be bad voltage/noise levels but no testing is required
There is a reason that for a very long time, product lists made by comsumer groups have existed and continues to exists. There are a couple PSU and SSD lists out there and they mostly deal with the products quality as opposed to its collection of badges, awards or its marketing.
They should test sample provided by manufacturer before mass production and then one they buy themselves from random retailer after launch. If the retail sample does not match manufacturer sample CPSC should issue recall and pass any and all costs on the manufacturer. That would quickly solve this issue.
When building a PC these days I just don’t think I can trust a single component especially when there are zero repercussions for intentionally selling faulty components and pulling the old bait and switch methods that have been reported on.
that's why it's important to have brand reputation and reviews
I once had a power supply short out and completely die on me before. Granted it was because I was using an underpowered PSU after upgrading the computer I was using, but still. It ran Division 2 for around 5 minutes before I saw a flash and everything powered off with a smokey smell emitting from the power supply. Only know it was a short because I opened it up and nothing looked damaged.
What was the outcome of this? Did your computer still work afterwards? Sometimes a PSU failure takes the the whole system with it. Or did you just replace the power supply and everything still worked.
@@MrR2185 I had to replace that power supply, it was 100% dead. I had concerns initially that it fried everything else (though it would have still been under warranty for the other parts), but after receiving the replacement which had higher wattage, everything worked fine, so I got really lucky.
I used to do Underwriter Laboratory testing for a manufacturer. The UL label only means that the device is unlikely to electrocute the user.
"if you're Phanteks buying a Seasonic power suply" - hey that's about my power supply!
In the end there are less manufacturers than brands. Seasonic is one of them who does both, build them for other companies and sell them under their own name. FSP is another one.
G'day Steve,
Company Marketing take every opportunity to Twist the meanings of Buzz Words & Phrases to make their Product seem the best, but I'm so impressed by the work GN does to help people understand even basics of Tech Specs & Quality Standards & how they Truly Apply in the PC & Digital Gaming space.
I think the most important part about this piece is the detailed explanation of the certification process, because a lot of certification processes across all sorts of industries share a lot of similarities to this. A person could make an entire career out of simply doing reviews on certification systems themselves. Somebody should.
Looks like you are really gaining traction on views and Subs . Congrats !
I feel the puns in your scripts are underrated, so commenting so you know they are very much appreciated :)
And the most wold about it is, 80PLUS is based on 120V but PSUs are more efficient on a 230V grid. So something that doesn't get an 80PLUS label can still reach >80% in almost everywhere outside of North America as everyone noton North and Central America or Japan runs on 220-240V
Newegg doesn't have a ripple current filter to sort by. :(
First lecture from GN I was able to fully make thru. Thanks for the in-depth research, as always.
Steve remains one of the best consumer advocates in this industry. Keep up the good work GN--proud to have one of the wireframe mouse mats on my desk.
I was actually considering a PSU upgrade recently. Thanks for this.
I learned long ago anything that delivers power, where money can be made by obscuring and obfuscating the specifications to fool people who don’t know any better, even well known brands are going to abuse that for profit. It’s been going on with PSU’s for as long as I’ve been building, 20 years. And it’s absolutely rampant with car audio amplifiers. Even brands like Sony straight up lie about their amps power output and hide the RMS rating in tiny text somewhere on the box if they show it at all. 800 watts “peak power”, 150 watts RMS. That’s a weak amp that sounds like a beast and it’s probably over priced.
I've been running Seasonics for over 10 years in my most important rigs. They are not cheap; however, the rigs keep running, no weird faults, no breaking down and they are pretty efficient.
Maybe they just need to hit 80% efficiency right before exploding
The reason why 80 plus always made sense to me is that: if the power supply had the 80 plus platinum or titanium mark on it then it have to be made with high quality components even golden ones usually comes in a higher quality.
It is not logical for the manufacturers to make such an expensive power efficient unit with poor quality components.
Poor quality components also can't even claim high efficiencies
BUT, you came and just proved that is no always the case and that is cool. Thank you steve
As a power systems engineer, I applaud your power factor explanation!! Perfectly done.
I view this as the same as with restaurants. Food may be good, but if the kitchen isn't clean, I don't wanna eat there. Or if the stuff is made with roadkill. Something like that.
With that in mind, there's a rating system our government uses for all establishments serving food or drinks. And that system uses 4 different emojis, with frown, neutral, slight smile and smile. And then the final big grade is equal to the worst grade they receive. This works for restaurants.
For technical stuff like this, think we can stick with the Bronze, Silver, Gold, Platinum, Titanium system, but just expand it to measure all of these things we want to keep track of and then worst grade is the certification. And you're not allowed to switch any components whatsoever - like with ESD certifications.
And as for how you're allowed to market the component, you'd be allowed to print out the certification details and the rating received, but on the item itself you couldn't have any rating other than the overall rating.
Thought that the level meant quality, till the bronze rated power supply from new egg had lasted me 8 years
Do you consider 8 years to be good or bad?
sounds decent to me, but i dunno, i dont have a psu with that age until now yet
I have a 1200 Watt Vulcan PSU 80 Plus Bronze, runs for over 10 years now. It powered 8 different machines, almost all SLI/Crossfire high-end setups and it is still running. I do clean it every time I build new stuff around it though. I guess it's pretty good :) It is massive and has incredibly thick components in it. I don't even know if the brand still exists lol, back then I never ever heard from it. I bought it directly from a distributor who could not get rid of them, so I got it for half the price, around 100 euro's. Best buy ever :)
@@kaldo_kaldo considering it was a off brand no name company yes, It could have lasted longer maybe but i replaced it
"80+" really says very little about efficiency too. In particular, a higher rating (silver, gold, whatever precious metal struck their fancy) doesn't mean the power supply will be more efficient for a particular workload - usually they're just the same power supply design rated for more power, so the overheads only get proportionally smaller when compared to the power rating. They still consume an equal amount of power themselves.
That aside, of course it's like most brands not saying anything about those who don't use the logo.
I wasn't aware they specified testing at 60Hz specifically. That's neat in that it means those aren't real world conditions for any power supply in Europe.
I only use 110+ Rhodium Pro certified PSUs in my builds
free electricity generator go brrr
Yooo hook me up with your over-unity PSUs
I personally put a LOT of Faith in EVGA PSUs. For some, it's Corsair or Silverstone. Could you do a piece Validating/invalidating the Avid PC builder's sometimes "Blind" faith in these popular and reputable brands? Are our preconceived self-assurances warranted? I am pretty sure that most of these companies don't own PSU factories of their own.
They are more like design houses. SO which ones use "Good" factories and components consistently? These are the type of questions I would like to see addressed.
A new standard is certainly in order NOW. Great informative Vid!!!
You are at a point in your career where you can honestly change the industry. Your voice is being heard. It would be nice if there were another certification that covers everything that 80plus doesn't. You should make a video covering all the psu's you have tested and rank them most recommended to least recommended without including costing as a factor of recommendation.
I would love for you to test a mining power supply (Octominer shown, or MiningCave)
Both rate their units as 90+ Gold but I think that's only if you run it on ~240v, but if you're pulling well over 1KW at the wall sustained with one system(6+ GPUs), I'd think the unit has to be at least decent not to explode on first use
Honestly I don't blame them for sticking 90+ Gold stickers on these things, the whole 80+ thing is clearly a scam designed to make companies give theses guys money for basically nothing...you can send them any unit you like for testing, if by some miracle one of your retail units is found wanting they are "encouraged" to send another for testing...at their own cost...and anyone who challenges them gets stuck with the bill when they send the higher quality prototype back for testing (probably the same one they sent in the first place). They don't even have the decency to post it back to them for free after they just paid them thousands for 20 minutes work!
There's zero benefit for any of this to the consumer, just another dirty marketing scam.
I have been looking forward to a video like this for so long, thank you! Your content is always fantastic
80+ is a mark of quality in a very limited extent, because there is a limit to how many corners you can cut specially when it comes to the actual design of the PSU while still achieving anything close to the rating (for example you can't label a 400W PSU as 750W PSU, because there is no way it would even reach the power level required nevermind efficiency). The PSUs that don't meet 80+ standard are so much worse that they make that Gigabyte unit that you exploded look like a luxury product. One of these days I hope you test one of those, I don't think you've exploded an early 2000s pre-Pentium 4 PSU design sold in 2021 yet. They do still sell those, though it took me a while to find one on Newegg. Seems like they are much less common today than they were years ago.
Speaking of which, I am submitting a request for a playlist of power supply reviews. Right now there is no playlist of that.
The Gold in 80+ Gold means it's good because that's the same color as the stars they give you at the end of Candy Crush levels.
With the Cybenetic's testing I personally see a potential problem, they do 5VSB which is arguably the most inefficient part of a power supply.
However are there not plans to phase out 5 volts and move that work to the motherboard, therefore making that testing moot.
The two most important components in any PC build for me is #1 The Case, and #2 The Power Supply. Those are the biggest causes of unstable, flaky builds where components get too hot because of crap airflow or a power supply with too much ripple and voltage creep under load and higher temperature. Bad PSUs can cause lots of odd problems that are very hard to diagnose..
As someone who has been building my own systems as well as part time for others, for decades I never depended much on the 80+ ratings, because I had used so many PSU over the years that were prior to the rating system, and I still have many of those PSU from systems I've retired, as well as repairing some.
The main things I learned are to get away from idealism, for example lots of people like to think a group regulated PSU is bad, but it really doesn't matter if the rails match the load of the system, none droop excessively and/or kick in voltage protection.
Another thing is that it doesn't fail, and I know that seems simple enough but I would rather a PSU not have a lot of money poured into higher efficiency before spend on highest quality capacitors and DUAL BALL BEARING fan. Don't give me some trumped up, specially named sleeve bearing fan and tell me that's okay for a horizontally mounted application because it really isn't if I want the fan to last the otherwise viable lifespan of the PSU. Dual ball bearing fan makes a little bearing whine but really not that much once PSU moved to 12cm or larger fans that don't need high RPM to get the job done, unless of course taking about a high wattage PSU with a large load. That is not the average system out there so we'd be talking exceptions rather than the norm.
Also, use a good surge protector. Transistors don't fail from overheating if you have enough margin in the design but are still susceptible to mains line spikes. It makes me wonder if the next evolution of PSU is going to be some built in surge protection. Even a mere couple MOVs would beat nothing and be little expense and real estate taken.
A question: My friend is an electrical engineer. He once (~ 2000-ish) told me a simple rule of thumb: If you have two PSUs (same specs, like 650W - in the "old" days you didn't know anything else), the one that is heavier is probably better. I adopted this rule, and never had any problems up to 2010. In 2010 was the "crossover" point when I started buying everything online. So in the last 10 years, I first read a lot of different reviews before buying. But I'm curious. If you have two PSUs, with the same specs, was it true that the one that's heavier was probably better?
It's still usable. The main thing that makes the weight difference is the size of main transformator (trafo). Trash PSU always uses cheapo-small-light weight trafo.
@@dhanarputra555 Thank you for an answer and quick explanation what was probably my friend talking about 2 decades ago...
Unless they put weights in the PSU, which some manufacturers have done, usually worse ones.
I did get 80plus titanium psu for my computer. Because of the 10% efficiency requirement when computer is idle. I did choose the power rating by theoretical maximum power of what the computer would draw with the components of my choosing. I wanted that max power draw to be in the ballpark of about 60% load of the psu. Maybe I did end up choosing too powerful psu, but at least it is almost always between 90-94% efficiency.
Steve, I recently got a new laptop with an amazing 4k mini led screen, you look different in 4k lol. your hair is even more majestic somehow.
I love that the Gigabyte exploding PSU footage just keeping popping up in any video they can make a rough connection
tfw 80 Plus Titanium Seasonic Prime Fanless 700
If there is one thing i will go overkill on, it's definitely the PSU.
Paying extra for a PSU that runs hotter due to marketing decisions is misguided. That's not overkill at all, rather a crippled unit that would be so much better with a fan strapped on anyway even if running at such low RPM that it's inaudible. Granted, such designs depend on unequal case pressurization so it just moves the burden of airflow to a different fan, more likely one mounted on a case wall so more noise escapes than would from a PSU fan itself which is kind of ironic. I suppose that doesn't make it "bad" per se but definitely a less universal fit if trying for low noise yet still need case fans to move air.
@@stinkycheese804 Not really. I tried a ton of PSUs over the years anything from cheap to expensive and "high quality" and the only PSU with a fan i would consider silent is the Corsair SF750 and only because the fans basicly never spin up even under load anyway and has no whine either. Any other fan PSU either had horrible fans, horrible fan curves or whine when under load.
I using the Seasonic Prime Fanless Platinum 500 and Titanium 700 and both of them never get hot even under complete load, not even warm to the touch, they basicly stay at room temp.
And both of them are 100% silent.
I know a lot of people don't really notice it that much and don't even care if their fans run at like 3000 RPM, but i basicly built my two PC around complete silence and no other PSU aside from these and the SF750 did the job.