Loads of testing of Ryzen 7000X cpus in "ECO mode" have already established that these CPUs are capable of great performance, even when TDP limited well below 120W. What on earth is Asrock doing to make the cpu panic throttle to 544mhz?
Great work. Don't underestimate the importance of reviewing bad - ahem (sorry) limited - products like this motherboard. Now we are informed and advised. Keep up the good work.
I consider Asrock is already doing a good job by releasing a BIOS fix in just 3 days after Steve's report. Somehow, that shows Asrock was taking input from external source seriously. They were wrong, but at least Asrock managed to correct them.
It would have been a fairly straight forward fix, change the power limit and raise the temp the VRM can hit, or it could have just been the latter. They've released another revision since then.
I can’t believe Asrock released this board in that state, Steve doing their darn R&D product testing for them fs. Man, good work exposing this and getting it fixed Steve!
You would think these mb companies would at least test configurations before putting out stated information, but no, we as consumers have to beta test all technology's these days. Keep up the benching!
@@mikem2253 I feel you! I used to buy to newest CPU/MB every 2 years, but it has bit me in the ass so many times in the last several years! I'm sticking with I have for the next 4 years at least and then I buy whatever chip is a generatio old at that time...I'm tired of beta testing $1000 CPU/MB for these companies!
Normal it have . Online you can see it My b450m msi have a 95watt tdp beult. Bit still i can use a 5800 serie on it And those cpu are 105watt tdp. Bios have it but the performance is just not thare. Funny thing is that automatic overklok on the x serie is not support.
You can usually tell, just by looking at it: - No heatsink on the VRMs: up to 65W with tower coolers and 95W with a topdown cooler - With heatsink on the VRMs: up to 95W with tower coolers and 125W with a big topdown cooler For anything beyond 125W, count the chokes: as a rule of thumb, each choke can easily handle 15W if cooled in a topdown manner or 10W if cooled by an AIO or a tower cooler. So basically for 170W, you should ideally have at least 12 chokes in total.
I hope you guys or someone else makes more videos on a620 boards. I know they will be commonplace for budget builds in the future. My used a320 board works well with my Ryzen 5600, and I managed to snag it for only $30 several months ago! I wish I saw some testing on the limits of the 65W boards. The "65W TDP" Ryzen 7000 CPUs actually consume about 90W without PBO on all core load according to GN. The "120W" 7800X3D is around the same power consumption at 90W on all core load, so if the boards are fine with a 7600 and 7700 - I would expect the 7800X3D to perform as usual. If they can't even handle "65W TDP" CPUs, then that's a story worth covering!
thanks for the in depth review. i assemble PCs for living. your videos always saves me hours of troubleshooting. i'll be recommending the b650m hdv for anything above 120w
Thank you for this absolutely crucial testing. Finding products behaving out of spec and reporting them in an independent and fair way is super necessary.
Good work Steve. Much appreciate the pain you go through to test these products so we don't have to. I'd like to see how the ITX AM5 mobos compare to each other. There's not that many of them and I've a sneaky feeling that they may actually provide better quality parts for the overall package. Not that I know for certain. I'm sure people will drop comments one way or another, I just hope they have facts rather than opinion to back it up :)
So similar things to this are happening on Intel 13th gen with 12th gen boards, especially with XMP RAM speeds and fast boot causing PC not to boot. Tons of BIOS revisions, at least on the Aorus Ultra board I have to try to address some of it, but really nothing much has changed. I'm honestly surprised someone isn't suing at this point over the RAM speed thing. It's almost impossible to get advertised speeds on some boards, even with RAM that has been "verified" to work on those boards.
At my work we regularly have to disappoint customers who want a PC built with 6600Mhz ram, simply because we can't get stable performance over 6200 or 6400Mhz. Sometimes even capping at 6000Mhz before failed Memtests. It's a gamble and no kit is the same. We send these kits back to the manufacturer with our complaint and we do get new sets, but even then it's a game of luck if it actually runs at the advertised speeds.
Intel is always terrible with RAM speeds in my experience. It's always just randomly how fast your cpu can handle while AMD so far could always run the RAM at the XMP speeds.
To be honest this would be good board to look at for budged builds after next gen boards will start to appear, low end boards often get really good discounts at the end of their lifetime.
Not in for an upgrade in the next two years, but I always love this content and thanks for going through with the testing. I'll consider Floatplane for this type of contgent!
I've always had a good experience with Asrock and their boards, although I've always been someone who builds mid-range computers.They added CPU BIOS features to their B450 boards that all other manufacturers only put on their B550 and above.
These are things that people need to know about. I can't believe they updated their website to support less cpus like you wouldn't have checked when buying a budget motherboard. Wow.
The 7950x is back on the list of supported CPUs. Even after this video. No mention of what TDP it will run at, on their site. So much for fixing this. They are back at it.
Reviewers shouldn't be weighed down with the task of QA testing. I'm sorry you have to deal with that Steve. I've been seeing this trend in tech companies for over a decade now. I worked as a sysadmin in a test lab for a tech company in 2010-2016. I saw them whittle down their testing departments repeatedly over the years. In 2010, I worked with over 50 people. By 2016, I worked with 8, and all those were outside the country and not direct employees of the company, outsourced to save money. (I bet they're having a great time testing now, considering their whole team was sourced from Kiev, Ukraine.) More and more companies are sending out untested hardware and software and "fixing" the issues after customers complain. It's lazy and cheap, and their miserly attitudes are hurting the entire industry.
It's not a trend. This stuff gets more and more complicated each generation. On top of that, ASRock is just 70 Engineers - it's not ASUS or Gigabyte - forget about couple of issues here and there, with time they fix it all. It's just they have to eat and sleep sometimes. I have always been impressed by level of technological wizardry ASRock has been able to pull out of their behinds over they years, and despite Platform Integration Documentation - basically a reference board with everything exposed with full documentation - creeping up to 2000 pages from measly 130 in Sandy Bridge era - they still drop insane products with same tenacity as giants. For the slimming of validation, I work myself in it, and I can only say you are wrong. In 2016 product my team works on line had about 330 tests, with 25 people running them in shifts to chase the release dates and bug escape analysis performed at least once per week. Now it's just 7 people, with over 46000 tests, and just few of them requiring manual intervention, everything else is automated. We still get customer reported bugs, but 99% of them are bizarre corner cases or eldritch combinations of hardware and users making up things and stating that 'by their judgement it should work, but why it doesn't?'. Of course things will slide under the radar unnoticed, for something we didn't think to test or due to specific combination of equipment used for testing purposes.
@@Vatharian To make my previous comment clearer: I'm not criticizing the testers. I'm criticizing corporate management for cutting back on testing staff. You even mention this with having previously 330 tests and 25 staff and now you have 7 staff and more tests to run. I'd like to smack certain management upside the head for their lack of sense in cutting back testing staff so much.
Usually I have criticism of the exclusion of MSFS in benchmarks of yours. Thanks for this great review of entry level hardware covering not only high end products Hardware Unboxed! Also thanks for out to the patreons 🖤
Apparently they have not changed. This reminds me of the infamous ASRock Fatality 970 Performance - a AM3+ board with cheaper AMD 970 chipset - supposedly supporting the 220W TDP FX-9590. Two of them simply burned out after a few weeks (a chip near the VRM scorched). That was because the VRM reached over 100° C when putting (simple gaming) stress on the CPU. But in the forum support kept insisting it is supported, but customers have to care for any extra cooling required, namely active VRM cooling. Others got it right without modding, though.
Much appreciated! Way more valuable than pre-release reviews as this shows the actual consumer experience. I know it's frustrating but more like these will make a big difference!
Such reviews are of the most importance not only because push the complacent vendors to fix their product descriptions and firmware, but exposes bad products warning people to dodge a bullet while saving $20.
Lost of talk about VRAM lately but what about DRAM? Steve's last 16GB vs 32GB video was 4 years ago, it would be interesting to see how it compares with modern games
In my anecdotal experience the only game that truly cares about having access to more than 16 gigs is Star Citizen. Now, I realize that that might not “count” as a game yet, and if so, my answer would then change to “none”.
It's baffling to me that ASRock didn't just enable ECO mode by default for the 170W CPUs on this board instead of dropping support. The 7900X and 7950X don't drop all the much performance at a 120W TDP limit, and it would avoid the huge drops from thermal throttling on the VRM. Obviously this would need to be made clear in the product listing and specifications so buyers know what to expect.
It's likely ASRock was just lazy and didn't bother with in-depth testing and optimizing. Once Steve notified them about it they promptly made changes improving performance and clarifying compatibility. That's great! Steve basically improved the product, I think he can be be proud of that.
In my potentially naive opinion, you salvaged your time spent. I hope this video performs well. It was very informative for me and thanks for the insight about what happens behind the scenes as well.
I been convinced that for my next build Asrock Motherboards aren't a bad start. I found this A520 for $200AUS, but I also found a B650M RS PRO for $240AUS but I haven't pulled the trigger becuase CPUs are still to expensive and then thier's buying windows 11, maybe when series 8000 CPUs come out people will start upgrading some cheap 7000.series CPU will come up on ebay.
I disagree with the mindset that nobody would pair a cheap board with an expensive CPU. For me, a board is basically just there to house the actually important hardware. So long as it can power the CPU and have all the connections i need which are not many and doesn't have any other serious issues, i'd always take the cheapest board available.
Thanks for being honest and showing the truth. They just hope people that go to purchase those boards realize what they're getting because most won't watch videos
I recently built am AMD system using ASRock using a mid range mobo chipset and was pleased with the quality and software support. Compared to my much more expensive PC with a mobo from gigabyte, it was a steal and worked far better out of the box
Asrock is entry level, if it was "midrange" it would already be ASUS otherwise they'd be eating into their own sales. Asrock are excellent and often pretty much just a solid ASUS board with a few less features.
As even Steve stated, the b650 from ASRock was a great board, whereas this baseline wasn't. In my experience a midrange ASRock (on sale at $120) gave me a better experience than a top end gigabyte (originally bought at $340).
@@givemeajackson I stand corrected then. They apparently are still owned by the same company but asrock no longer just parent company but a straight competitor. They still kind of own the low end segment.
@@SoftExo no they're not, asrock is owned by pegatron. and asus is the main brand of Asustek. pegatron split off from asustek in 2010, they're completely independent companies.
@@DebasedAnon Depends on the Resolution. 1080p 8GB VRAM would be enough. 1440p you need to lower down the textures to high or medium depending on the game.
@@DeadPhoenix86DP Even in 1080p you need to lower it in newer games... Only reason 8GB was enough was because they were still developing with PS4 in mind, now they're switching to focusing on PS5 and they wont limit VRAM usage anywhere near as much and you can already see that with Hogwarts and how its literally unplayable even at 1080p ultra on a 3070...
@@DebasedAnon Never seen any game go above 8GB on 1080p Resolution. 1440p yes but 1080p no. And the PS5 is using 10GB at best for games, while the rest is reserved for the OS.
7800X3D with a cheap board is for many a perfect combination, it will last a loooong time and thanks to its limited power consumption the need for the high end VRM cooling etc just isn't needed. So is it worth it?
@@JagsP95 I watched the video and was honestly surprised that it was a test that focused more on the motherboard and its shortcoming (who would brings this to market in this state??) with all CPUs above 120 watt. I did expect a 7800X3D focused test on a super cheap board and a conclusion that would say something about it.
@@annoyingguyoninternet1631 It's hard to see how this'd be worth the trouble anyway, since you can get proper 170W support for an extra $20-30. The temps are just fine for the CPUs you should realistically use in a board like this.
I definitely wouldn't use one data point at the high end of a620 motherboards from a flaky motherboard manufacturer to completely look down on the idea of a a620 mixed with a higher end CPU. While $20 isn't worth it, when you start talking about the $85 boards... you're looking at actual trade offs that make sense, such as with the MSI PRO PRO A620M-E.
I do believe it is nothing absurd to see an entry level board not able to run a 170W cpu, as I believe no one would ever use a combo like that. Even if am5 ends up supporting another 3 generations like am4 did, there will always be lower consumption cpu. Even the 7950x itself has the bios option to limit consumption to 95 or 65W in ECO mode
@@Deathscythe91 how much more needed to be said about the 7800X3D? The review is mainly about the this A620 board so the content reflects that. The 7800X3D works fine in this board and he showed that thoroughly.
Nothing but problems with multiple x670 chipset MB's I have used and tested. Zen 4 CPU's are exceptional, truly amazing. The motherboards as a whole are without a doubt, the worst generation I have ever seen. They are FULL of problems. We now have the x3d cpu's frying the sockets of ASUS/MSI mb's.
I know of one case and even if there was a second one it doesn't mean anything in relation to the number of chips sold, time will tell. Also, it's unclear what the problem was, the CPU or the motherboard.
17:58 "look like it has the chance to support parts like the 7950X" Please teach us how to also judge by the looks 😄 In all seriousness, though, when I'm looking for ITX boards "looks" is the only angle I have. And when it comes to power stages I'm lost and would likely pick the wrong board. A "buildzoid light" eduacational piece would be awesome.
As a system builder since the early 2000's, i can 100% concur with your sentiments (i.e. Steve) completely regarding ASRock. I think i first purchased an ASRock board in 2007-'08, knowing fully well that they were a relatively new(b) OEM at the time. Needless to say, there were issues from the outset: from failing drivers (audio, system BUS drivers, etc). At the time, i thought i was "doing it wrong", until my online investigations into AsRock began to expose a pretty dismal record of low quality m/b builds. At any rate, fast forward some 18-20 years, and all i can truthfully say to any would be system builder: DO NOT BUY ASRock m/b's if you can at all afford alternatives. In fairness the overall quality may have improved over the years, but my gut feeling is Steve's experiences here point to the fact they (ASRock) clearly still take shortcuts in the m/b production processes. Spend a few dollars more on (my recommended) brands, such as: ASUS, Gigabyte or MSI, and avoid (most) the hassles you'll get with ASRock.
These A620 motherboards, especially the 120W versions, will make a lot more sense when the R3 7000 CPUs and AM5 APUs roll out. Pairing an A620 motherboard with something like an Athlon 7000G, R3 7100, or R3 7200G would make for a great starter PC or home office system. Right now, I totally agree, there is no market for these things. But really, in most cases, it would make more sense to save up and budget enough money for the cheapest B650 motherboard available.
16:00 Just adding a parenthetical with "up to xxxW" after the "Supports AMD Socket AM5 .... Processors" line on top would go a long way. Then add an asterisk with "*) Processors above xxxW may introduce performance issues" to make it clear they'll run, but poorly.
11:20 I wonder if the 7950X would perform better by manually setting lower power/electrical limits (PPT, EDC, TDC), so it keeps all cores at a lower but steady frequency rather than some of them running full speed and others clocking down way too much.
I remember Asrock doing this same thing with AM3+ and Bulldozer. Initially some boards listed compatibility with 125W processors which were later removed from the support list because they would throttle the 8120. It was very unfortunate for buyers who bought them to pair with the 8120s.
I almost ordered a a620 but I am about to stick a 7700x into a 2u chassis and not having voltage control can get messy real fast. So I got a Gigabyte B650I but these boards a crazy expensive.
@@hyperturbotechnomike Yes, I gave up on that plan (2U + 7700X) for anything powerful. I have a 4U in my home rack with a 5900X and the Noctua NH-D12L which works very well. At my office I did a 2U with Ryzen 5600 and the wraith stealth cooler fits in there comfortably, it runs as a Windows file server and FreePBX in Hyper-V. I tried to run a Noctua NH-L12S on the 7700X and it was crash city even at very low clock speeds, I guess it needs a minimum amount of voltage, but heat kills ask any Intel owner, lol.
Looks like a perfectly suitable board for a budget build, its now 37% cheaper than the b650. Testing it with a 7950x in cinebench was nonsense. 7800x3d or 7600(non-x) fps testing vs a b650 would have been much more valuable, as they much closer fit its use case /pricepoint.
Good to see that Asrock, at least, reply to emails. On the other side you have Gigabyte, which only replied to me with a copy-pasted message when I asked help with their faulty B650-DSH3. And the worst thing, after 1 month I'm still waiting to resolve my RMA.
4 months are passed and A620m HDV/M. 2+ performance is reflected in price, now it costs around 25% less than B650M Hdv here in Europe and the difference is even more in US 79$ vs 125$
Well here in Germany we have to pay 117 Euros for the A620 + and same price for non + and 156 Euros for the A650, so its quite a difference that you could consider the A620+ with a 7600 I guess
I learned the hard way to avoid the low end ASRock boards esp the HDV models when I built my first Ryzen system. There were so many little issues that were remedied when I went with something of higher quality and my system was actually more stable.
"no one is going to pair a 7950x with a cheap mobo"
Amazon prebuilts: Allow me to introduce myself
So accurate
most prebuilt duing that including asus, dell, etc
Any and all prebuilts, along with cheap RAM and cheap PSU
@@rattlehead999 insane the margins they must get selling $1000 components for $1700
@@rattlehead999 Cheap SINGLE channel ram, with poor speeds 😂
Loads of testing of Ryzen 7000X cpus in "ECO mode" have already established that these CPUs are capable of great performance, even when TDP limited well below 120W. What on earth is Asrock doing to make the cpu panic throttle to 544mhz?
Asrock,
@@mrfaern if thats the case then Asus is better off to not exist.
The budget ASRock boards only exist as an exercise in getting familiar with the RMA process.
@@creativityfails1 that's funny
@@creativityfails1 😆👌
0:00 - Intro
1:51 - Sponsor - Gigabyte Aorus Laptops
2:47 - AM5 platform review
5:45 - Cinebench R23 -- Testing [Ryzen 9 7950X]
6:38 - Cinebench R23 -- Testing [Ryzen 7 7700X]
7:02 - Cinebench R23 -- Testing [Ryzen 5 7600X]
7:13 - Cinebench R23 -- Testing [Ryzen 9 7900]
7:25 - SoTR gameplay [Ryzen 9 7900]
8:12 - Contacting Asrock and further troubleshooting
8:52 - Vendor contact, new motherboard BIOS
9:08 - Revocation of 170w TDP support and resolution
10:27 - Benchmark - Cinebench R23
11:57 - Benchmark - Cinebench R23 (10m loop)
12:07 - Benchmark - SoTR
12:25 - Benchmark - Watchdogs Legion
12:45 - VRM thermals - 1 hour test
13:31 - VRM thermals - comparison [Ryzen 9 7950X]
13:52 - Benchmark - comparison CB R23 [Ryzen 9 7950X]
14:07 - Comparison between B650M-HDV/M.2
15:40 - Conclusion & “Caveat Emptor” between this board and non-plus variant
Great work. Don't underestimate the importance of reviewing bad - ahem (sorry) limited - products like this motherboard. Now we are informed and advised. Keep up the good work.
i still have my hopes up for other A620 boards. at least to Buildzoid levels of success...
I would watch a video on the suspicious Asus and MSI boards
Funny how bios fixes only take 3 days when a journalist reports a problem, but regular people wait weeks or month for fixes on reported problems.
That's the reason the news media is known as "the fourth power"
Well said, when making incremental upgrades on budget different ideas come to mind.
I consider Asrock is already doing a good job by releasing a BIOS fix in just 3 days after Steve's report. Somehow, that shows Asrock was taking input from external source seriously. They were wrong, but at least Asrock managed to correct them.
it does seem like they were working on it already, 3 days seems tight for such things..
@@yukisnoww Yep, his feedback most likely got it pushed up the priority list...they know he will have a review
It would have been a fairly straight forward fix, change the power limit and raise the temp the VRM can hit, or it could have just been the latter. They've released another revision since then.
@@Hardwareunboxed i see, but the timeline for straightforward fix can be anywhere from couple minutes to a month 🤣
Not when they know a video is coming ;)
I can’t believe Asrock released this board in that state, Steve doing their darn R&D product testing for them fs. Man, good work exposing this and getting it fixed Steve!
You would think these mb companies would at least test configurations before putting out stated information, but no, we as consumers have to beta test all technology's these days. Keep up the benching!
What do you mean these days? Its way worse back in the day lol
Its why i’m still AM4. I’ll let everyone else beta test a while longer before I jump in.
@@mikem2253 and in that time prices goes down, I see this as a win
@@HSG4meR Will the prices go down, or will the profits go up. That's the real question.
@@mikem2253 I feel you! I used to buy to newest CPU/MB every 2 years, but it has bit me in the ass so many times in the last several years! I'm sticking with I have for the next 4 years at least and then I buy whatever chip is a generatio old at that time...I'm tired of beta testing $1000 CPU/MB for these companies!
Thanks for Steve’s effort for making Asrock’s product better. They should definitely pay the QC officer’s salary to you instead!
What QC?
Sorry, I just had to.
@@PQED quality control
@@patlotpotlot6154 i was being facetious (joking).
Don't worry about it. :)
I think thats something that most brands have problems with, new tech! Always something new to tweak.. you can see Asus having issues of their own!
Asrock improve thanks to the Engineer and people Freedback in the forum. Steve did good job, but all improve is beacause customers.
Thanks for your excellent fast review as always Steve. Very informative. And I've come to watch it fast to show my respect again!
I think labeling CPU max wattage support for motherboards is really useful. Researching for it on the internet is bloody brutal.
Normal it have .
Online you can see it
My b450m msi have a 95watt tdp beult.
Bit still i can use a 5800 serie on it
And those cpu are 105watt tdp.
Bios have it but the performance is just not thare.
Funny thing is that automatic overklok on the x serie is not support.
It would make sense, if CPUs actually only pulled as much power as their "TDPs" say they do.
You can usually tell, just by looking at it:
- No heatsink on the VRMs: up to 65W with tower coolers and 95W with a topdown cooler
- With heatsink on the VRMs: up to 95W with tower coolers and 125W with a big topdown cooler
For anything beyond 125W, count the chokes: as a rule of thumb, each choke can easily handle 15W if cooled in a topdown manner or 10W if cooled by an AIO or a tower cooler. So basically for 170W, you should ideally have at least 12 chokes in total.
@@sassukiwhere did you obtain this information?
I hope you guys or someone else makes more videos on a620 boards. I know they will be commonplace for budget builds in the future. My used a320 board works well with my Ryzen 5600, and I managed to snag it for only $30 several months ago!
I wish I saw some testing on the limits of the 65W boards. The "65W TDP" Ryzen 7000 CPUs actually consume about 90W without PBO on all core load according to GN. The "120W" 7800X3D is around the same power consumption at 90W on all core load, so if the boards are fine with a 7600 and 7700 - I would expect the 7800X3D to perform as usual. If they can't even handle "65W TDP" CPUs, then that's a story worth covering!
thanks for the in depth review. i assemble PCs for living. your videos always saves me hours of troubleshooting. i'll be recommending the b650m hdv for anything above 120w
Great reviews, congratulations, you gained a subscriber from Brazil.
Thank you for this absolutely crucial testing. Finding products behaving out of spec and reporting them in an independent and fair way is super necessary.
AsRock HDV back on top of their game LOL
Great video Steve. Way to be on top of these board manufactures.
Good work Steve. Much appreciate the pain you go through to test these products so we don't have to.
I'd like to see how the ITX AM5 mobos compare to each other. There's not that many of them and I've a sneaky feeling that they may actually provide better quality parts for the overall package. Not that I know for certain. I'm sure people will drop comments one way or another, I just hope they have facts rather than opinion to back it up :)
ITX boards are usually a good bit more expensive and of a good quality because ITX is a niche people will pay a premium for.
So similar things to this are happening on Intel 13th gen with 12th gen boards, especially with XMP RAM speeds and fast boot causing PC not to boot. Tons of BIOS revisions, at least on the Aorus Ultra board I have to try to address some of it, but really nothing much has changed. I'm honestly surprised someone isn't suing at this point over the RAM speed thing. It's almost impossible to get advertised speeds on some boards, even with RAM that has been "verified" to work on those boards.
At my work we regularly have to disappoint customers who want a PC built with 6600Mhz ram, simply because we can't get stable performance over 6200 or 6400Mhz. Sometimes even capping at 6000Mhz before failed Memtests. It's a gamble and no kit is the same. We send these kits back to the manufacturer with our complaint and we do get new sets, but even then it's a game of luck if it actually runs at the advertised speeds.
Intel is always terrible with RAM speeds in my experience. It's always just randomly how fast your cpu can handle while AMD so far could always run the RAM at the XMP speeds.
@@tilapiadave3234 Both have problems, dont mention Asus history in the crosshair formula in Iintel and amd
To be honest this would be good board to look at for budged builds after next gen boards will start to appear, low end boards often get really good discounts at the end of their lifetime.
Only for the Ryzen 5 7600 and any future entry-level AM5 CPUs though. Even if it's cheap, there's no point in wasting your CPU's potential.
7500F runs fine on A620. I have one in a microserver.
Not in for an upgrade in the next two years, but I always love this content and thanks for going through with the testing. I'll consider Floatplane for this type of contgent!
I've always had a good experience with Asrock and their boards, although I've always been someone who builds mid-range computers.They added CPU BIOS features to their B450 boards that all other manufacturers only put on their B550 and above.
Same here.
Thanks for the great work with testing these boards! Looking forward to the Gigabyte MK one.
These are things that people need to know about. I can't believe they updated their website to support less cpus like you wouldn't have checked when buying a budget motherboard. Wow.
The 7950x is back on the list of supported CPUs. Even after this video. No mention of what TDP it will run at, on their site. So much for fixing this. They are back at it.
Reviewers shouldn't be weighed down with the task of QA testing. I'm sorry you have to deal with that Steve. I've been seeing this trend in tech companies for over a decade now. I worked as a sysadmin in a test lab for a tech company in 2010-2016. I saw them whittle down their testing departments repeatedly over the years. In 2010, I worked with over 50 people. By 2016, I worked with 8, and all those were outside the country and not direct employees of the company, outsourced to save money. (I bet they're having a great time testing now, considering their whole team was sourced from Kiev, Ukraine.) More and more companies are sending out untested hardware and software and "fixing" the issues after customers complain. It's lazy and cheap, and their miserly attitudes are hurting the entire industry.
It's not a trend. This stuff gets more and more complicated each generation. On top of that, ASRock is just 70 Engineers - it's not ASUS or Gigabyte - forget about couple of issues here and there, with time they fix it all. It's just they have to eat and sleep sometimes. I have always been impressed by level of technological wizardry ASRock has been able to pull out of their behinds over they years, and despite Platform Integration Documentation - basically a reference board with everything exposed with full documentation - creeping up to 2000 pages from measly 130 in Sandy Bridge era - they still drop insane products with same tenacity as giants.
For the slimming of validation, I work myself in it, and I can only say you are wrong. In 2016 product my team works on line had about 330 tests, with 25 people running them in shifts to chase the release dates and bug escape analysis performed at least once per week. Now it's just 7 people, with over 46000 tests, and just few of them requiring manual intervention, everything else is automated. We still get customer reported bugs, but 99% of them are bizarre corner cases or eldritch combinations of hardware and users making up things and stating that 'by their judgement it should work, but why it doesn't?'. Of course things will slide under the radar unnoticed, for something we didn't think to test or due to specific combination of equipment used for testing purposes.
@@Vatharian To make my previous comment clearer: I'm not criticizing the testers. I'm criticizing corporate management for cutting back on testing staff. You even mention this with having previously 330 tests and 25 staff and now you have 7 staff and more tests to run. I'd like to smack certain management upside the head for their lack of sense in cutting back testing staff so much.
Google started this with Android and everybody followed.
Awesome piece! Thanks for the time invested Steve.
I hope Asrock is grateful for your findings, you brought a huge issue to their attention
Usually I have criticism of the exclusion of MSFS in benchmarks of yours. Thanks for this great review of entry level hardware covering not only high end products Hardware Unboxed! Also thanks for out to the patreons 🖤
Apparently they have not changed. This reminds me of the infamous ASRock Fatality 970 Performance - a AM3+ board with cheaper AMD 970 chipset - supposedly supporting the 220W TDP FX-9590.
Two of them simply burned out after a few weeks (a chip near the VRM scorched). That was because the VRM reached over 100° C when putting (simple gaming) stress on the CPU.
But in the forum support kept insisting it is supported, but customers have to care for any extra cooling required, namely active VRM cooling. Others got it right without modding, though.
That comparison shot of the dead guy on quake and cyber punk in the same pose was legendary!!!
My take from this review is that the 7800X3D is a god among CPUs. Good job AMD. Maybe next time release the successor sooner ?
Great video, thank you for dealing with all the issues and testing it out!
HUB being migrated from the ASRock blacklist to the BETA testers group is at least a move in the right direction.
Sadly though, it was a paid program, and we were the ones paying.
@@Hardwareunboxed It's as if they didn't learn from their previous mistakes and thought no one would notice. Continue keeping them honest.
Good info man thanks a lot
Much appreciated! Way more valuable than pre-release reviews as this shows the actual consumer experience. I know it's frustrating but more like these will make a big difference!
Such reviews are of the most importance not only because push the complacent vendors to fix their product descriptions and firmware, but exposes bad products warning people to dodge a bullet while saving $20.
Lost of talk about VRAM lately but what about DRAM? Steve's last 16GB vs 32GB video was 4 years ago, it would be interesting to see how it compares with modern games
In my anecdotal experience the only game that truly cares about having access to more than 16 gigs is Star Citizen. Now, I realize that that might not “count” as a game yet, and if so, my answer would then change to “none”.
@@124thDragoonThen you haven't met my good old friend... 7 Days to Die. It sucks DRAM like no other 😅😂
@@xalderin3838 the last time I played that game was ~6 years ago. It was fine at the time. Weird.
@@124thDragoon Ahh, so you haven't played any recent updates.
I have 32gb of Ran, and it sucks about 25+ relatively easily, when I play.
@@xalderin3838 Jesus Christ
Always asking the right questions, awesome!
It's baffling to me that ASRock didn't just enable ECO mode by default for the 170W CPUs on this board instead of dropping support. The 7900X and 7950X don't drop all the much performance at a 120W TDP limit, and it would avoid the huge drops from thermal throttling on the VRM.
Obviously this would need to be made clear in the product listing and specifications so buyers know what to expect.
Thank you Hardware Unboxed for keeping Asrock honest. Reviews like this greatly influence my buying decisions.
It's likely ASRock was just lazy and didn't bother with in-depth testing and optimizing. Once Steve notified them about it they promptly made changes improving performance and clarifying compatibility. That's great! Steve basically improved the product, I think he can be be proud of that.
Thank you for your time on this subject.
Thank you for testing the boards at the budget end of the scale.
In my potentially naive opinion, you salvaged your time spent. I hope this video performs well. It was very informative for me and thanks for the insight about what happens behind the scenes as well.
I'm just glad that the board is in a working state now. It's good to know that one of the lowest budget boards can properly run a 7950X3D
Thank you for making this video. Great content!
1:43 oh wow, I thought Asrock had blacklisted HUB?
We're back in their good books this year ;)
Thats nice, all Asrock have to do is put out good products and do away with the marketing BS and they'd get good reviews from HUB anyway
@@Hardwareunboxed I can tell by first minute thats not gonna be long lasting 😅
I been convinced that for my next build Asrock Motherboards aren't a bad start. I found this A520 for $200AUS, but I also found a B650M RS PRO for $240AUS but I haven't pulled the trigger becuase CPUs are still to expensive and then thier's buying windows 11, maybe when series 8000 CPUs come out people will start upgrading some cheap 7000.series CPU will come up on ebay.
I disagree with the mindset that nobody would pair a cheap board with an expensive CPU. For me, a board is basically just there to house the actually important hardware. So long as it can power the CPU and have all the connections i need which are not many and doesn't have any other serious issues, i'd always take the cheapest board available.
Very useful information and is appreciated!
What if we undervolt and apply PPT edc values ? for 7950x
Thanks for being honest and showing the truth. They just hope people that go to purchase those boards realize what they're getting because most won't watch videos
I recently built am AMD system using ASRock using a mid range mobo chipset and was pleased with the quality and software support. Compared to my much more expensive PC with a mobo from gigabyte, it was a steal and worked far better out of the box
Asrock is entry level, if it was "midrange" it would already be ASUS otherwise they'd be eating into their own sales. Asrock are excellent and often pretty much just a solid ASUS board with a few less features.
@@SoftExo ah yes, the 550 dollar entry level asrock x670 taichi...
asrock hasn't been part of Asus since 2002.
As even Steve stated, the b650 from ASRock was a great board, whereas this baseline wasn't. In my experience a midrange ASRock (on sale at $120) gave me a better experience than a top end gigabyte (originally bought at $340).
@@givemeajackson I stand corrected then. They apparently are still owned by the same company but asrock no longer just parent company but a straight competitor. They still kind of own the low end segment.
@@SoftExo no they're not, asrock is owned by pegatron. and asus is the main brand of Asustek. pegatron split off from asustek in 2010, they're completely independent companies.
Thank you for the warnings!
i have an msi rtx 3070 gaming x and 5800x.. i think im good..
Not for long with the 3070...
@@DebasedAnon Depends on the Resolution. 1080p 8GB VRAM would be enough. 1440p you need to lower down the textures to high or medium depending on the game.
@@DeadPhoenix86DP Even in 1080p you need to lower it in newer games...
Only reason 8GB was enough was because they were still developing with PS4 in mind, now they're switching to focusing on PS5 and they wont limit VRAM usage anywhere near as much and you can already see that with Hogwarts and how its literally unplayable even at 1080p ultra on a 3070...
@@DebasedAnon Never seen any game go above 8GB on 1080p Resolution. 1440p yes but 1080p no. And the PS5 is using 10GB at best for games, while the rest is reserved for the OS.
nah 8 gb is still good. those games are just unoptimized. and people are blowing it pjt of proportion.. u can still do good with 8gbs..
that's a new and different review for motherboards you have been tested. Good job
could you take a look at the recent fiasco over asus boards burning 7800x3d cpu's?
This is news to me. What is your source for this please?
@@Squashed8Ballit’s a popular thread in the Amd subreddit. Gamers nexus offered the person to purchase it from the person who’s board and cpu died
@@Squashed8Ball On videocardz you have article about it
@@adamadamx5464 the German did a video he does speak English
That's how rumors start, we know of 1 (ONE) case and now it's already "asus boards burning 7800x3d cpu's".
THANK YOU!!!! I’d been considering the A620 board for a cheap build
7800X3D with a cheap board is for many a perfect combination, it will last a loooong time and thanks to its limited power consumption the need for the high end VRM cooling etc just isn't needed.
So is it worth it?
Why don't you watch the video first lmao
Yes
@@JagsP95 I watched the video and was honestly surprised that it was a test that focused more on the motherboard and its shortcoming (who would brings this to market in this state??) with all CPUs above 120 watt. I did expect a 7800X3D focused test on a super cheap board and a conclusion that would say something about it.
Great video, thanks for the review.
Perhaps adding a heatsink and even a fan could improve vrm temps?
It already have a heatsink so adding would be hard.
@@annoyingguyoninternet1631 It's hard to see how this'd be worth the trouble anyway, since you can get proper 170W support for an extra $20-30. The temps are just fine for the CPUs you should realistically use in a board like this.
@@ozzyp97 absolutely this and A series traditionally drop prices down to $50 overtime
I definitely wouldn't use one data point at the high end of a620 motherboards from a flaky motherboard manufacturer to completely look down on the idea of a a620 mixed with a higher end CPU. While $20 isn't worth it, when you start talking about the $85 boards... you're looking at actual trade offs that make sense, such as with the MSI PRO PRO A620M-E.
You're right , enough for A620 . We need more B650 and X670 hardware unboxed review .
No! I'm waiting for gigabyte a620 boards
I appreciate the testing.
I do believe it is nothing absurd to see an entry level board not able to run a 170W cpu, as I believe no one would ever use a combo like that. Even if am5 ends up supporting another 3 generations like am4 did, there will always be lower consumption cpu. Even the 7950x itself has the bios option to limit consumption to 95 or 65W in ECO mode
No, but it is absurd to claim that it does in fact do that, when it actually doesn't.
@@memitim171 sure, I agree. At least now it is correct
It's refreshing to see NewEgg being a good guy for once. Well done.
so bad that it pulverized the oc world records on a 7950x MDR
Steve thanks for taking a lot of time to provide info for average buyers.
Interesting to hear how much Asrock apparently tried to gaslight you
Great work mate! Fighting the good fight!
video was barely about the 7800x3d
It's called motherboard review.
@@Smallman647 then dont name the title 7800x3d +a620 is it worth it?
Makes it sound like the video is about testing the x3d chip on a620??
@@Deathscythe91 how much more needed to be said about the 7800X3D? The review is mainly about the this A620 board so the content reflects that. The 7800X3D works fine in this board and he showed that thoroughly.
Thanks to your tests, I was able to decide on the choice of motherboard for AM4. And I'm not the only one
Nothing but problems with multiple x670 chipset MB's I have used and tested. Zen 4 CPU's are exceptional, truly amazing. The motherboards as a whole are without a doubt, the worst generation I have ever seen. They are FULL of problems. We now have the x3d cpu's frying the sockets of ASUS/MSI mb's.
I know of one case and even if there was a second one it doesn't mean anything in relation to the number of chips sold, time will tell. Also, it's unclear what the problem was, the CPU or the motherboard.
17:58 "look like it has the chance to support parts like the 7950X" Please teach us how to also judge by the looks 😄 In all seriousness, though, when I'm looking for ITX boards "looks" is the only angle I have. And when it comes to power stages I'm lost and would likely pick the wrong board. A "buildzoid light" eduacational piece would be awesome.
Just look at the components on the board ;)
Asrock makes the worst hardware. Period.
My AB350 PRO4 from Asrock is serving me well. Upgraded the cpu from R5 1600 to 5800x3D and it's still working perfectly.
Yep Asrock is trash.
That's not true although they do seem to be rather inconsistent but Steve himselv said their B650 is the best in that class.
Surprisingly Asus entry level ones are worst on b650
@@skyryudo311 Not suprising to me, tbh. Asus always seems to be lacking with AMD hardware.
I'm going to upgrade to AM5 this summer so I'll be eagerly waiting for your AM5 motherboard reviews!
Ok 🥲
As a system builder since the early 2000's, i can 100% concur with your sentiments (i.e. Steve) completely regarding ASRock.
I think i first purchased an ASRock board in 2007-'08, knowing fully well that they were a relatively new(b) OEM at the time. Needless to say, there were issues from the outset: from failing drivers (audio, system BUS drivers, etc). At the time, i thought i was "doing it wrong", until my online investigations into AsRock began to expose a pretty dismal record of low quality m/b builds.
At any rate, fast forward some 18-20 years, and all i can truthfully say to any would be system builder: DO NOT BUY ASRock m/b's if you can at all afford alternatives. In fairness the overall quality may have improved over the years, but my gut feeling is Steve's experiences here point to the fact they (ASRock) clearly still take shortcuts in the m/b production processes.
Spend a few dollars more on (my recommended) brands, such as: ASUS, Gigabyte or MSI, and avoid (most) the hassles you'll get with ASRock.
Glad to see this time ASRock responded adequately! There is hope.
These A620 motherboards, especially the 120W versions, will make a lot more sense when the R3 7000 CPUs and AM5 APUs roll out. Pairing an A620 motherboard with something like an Athlon 7000G, R3 7100, or R3 7200G would make for a great starter PC or home office system. Right now, I totally agree, there is no market for these things. But really, in most cases, it would make more sense to save up and budget enough money for the cheapest B650 motherboard available.
This was great, thank you.
❤ Many thanks just me from making a costly choice. Great A620 review.
Tests like this one are the absolute gold standard, well done!
3:08 +XD! No timestamps means Steve is simply expressing his frustration, XD! Thumbs up!
Shout-outs to all people writing an Email to support about issues, you really help everyone 👍
16:00 Just adding a parenthetical with "up to xxxW" after the "Supports AMD Socket AM5 .... Processors" line on top would go a long way. Then add an asterisk with "*) Processors above xxxW may introduce performance issues" to make it clear they'll run, but poorly.
11:20 I wonder if the 7950X would perform better by manually setting lower power/electrical limits (PPT, EDC, TDC), so it keeps all cores at a lower but steady frequency rather than some of them running full speed and others clocking down way too much.
Then it would still be hobbling the 7950x performance.
Excellent work!
I remember Asrock doing this same thing with AM3+ and Bulldozer. Initially some boards listed compatibility with 125W processors which were later removed from the support list because they would throttle the 8120. It was very unfortunate for buyers who bought them to pair with the 8120s.
Great review, as always.
I almost ordered a a620 but I am about to stick a 7700x into a 2u chassis and not having voltage control can get messy real fast. So I got a Gigabyte B650I but these boards a crazy expensive.
@@hyperturbotechnomike Yes, I gave up on that plan (2U + 7700X) for anything powerful. I have a 4U in my home rack with a 5900X and the Noctua NH-D12L which works very well. At my office I did a 2U with Ryzen 5600 and the wraith stealth cooler fits in there comfortably, it runs as a Windows file server and FreePBX in Hyper-V. I tried to run a Noctua NH-L12S on the 7700X and it was crash city even at very low clock speeds, I guess it needs a minimum amount of voltage, but heat kills ask any Intel owner, lol.
Good stuff!, thanks.
For some reason i find it always interesting when someone is describing computer problem they did go trough. Its very relatable.
Looks like a perfectly suitable board for a budget build, its now 37% cheaper than the b650. Testing it with a 7950x in cinebench was nonsense. 7800x3d or 7600(non-x) fps testing vs a b650 would have been much more valuable, as they much closer fit its use case /pricepoint.
Good to see that Asrock, at least, reply to emails. On the other side you have Gigabyte, which only replied to me with a copy-pasted message when I asked help with their faulty B650-DSH3. And the worst thing, after 1 month I'm still waiting to resolve my RMA.
4 months are passed and A620m HDV/M. 2+ performance is reflected in price, now it costs around 25% less than B650M Hdv here in Europe and the difference is even more in US 79$ vs 125$
I think that I bought the last one on Newegg😭
Well here in Germany we have to pay 117 Euros for the A620 + and same price for non + and 156 Euros for the A650, so its quite a difference that you could consider the A620+ with a 7600 I guess
Yeah here the a620 goes for 95€ , while b650 goes for 165-170 xD its a big difference
You saved a lot of people who got this mobo
I learned the hard way to avoid the low end ASRock boards esp the HDV models when I built my first Ryzen system. There were so many little issues that were remedied when I went with something of higher quality and my system was actually more stable.
Thank you, Steve. Your sacrifice will not be forgotten.
Thanks for the top tier content 👍 Can you please do the midrange b650 boards as well, Steve 😭😭