AMD Ryzen 9 7900X CPU Review & Benchmarks vs. R7 1700, 7950X, & More
HTML-код
- Опубликовано: 15 июл 2024
- Sponsor: HYTE Y60 computer case on Amazon - geni.us/3ngWP
Our review & benchmarks of the AMD R9 7900X CPU, a new 12-core part, include tests vs. the R9 7950X, Ryzen 5 7600X, and all the older parts. The R9 5900X, R9 3900X, and R7 1700 are all featured in these benchmarks of some of the best CPUs in 2022 (so far, anyway). Intel still has 13th Gen on the way, but for now, we tested the R9 7900X vs. the i9-12900K and i7-12700K, alongside others.
Grab a GN Anti-Static Modmat for PC Building, a Tear-Down Toolkit, a 3D PC-themed coaster pack, or other unique item on the GN store while also helping fund our next tests: store.gamersnexus.net/
Watch our AMD R7 7700X CPU review: • AMD Ryzen 7 7700X 8-Co...
Watch our AMD R9 7950X CPU review: • 95°C is Now Normal: AM...
Watch our AMD R5 7600X CPU review: • AMD's R5 Mistake: Ryze...
And general specs on the AMD Zen 4 CPUs: • AMD Ryzen 7950X, 7900X...
RELATED PRODUCTS [Affiliate Links]
AMD Ryzen 9 7950X on Amazon: geni.us/6X1RO
Intel Core i9-12900K on Amazon: geni.us/YAlrh9
AMD Ryzen 9 7900X on Amazon: geni.us/FIyuhcS
AMD Ryzen 7 7700X on Amazon: geni.us/udIJ
AMD Ryzen 5 7600X on Amazon: geni.us/6AdNkw
Intel Core i7-12700K on Amazon: geni.us/MP2tSl
Like our content? Please consider becoming our Patron to support us: / gamersnexus
TIMESTAMPS
00:00 - AMD Ryzen 9 7900X CPU Review & Benchmarks
01:10 - Recapping the Ryzen 7000 Important Details
01:53 - 7900X Thermals & Frequency
04:00 - Core-to-Core Deltas
04:35 - Single-Core Boosting Frequency 7900X
04:52 - R9 7900X Power Consumption Benchmarks
06:28 - Production Benchmarks
07:15 - Blender Rendering Benchmarks (7900X Cycles)
08:29 - Power Efficiency Testing (7900X vs. 12900K & 5900X)
09:58 - CPU Efficiency Chart (7900X, 12700K, 3700X, etc.)
10:38 - Chromium Code Compile Benchmarks (R9 7900X Compile Time)
11:03 - Compression & Decompression Best CPUs
12:44 - Adobe Premiere CPU Comparison (7900X Benchmarks)
13:29 - Adobe Photoshop Best CPUs
13:58 - CSGO CPU Benchmarks (1080p, 1440p on R9 7900X)
15:14 - Discussion of Gaming Results So Far
16:50 - Rainbow Six Siege CPU Benchmarks
18:20 - Far Cry 6 CPU Benchmarks
19:31 - Tomb Raider CPU Benchmarks - 1080p, 1440p
20:30 - F1 2022 CPU Benchmarks
21:50 - Cyberpunk is GPU-Bound and Useless Here
22:24 - Conclusion
** Please like, comment, and subscribe for more! **
Links to Amazon and Newegg are typically monetized on our channel (affiliate links) and may return a commission of sales to us from the retailer. This is unrelated to the product manufacturer. Any advertisements or sponsorships are disclosed within the video ("this video is brought to you by") and above the fold in the description. We do not ever produce paid content or "sponsored content" (meaning that the content is our idea and is not funded externally aside from whatever ad placement is in the beginning) and we do not ever charge manufacturers for coverage.
Follow us in these locations for more gaming and hardware updates:
t: / gamersnexus
f: / gamersnexus
w: www.gamersnexus.net/
Steve Burke: Host, Writing, Test & Lead
Patrick Lathan: Testing, Research
Andrew Coleman: Video Production
Mike Gaglione: Testing, Editing
Jeremy Clayton: Additional Writing
Robert Hoang: Product Footage 
Игры
Watch our AMD R9 7950X CPU review: ruclips.net/video/nRaJXZMOMPU/видео.html
Watch our AMD R5 7600X CPU review: ruclips.net/video/JM-twyjfYIw/видео.html
And general specs on the AMD Zen 4 CPUs: ruclips.net/video/hVnJbiYOCq4/видео.html
Grab a GN Anti-Static Modmat for PC Building, a Tear-Down Toolkit, a 3D PC-themed coaster pack, or other unique item on the GN store while also helping fund our next tests: store.gamersnexus.net/
Just one note. On workloads where lower is better, its a bit misleading to say how much faster it is in %, since its calculated differently than the usual higher is better workloads. For instance in the chromium compile benchmark you have 7900x finishing in 46.5 seconds and the 5900x finishing in 61.7 seconds. You said this is a 25% improvement because 1-46.5/61.7 = 0.25. However if you instead calculate how much each cpu can do per second you get 100/61.7 = 1.62 and 100/46.5 = 2.15 respectively. Which means the 7900x does 32% more work per second. I think its fine to calculate the improvement in terms of reduced compile time, however the problem is that the different benchmarks are calculated differently.
Good morning! Thank you for the detailed reviews. When will the AMD R7 7700X review be published?
Steve motherboards PLEEEEEEEEEEEEEEEEEEEEEEEEEEEEEASE!!!!!!!!!!!!!!!!!!!!!
With regards to things like source engine and it's stuttering and older titles etc.. I wonder given the how high the frame rate is on average if it might be worth looking into a even smaller fractional percentage than 0.1% lows? I'm not super familiar with CS:GO, but just in general with that old engine.. it has a lot of stutter, and it seems to be not so much brute forcible.. but if it could be I would be interested just conceptually in that as it seems since efficiency is kind of out the window in this regard.. and efficiency was never going to get you good performance on an engine not designed around the last 15 years of CPU concepts comparatively obviously.. if anything the rainbow six scores next to the cs go scores given the graphical Fidelity and internal complexity of both, the fact that rainbow six is almost double while being computationally more complex to a degree.. I mean that's that's just a good marker in general as to how much you're giving to source engine for what you're getting.
But those stutters, the more you're doing things that the engine is capable of but wouldn't have been coded for in modern terms.. those are well for one they're far more interesting than anything like a seat a game of cs go, which is the most stripped down of source engine source engineiness.. outside of the visuals of course.. and even in CS:GO those stutters are brutal and usually consistent throughout rather than the ones that are first time loads etc, although it has both.. that's why I'm interested possibly in in a further low marker than 0.1% and I wonder if that might actually apply to other things as far as the actual noticeable stutters.. when the frame rate is that high I feel like the percentage might actually be lower as to what is actually perceivable given that on a good day, if I can get my minimum frame time to not drop to below 30 milliseconds I'm having the best day ever or not doing anything special it seems LOL.
LOL the computer yeah yeah she's had a lot of lot of socketing.. not something you need to worry about she's a professional... LOL
Would like to see some ECO mode testing. Apparently you can get 90% of the performance at just half the power.
Is this the function using PBO/curve optimizer for lower voltages/temps? I've seen a few mentions of that getting 10-15 degrees lower under full load on a 7950X
Same here. I like to run things cooler and much quieter.
That would be required to make Ryzen 7000 be of any interest
He did say there'll be a dedicated video on ECO mode
@@ultravisitors thats good to hear
Steve and GN are absolutely smashing the early release of these videos. Making a big impact for consumer awareness on launch day 😁👍
@@glitter_fart And I NEED IT!!!! (Spongebob)
@@glitter_fart I have both the 3600 and the 5600X, but I am going to upgrade the 3600 cause it is so much weaker than the 5600X, so it's either going to be a new 5600X or 12600K or maybe 13600k, so I'll decide after the intel 13 tests are ready in the near future.
Launch day? Where the fuck are they then?
@@zzz4854 PINKY. OUT.
I was saddened when HardOCP shut down, but GN are a great replacement.
Would love to see a detailed video on the performance of the ECO modes. Centered around something like the absolute draw under each eco mode and the CPU efficiency chart with the various 7000 series chips under their eco modes (alongside a smaller selection of the benchmarks like the Blender rendering test to monitor absolute performance loss).
Checkout Wendell's Linux vid: ruclips.net/video/LENj4uxjnRs/видео.html
Basically it's INSANELY effective if you lower the wattage. I wanted to buy a laptop this month, but I am certainly waiting for Zen 4 laptops!
I also like to see a total kWh consumed by several systems running some light (home type) server workloads, a few hours of gaming, and basically idle overnight. What is the balanced power consumption for a person that leaves one system on 24x7 and also uses it a few hours a day.
@@kwrzesien17 Yeah would be interesting. PCWorld kinda touch on it a little near the end of their video yesterday but an expansion on it by someone would be cool (doesn't have to be GN doing all parts of it).
ruclips.net/video/uks4qQ2MXrM/видео.html&ab_channel=PCWorld
@@alastairtemple Thank you, that was exactly what I needed!
Me too. The tdp seams to be high since these cpu basically are very smart at... overclocking themselves. It would be cool to see how they fare when you try to find a more reasonable balance. I suspect you waste quite a lot of power for relatively small gains letting them go full speed
“Although most people will be bottlenecked on skill before they’re bottlenecked on the CPU.”
- Steve Burke 2022 (16:45)
Savage 😂😂😞✌️
🔥
9-1-1, yeah, I'd like to report a murder!
So 5800X3D users have been enjoying Zen 4 top tier performance (and even higher) all this time? Such a beast gaming CPU, 1080 Ti legendary tier.
Thats the power of cache. Zen 43D will be just as insane.
$300 open box on ebay right now, makes more sense then dishin out $1000 for new ram and motherboard and cpu. Hope they keep making it and it hits $250 some day...
I just bought a 12700k and ddr5 5600. I want this computer to last 6-8 years, so I wanted to invest in something that will have an advantage on things we don’t really need right now like faster ram bandwidth. Maybe I was wrong and should have bought a 5800x3d. 🤷🏼♂️ I’m still pumped. Shows up on Wednesday. Mildly nervous about installing the thermalright contact frame. Shouldn’t be more touchy than installing a heat sink on the open die athlon thunderbird 1.2ghz when I was much younger and inexperienced.
@@Akkbar21 12700k is a good choice, I am thinking of it or maybe a 12600k/700k.
the 5800x3d is only an 8 core while the 12700k is what a 12 core 20 thread so sacrifice some gaming performance for extra cores isn't bad for 5 years down the line
@@charleyweinhardt more like 420
Your contributions, your high standards and how you stand up for us consumers is greatly appreciated. Thank you Steve and everyone at GN.
Thank you again for the coverage with such thorough testing and review! Very much appreciated, I've been looking at upgrading my 3900X and this was very insightful!
Glad to provide that for you! It's been a crazy week, but we've loved doing the work on these!
I have the same CPU but run at 1440p ultrawide so I'm not sure it would be worth it this time around.
I'm even thinking about the 5800X3D and keeping my existing platform (x570 board and 16gb memory) yet achieve the same gains at 1440p
@@GamersNexus Does using a 6950xt with SAM affect the results ?
@knives that would be smart decesion imho.
This launch gives Intel an amazing chance to take the crown again when raptor lake launches.
The next gen parts though zen5 and meteor lake are gonna be the real tale if the tape imho.
Kudos for having separate dedicated videos for each model of AMD's 7k series.
It seems to me that the 5800x3d is still highly competitive and is able to keep up very easily in a lot of games.
indeed, the extra cache probably makes it better performance for cheaper price than any of the current 7th gen in many of the games that arent tested (highly crowded mmo towns of ff14/eso/new world) sims like microsoft flight sim and acc, etc but those workloads are always extremely hard to comparatively test for display purposes in videos and need to be approached with much more time sank into it to get better empirical picture
for gaming only for somebody who already has an am4 board, it makes no sense at the moment to upgrade to anything else than x3d especially when considering the price tag of changing platform
until the 3d variant of this gens comes out, which is propably after intel launches their new prodcus for counter attack
Why wouldn’t it be, that chip is released recently, it’ll be competitive for a while longer no?
@@heroicsquirrel3195 yah but it was built on older tech. I would have assumed that the 7950x would beat it rather soundly but that doesn’t seem to be the case. In fact in other videos it seems that it still not only keeps up but beats the entire 7000 series lineup in multiple cases. Until we get more diverse benchmarks we can’t say for sure but for now something is not adding up.
@@theghost8914 Zen 4 doubled the L2 cache. The 5800X3D has 64 MB of L3 cache stacked on the CCD, tripling the total L3 cache. Those will affect performance in different ways.
Zen 4 will get its own 3D V-Cache SKUs next year.
I love that FX landmark for insane power consumption.
Great reviews from the finest of channels.
Thank you for all of your hard work, it's amazing what you line up for us this week.
Thank you for your work! Super detailed coverage like always. The team at GN does absolute awesome reviews. Thanks for letting us make an informed buying decision.
Props to you and the rest of the GN team, Steve! You guys have obviously been working overtime for this release!
Thanks for the great content, and please get some rest! o7
Man, you guys are on the grind 💪. Thank you for the coverage
Seeing all this awesome new hardware being released gets me all hyped up. Its good to see you guys take a step back, put things in perspective and reiterate that it is not needed to have the latest stuff to enjoy games. Thanks for that! :)
Congrats GN team on the channel really taking off over the last 2 weeks. Over 5 million views by the looks of it so far. Thank you GN for the incredible work you do.
Who didn't know of GN before this?
@@elmalloc I'm pretty sure at least someone didn't know lol
tis' the season! the channel spikes around GPU/CPU releases
@@elmalloc I don't know everyone to tell you everyone who didn't know GN before this.
According to statistical websites... no. For example in Sept 2020 they had 18.3 million video views.
There should be more eco mode testing, especially on the higher end amd cpus where it does little difference in gaming but massive per watt performance difference in production tasks. In 2022 with high energy costs in a lot of places , this should be more talked about.
Not only that but my room turns into a hot box
Eco does not make sense in the high-end, I would say PBO2 makes more sense.
@@13thzephyr It makes sense for any chip. You can cut power consumption in half at the cost of only a bit of performance.
@@reddyuda yeah i felt the difference in the room this summer after upgrading from a 3600 to a 5900x on my private pc. I get that this a a gamers channel and we want more fps, but i still think a lot of us are more inclined to run our hardware nearer to an efficiency sweetspot nowadays.
@C what? most (graphics intensive) games are cpu bound, getting more power for games is absolutely right thing to do
Always appreciate the unbiased, professional and thorough testing of these products.
Thank you so much for all of the hard work that it takes to do all of this in-depth testing! We love you guys!
I ordered the GN coasters to help support the channel. GN and Steve are always the first place I go for information and reviews concerning computer tech, and I love all the information that is provided! This information really helps to make an informed decision when purchasing new tech.
I found Der Bauer's delidding interesting and it does support your findings here. He was able to drop the heat by 20 degrees C, though it did take a lot of effort and some fabrication to accomplish. It would be interesting to see what lapping does. Even if it gained 5 degrees C it would impact boost speeds.
I love the analytical, data based approach to the definition of what the results actually represent and what "better" actually means.
14:49 "whether or not you can feel that change depends on how good of a player you are"
16:40 "ultra competitive types will get benefit from the 7900x and 7950x alike, although most people will be bottlenecked on skill before they are bottlenecked on the CPU"
Thanks for the hard work on this. Been looking on upgrading from a 7700k relatively soon and it has been useful in seeing all this data.
Excellent video. Keep up these reviews. I remember a time when you would see all Intel CPUs at the top of review charts, year in, year out.
AMD Nexus, too many AMD reviews on this channel. Unsubbed.
(also great review, please don't hurt me it was just a prank bro)
HoW cOmE yOu ArE aLl ReViEwInG aMd PaRtS tOdAy?!
Oh hi AMD Unboxed!
That's AMDers Nexus to you pal....
Go review a monitor or stfu... 😁
Still on my Sandybridge 2600k, still getting the job done^^
The 2600K is the GOAT!
Appreciate the video on this one, thank you! Awesome work as always!!!
Love that you included the FX 9590. I had a lot of fun with that CPU back in the day, trying to cool it and the power phases on the mobo LOL.
These Blender Power Efficiency charts are just great, GN. Amazing work everyone
Really appreciate the comparisons to the Zen 2 parts. As somebody who generally only upgrades every 2 generations, sometimes it's difficult to find comparisons to the parts I use, since most only focus on new gen vs previous gen or competitors.
I was glad when they pity-included 1700 users like me in a few of the charts throughout these videos.
@@klobiforpresident2254 I’m not much better off still rocking my 3600 lol. Glad they were included in the charts.
Awesome in depth coverage as always. Thank you.
Thanks for all your thorough benchmarking work, Steve - and company :)
It's hard to find the right words to express how much I appreciate the excellent work you do on these reviews. For me, you are the absolute point of reference when it comes to PC building.
Watched 7600x review, went to sleep. Woke up, opened youtube 7900x review is up.
Life is good
You guys never rest!! 3 videos one after another and retaking all your measurements in windows 11 and with blender's cycles X for every CPU. You are all wonderful madmen ❤️
Great on the ball technical review of the latest range Steve, you and the team have excelled on this and your recent EVGA story.
Hey GN, I hate to add ANOTHER test for you guys, but I'd reallly like to see testing data on CPUs for video rendering - particularly using FFMPEG, encoding libxh265 (or other newer codecs). Along the same line, some AI workloads like Topaz Gigapixel or Flowframes for GPU testing would be very much appreciated. I hope we can see stuff like that in the future!
Regardless, im super appreciative of all of the work you guys do, keep it up
Really excited to see the 7700x... And even more excited to see the 3d cache variants down the road. I think a 7800X3D will be absolutely mind blowing paired with a new gen GPU
I'd love to see gzip benchmarks alongside the 7-zip benchmarks included currently. Many webpages will compress the data they send to clients with the gzip algorithm, and for the more "production" use-cases like programming on these CPUs, it would be nice to know how much improvement we can expect to see in this area. Thanks for the great content as always!
Good work guys, dedicated to fair testing and providing the best option for the consumers!
Has anyone told you lately that you and your team do fantastic work? Thank you GN, for truly unbiased tech news.
Thank you!
GN is without a doubt the best creator of all time for PC hardware.
The new thermal behavior really makes me wonder if it's even possible to have a CPU fan curve accommodated for the frequency instead of thermals.
I can't imagine being on constant 100% fan speed to be good for the cooler's power draw (however low compared to the CPU it may be) and its longevity.
Ideally we want Package Power (factual W consumed) used for fan speed.
That's actually one of the ways I was silencing my systems -- ignore the temps and let throttling mechanism act as a safety net, just set constant fan speed enough to cool package power I need (Noctua C14S did good with just 800 RPM for 5800X@100W, practically silent for normal home use)
It seems AMD went wild with planned obsolesence in this new gen.
Same. I won't buy this as I don't want a space heater. This and Nvidia is getting redicilous!
@@Morpheus-pt3wq It's more like "bigger bar better" kind of move to be easier to sell for the average consumer, the same thing Intel did with the i9 12900K to make it look more impressive and a bit better than the R9 5950X albeit consuming twice the power. At least the 7000 series has ECO modes to lower the TDPs to make the CPU more efficient while sacrificing a bit of performance.
@@ismaelsoto9507 ECO mode isn't new, I've used it on my 5950x since release and with manual RAM OC and per core undervolt you can retain stock performance at 65W
Thanks for the great video mate. Just picked up my 7900x and am excited for it to arrive!
There has been a content waterfall the past week, we are spoiled lol
Thanks for all the hard work you put into making these videos!
Thanks guys , sticking to the 5000 series for the upgrade -- cheap cheap
I feel like Christmas receiving so many videos from your channel .with a lot of interesting information, I like that this channel always lowers the emotion of hype and reviews what these new products really do for the average consumer and how they affect us. greetings from Argentina :D
Thanks doc
Thanks for the fairly comprehensive review. This is great for informing me to whether to upgrade from my 2600 to the new platform or 5800X3D. Keep up the great work 👍
thanks for all your hard work testing much appreciated
I would be really interested in how the performance scales with the downgrading of cooling. Like what happens with something like an SFF-grade 60-70mm hight-limited CPU cooler
Yes! I'd also like to see the 7600X tested with a 120mm single tower cooler to see how it would perform in a budget build.
Let´s boil some eggs or cook few steaks!
With these temps and wattages it wont be long until we start strapping steam generators to our CPUs just to do something with all this heat
Just ordered mine!! Super stoked
Thank you for the comprehensive review.
I’ve recently upgraded to the Ryzen 9 7900X and it is absolutely AMAZING! From a Ryzen 7 3700X that is.
yeah this basically confirms i'll stay with my 5900x until the initial kinks (and there absolutely will be) that a whole new platform + ram brings, far less of an improvement than i expected tbh.
You expected more than 22% single core and 40% multicore? Those were some high expectations :)
Thank you for the video!
Massive work mate. Kudos
Thanks for including the 1700, currently have it so good so what an upgrade would do.
Same, been using it since 2017
Microcenter has the 7000 series in stock, they sent me the email flyer today. The motherboards they have listed start at $479 for the Gigabyte X670E AORUS eATX.
As a gigabyte aorus owner I have to say it's in my do not buy list in the future. I am on my third z390 board and you remember the exploding psus last year?
@@timothygibney159 Juuuuust referencing lowest price. The others were 600 and 900 respectively. You do you.
Thanks for all the work! 👍
Thanks! Great review.
So does this new first to thermal ceiling behavior also means that my tower cooler (NH-D15) will constantly ramp up and down as I "open new Firefox tabs"? This was an issue with 1st gen Ryzens. How is this solved or experienced?
Yeah... unless they backpedal to default human-friendly behavior now that the pretty benchmarks they wanted are already out, I expect a lot of negative feedback from regular users with enough money to buy these, but not enough experience to tame them.
Gamers Nexus is a jewel for computer hardware reviews and testing. When I think back to what we had in the 90s compared to highly reputable outlets like this, I feel really happy. Proper testing that you strive to improve constantly and unrelenting impartial standards. Keep it comings guys. 🙏👏👏
I want Steve coming in me
@@BBWahoo eww.
@@BBWahoo 😳
@@BBWahoo 🥴
Thanks for all the work you guys put into these reviews. A lot of fun to watch and learn little by little each video.
I do a bit of video upscaling using video2x, I'd love to see a chart thats applicable to that work as new GPUs launch this winter.
I love the watt-hour comparison on the Blender render. It would be good to see that in the future for GPUs.
I am staying with my 5950X for a year or 2 yet no reason to upgrade for a while. And I really do like the power efficiency of it.
Looking at the 7900, it makes me want to put thermal pads between the IHS on those exposed parts of the chips so they get some contact with the cooler, even if indirectly
not even a need to do so. On intel chips those are on the contact side of the cpu and also no not have any contact with the ihs.
@@chaous2000 don't need to do so, no, most graphics cards going back don't either. But it makes me want to
Thanks for making the videos!
Dude this is the stuff that makes you a legend in our communities.
Every single detain hammered out.
Absolutely astonishing work.
I have mixed feelings with AMDs new approach to power/boosting. I like a cool and most importantly quiet system; I'll probably never go beyond a 6 core part, but I'm curious how programs like speedfan are going to interpret AMDs new approach. Speedfan is how I ensure a quiet system since I use my PC as a TV and media consumption as well as gaming. Either way I'm waiting at least until the Ryzen 8000 series before I decide to upgrade.
See you in a month when we do it all again for for 13th Gen Intel👌
Just wanted to say thank you for including compiling code (Chromium) in your tests.
Steve and team, you're doing amazing work for all consumers. Thanks a lot, you deserve all the best :)
Can't wait to see the Zen4 chips with 3D cache. Robert Hallock has stated that the 1.350vcore limit of the 5800X3D will NOT be present on the Zen4 versions - meaning we could potentially see a "7800X3D" with 128MB of cache, 5GHz+ clockspeed and very high speed Infinity Fabric and RAM speeds, negating the largest weaknesses of the 5800X3D
The weakness is temps. That's why they have all the limits on 5800x3D
And since zen4 is already pretty hot out of the box, it will most likely be only on mid tier chip like 7700x3D, maxing out gaming but sacrificing compute performance.
These detailed specs are incredibly usefull to anyone who wants to be ready to go on launch day... That being said, the highlight of the video for me was 8:47 where you placed the CPU box *exactly* in the middle of the stand without even looking. This pleases my OCD greatly.
You are not ocd enough then, it wasnt exactly in the middle, it is slightly offset to the right. The wood bit between the two metal bits at the bottom is lined up closer to the edge of the inner square of the packaging than the center. It is more than can be accounted for with parallax error.
@@irishRocker1 That was noticed, worry not. With age, my OCD has deemed it necessary not to obsess over the small small details, but over the completely unnecesary small details in general.
Thank you GN for the early and thorough reviews! I'm planning to build a VR Flight Simulator PC and need all the help I can get selecting a cpu and gpu that are good enough but not overkill
Honestly I'd buy Alder Lake and RDNA2 if you were to build right now.
Motherboards are priced fairly, can use DDR4 if you want, 12400 or 12600. Or you could wait a few months for Raptor Lake.
Pair with something like 6800xt. Really solid Flight Sim build.
@@libertyprime9307 Thank you!!
thanks for once again answering my question, been thinking about upgrading from the 3900x to the 7900x big time enthusiast, sometimes run multiple instances of a game at once while listening to music watching streams etc love having the extra power for multi tasking and it seems this chip will be the perfect upgrade, hoping i can get some sick black friday deals and build another awesome machine to last several years
I'd love to see someone chart the change in ambient room temperature with a Ryzen 7000 series processor v. previous gen Ryzen and Intel. 95 degree air blasting out the back of one's PC is sure to cause a biblical flood of gamersweats.
The temp itself is not relevant, but power (W) is. When delidded the temp will drop 20 degrees, but power consumption is virtually the same. None of these CPUs pull these figures when gaming. Be more concerned about GPU power draw for those gamer sweats.
I think the power consumption and cooling factor is why they added the ECO modes - if you have a crappy PSU and cooler, you can just set the ECO mode appropriately until you can afford to upgrade those components ... or not. PCWorld showed the performance of 7950x at "65w" was still beating 12900k in some tests.
Now that is pretty crazy.
Fantastic coverage as always. Just a quick note, at 17:23 in R6S, the 7900x leads the 5900x by about 2%, not 12%.
Thanks GN. I really wish a reviewer would test music workloads, such as tracks/plugins. I understand there really is no workbench that would do this, nor is it a gaming wheelhouse thing.
In fact, the default mode (almost PBO no limit?) looks screwing up the whole value of Zen4.
Looking forward some ECO mode testing to save Zen4.
Debauer had lots of trouble even coolung it to 70c. This is unacceptable and gn ignored this 😡
Hard Pass on these 7000 series, I'll wait for next year.
7800X3d perhaps?
@@camy205 that will run 150c prolly
Love the honest reviver level scale!
And plz do some
Eco testing. With current energy prices in the EU…. Yeha we need it
Great Vid! Are you guys planning on how the 90-95°C affects other Parts (temperature-wise) in a Case?
Will you check ECO mode? Cuz it shows impressive performance with only1/3 TDP !!!
3900x gang rise up
Aye 🫡
Can 5900x gang join
@@lochl2 lol yes please
@@lochl2 exactly my thought. Cheers for the x900x gang
and go to microcenter to grab a 5800x3d?
thanks for these excellent videos!
Really great work on the release content as always! :)
Correct me if I'm wrong, but I think there is an error at 7:30, where you say the gain is non-linear between the 12core 7900x and 16core 7950x in the blender test.
If the work took 6min for 16core and 8min for 12core. Then in 1min, the 16core did 1/6th work and the 12core did 1/8th work. Therefore in 1min, each core for the 16core did 1/96th work (1/6 / 16) and each core for the 12core did 1/96th work (1/8 / 12).
Each core on the both CPUs are doing the exact same amount of work per minute which means an exact linear increase in work relative to the number of cores.
I think this error is from using the comparison of 25% less time to 33% increase in cores. This would be a problem in the following comparison where the 16core took 6min and an 8core took 12min. Clearly the core scaling is linear as the 16core can do the work in half the time due to double the cores. But using the percentage approach would come out as 50% less time for 100% increase in cores. So comparing the value of these 2 percentages can be a bit misleading.
Scared to jump on am5 got burnt with a b350 board that didn’t support pbo and other features for the newer cpus
Hardly a burn, it still worked as a back up and it didn't need to go to landfill 👀
I'm most likely going to stick with my 5900X unless I build a second PC
Thank you for the very technical reviews as usual, really like your videos!
I have 2 comments, hope they are not too dumb..
1) The efficiency test (w/h for each frame) is really useful because it objectively serves to prove/counter claims of efficiency from amd and intel, so that's a great addition.
I was wondering though if you could add other efficiency tests like these, say for certain games or other production tools, since cpu performance may vary a lot depending on what it's running
2) Sorry I am mostly a gamer and not the least knowledgeable as you guys are, but would it be possible to add other production tests? Even if not a usual software people use, but just to again show the scalability between CPUs
I have an 7700x.I have set PBO, 85c TJ MAX and -30mv to voltage curve. For cooler I have a be quite silent 2, 280mm AIO. I saw an increase in performance and much lower temperatures. Geekbench from 2100 and 14500 to 2240 and 15450. Cinebench from 19900 to 21100. I have also overclocked memory from 5600 36-38-38-80-118 to 6000 32-38-36-76-112.
I think a reason AMD isn't making low end parts is because they have the 5000 series as well as plenty of chips on the used market that are simply better value. I'd pretty much never consider a new chip sub 300$ when I can just look on eBay and score a deal. When raptor lake launches AMD can just drop prices, they have the thick profit margins now to do so.
There's also the issue of yields. With TSMC's absurdly good yields on the N7 and N5 nodes, AMD almost _needs_ to target R3s, hence the R3s basically becoming unicorns since Zen 2 (Zen 3 R3s are all APUs, none of which are DIY; Zen 3+ doesn't even *have* R3s to speak of, to say nothing of it being laptop-exclusive).
@@sonicboy678 lol, AMD is suffering from success.
Releasing heavily overclocked CPU's seems not to be the smartest move at a time where energy prices going through the roof in some parts of the world. We know Intel is doing that but that's one reason they fell behind. It's disapointing seeing AMD doing the same thing now.
At the end the customer gets a energy hungry, loud product - just to win the benchmarks at release. Personally I want the opposite.
It's still pennies.
The delta is something like two 60w lightbulbs at most.
One can probably afford that if one can afford a high-end computer.
It's not the case for data centers, though.
@@Vladislav888 Yea, maybe. It is still a waste of energy. Most users would be totally ok with efficient settings as standard. As they've always been. And the small number of overclockers went higher - with all the disadvantages and further costs. Everyone was happy and it was "the right thing". Now it's the opposite - you need to adjust it on your own to eco settings. However the majority of the users will not even know that it exists.
great analysis
thanks for the video
“who on earth would want a CPU review at 3 in the morning?” “OH BOY 3AM!!!!”
Just the rest of the world.
midnight on west coast… 8-9 am in Europe…. and who cares when one posts a prerecorded video… it is there now, tomorrow, next month…
@@macking104 not hating at all, i was just bored and this immediately popped up in my sub feed :)
The vibe on this one seems a bit more positive than either the 7950x or 7600x. My concern with this new generation is about heat and power tho, I went to a fair bit of effort to build my current 2700x system to be small (ITX + SFX-L) and quiet (Noctua U12S), really not sure how any CPUs in this family are going to play well in such builds.
Steve kind of went off on the 7600x, which is a really good cpu, every one else out there was really positive, check out 'hardware unboxed's coverage
That why I never went itx. Of course your not going to be able to just swap parts with new gen. To me, itx was a minimalist approach to balancing case options with the thermals of your hardware. Its basically a custom case.
All of them are better for an ITX build if you set the power limits to a reasonable level. The stock power settings are pushed to the limit because of the marketing importance of being top.
Reasonable power limits will have them much more efficient and a good match for ITX.
@@egalanos am I missing something? 🤔 a 7600 goes toe to toe with a 58x3d. If he drops performance why buy it then? 🤔the buy in for a 7600 is 1200 bucks. New board. New ddr5. New AIO. Well no new aio in an itx....so hes probably stuck at a 240 rad. I honestly wouldn't buy it.
@@egalanos There seems to be some kind of "Power" vs "Eco" mode, would love to see more details about this in the reviews.. Given that top-end performance seems to take exponentially more power/heat .. it's often not that big of a compromise to sacrifice a little in order to have something that's still really good but keeps power consumption and heat (and thus noise) down.
GN is the only channel I don't scrub through the timeline on when he mentions the sponsor. I actually would buy and have bought some of that stuff.
Hi GN! I was wondering if you guys could do (if this hasn’t been suggested before) a test of ryzen 7000 based on various cooler sizes as not everyone is able to throw a 360 AIO in all cases. It would be cool to see how the CPU underclocks in comparison to smaller coolers