AMD Ryzen 9 7900X Review: Needs To Be Cheaper, Just Get 7950X

Buy relevant products from Amazon, Newegg and others below:

Video Index:
00:00 - Welcome back to Hardware Unboxed
02:12 - Test System Specs
03:16 - Clock frequency behaviour
03:57 - Cinebench R23
04:39 - 7-Zip File Manager
05:14 - Blender Open Data
05:27 - Corona 1.3 Benchmark
05:41 - Adobe Premiere Pro 2022
05:57 - Adobe Photoshop 2022
06:16 - Adobe After Effects 2022
06:33 - Chromium Code Compile
06:51 - Factorio
07:14 - Watch Dogs: Legion
07:25 - Tom Clancy's Rainbow Six Extraction [Vulkan]
07:39 - Hitman 3
07:54 - Tiny Tina's Wonderlands [DX12]
08:24 - F1 22 [DX12]
08:46 - Spider-Man Remastered
09:05 - Shadow of the Tomb Raider
09:22 - Horizon Zero Dawn
09:43 - Cyberpunk 2077
10:07 - Assetto Corsa Competizione
10:21 - The Riftbreaker [DX12]
10:41 - Counter-Strike: Global Offensive
10:57 - 12 Game Average
11:16 - Power Consumption
11:56 - Cooling Performance
13:46 - Cost per Frame CPU
14:11 - Cost per Frame CPU + RAM + MB
15:09 - Cost per Frame CPU + RAM +MB (Cinebench)
15:39 - Cost per Frame CPU + RAM + MB (Premiere)
15:58 - Final Thoughts

AMD Ryzen 9 7900X Review

Disclaimer: Any pricing information shown or mentioned in this video was accurate at the time of video production, and may have since changed

Disclosure: As an Amazon Associate we earn from qualifying purchases. We may also earn a commission on some sales made through other store links

FOLLOW US IN THESE PLACES FOR UPDATES

Outro music by David Vonk/DaJaVo

[yapay_zeka_gundemi]

The title itself alone says how different and valuable this channel then most of the rest. No clickbait, clear sentence summarizing the the whole video. You are one of the best ...

[DATBOCK]

You keep saying gamers would be better off just going with a 7600X or 7700X but wouldn't they really be better off with a 5800X3D or 12700K/ 12600K and save hundreds going with a budget board and DDR4? That they could then put those savings towards a better GPU which would actually make a considerable performance difference in gaming? Idk just seems Zen4 for gamers right now isn't worth it whatsoever at least the value gamers that's for sure.

[evildoer]

Thanks for not rushing in with reviews, and taking it one by one. Great, detailed job, as always! Looking forward for the motherboards :)

[peterjansen]

A power-target benchmark would be very interesting, seeing how those Zen4-CPU's hold up with 105, 90 and 65W, something around those numbers. I also would be interested in a deep-dive into coolers. Do the old towercoolers suffice (also taking sound into consideration)? Can you put a Noctua D15 (AM4 edition) on AM5 without a extra kit? Such things.

[Sam-hjhy]

One of the reasons why the 5900x was so popular was that it had the cheapest price/core ratio and was great at everything.

[UnderageBeerHere]

11:35 - At 242 watts total system power consumption for the 5900x, the 7900x at 302 watts is NOT 47% more power consumption. It's not even close. 302w/242w = 1.2479 (a 24.79% higher power consumption for the 7900x).

47% is the difference between the 5900x and the 7950x.

[wanotai]

just got my 5900x for 300usd... guess i wont be upgrading until socket AM6

[conza]

The B-roll is a testament to "Hey look! No pins!". Looking forward to seeing if these IHSs can be shrunk/thinned, would be great if we can do that by Zen 4 3D

[la]

I would love to see a full benchmark run from any single ryzen 7000 chip with the same undervolt optimum tech published. So can getting that 20c drop impact Performace either way.
Keep up the great work!

[fVNzO]

Hope you understand but i don't usually have time to watch the whole video here, and i feel like watching one of the Zen 4 reviews gives me enough context to be able to comfortably skip the benchmarks and get straight to the cost per frame and conclusion. The timestamps are a godsend is what I'm trying to say.

[eNoqfOu]

At the Cost per Frame chart, you should take $100 off of the RAM for the 5800x3d as it doesn't care much about RAM speed, DDR4-3600 only gained 7.2% performance over DDR4-2133. 3600 c14 vs c16 is less than a percentage difference in performance, and a c16 can be picked up for $100 instead of $200 for c14. This would put the cost per frame at 4.49 for the 5800x3d.

[KeplersConjecture]

Love to see the 5800X3D still holding its own on the charts.

[CommanderGinyu]

Man getting the 5800x3D was such a great choice especially considering how cool it runs even on heavy loads (never gets hotter than 50C)

[soorajj]

4:10 Math is wrong . 7900x approx 24% slower than 7950x in Cinebench R23 multi

[RchamTV]

Im in dire need of a new workstation / gaming rig and am torn between the 7900x /7950x or using a 7700x temporarily for streaming and building a cheap stream PC later

[Mad.Maxx.]

I'm going with AMD for my new build. This will be my first ever non-Intel build. Already have my mobo (X670E Crosshair Hero). Now, just need to decide on the CPU. Great video, mate!!

[deXXXXter]

I would love to see the same ryzen 7000 processors review at 65W.

[FeTiProductions]

I truly appreciate the direction of the vid titles these days. Open and honest from the get go

[TrueThanny]

Zen 4 is not worse than Zen 3 in performance per watt. To properly compare that metric, you have to run them at the same power level. Since you can't get Zen 3 up to a 170W TDP, that means running Zen 4 at a 105W TDP, where it will still greatly outperform Zen 3. It will still outperform Zen 3 at a 65W TDP as well, even when Zen 3 has a 105W TDP.

Being able to squeeze out an extra 10% performance by using an amount of power that Zen 3 can't reach doesn't make Zen 4 less efficient. It makes it more scalable.

[polygons]

@Steve, 5:23 maybe there is a typo or something... In the Blender test, it is said that the 7900X is a 30 % faster than the 12900K, but I've made the operation with those numbers, and unless I'm terribly wrong, here it's giving me a 18% faster, so, much less. Or maybe am I not considering something ? (btw, Blender is the app I care more about ;) )
Also, in the Cinebench test, multiple core, 4:10, It is also said "it was also just 12% slower than the 7950X". But 29.358 is the 76.92% of 38.165, so, the difference is a bit more than 23%, not 12%. At this point, I'm starting doubting myself, I must be doing the math wrong... I only did look at it as the size of the bars didn't match with what I was hearing.
If it's a typo, understandable, with that mountain of data and tests.... Sorry if it is my mistake...