AMD's Reviewers Guide for the Ryzen 9 7950X3D Leaks

[TheLostSwede]

AMD's Ryzen 7000-series CPUs with 3D V-Cache are set to launch next week and alongside the launch, there will obviously be reviews of the upcoming CPUs. As with many other companies, AMD prepared a reviewers guide for the media, to give them some guidance, as well as expected benchmark numbers based on the test hardware AMD used in-house. Parts of that reviewers guide has now appeared online, courtesy of a site called HD Tecnologia. For those that can't wait until next week's reviews, this gives a glimpse of what to expect, at least based on the games tested by AMD.

AMD put the Ryzen 9 7950X3D up against Intel's Core i9 13900K, both systems were equipped with 32 GB of DDR5-6000 memory and liquid cooling. Tests were done with both AMD's own Radeon RX 7900 XTX and an NVIDIA GeForce RTX 4090 graphics card. We won't go into details of the various benchmarks here, as you can find those below, but according to AMD's figures, AMD came out on top with a 5.6 percent win over the Intel CPU, at 1080p using the Radeon RX 7900 XTX and by 6 percent using the GeForce RTX 4090. This was across 22 different games, with Horizon Zero Dawn and F1 2021 being the games favouring the AMD CPU the most and Far Cry 6 and the CPU test in Ashes of the Singularity being the games favouring the AMD CPU the least. TechPowerUp will of course have a review ready for your perusing by the time the new CPUs launches next week, so you'll have to wait until then to see if AMD's own figures hold true or not.



View at TechPowerUp Main Site | Source

 

[Zunexxx]

More interesting comparison would be faster ram on the Intel side and what it is on the AMD side. Intel can support faster ram while AMD can't. Tbh, it seems to be quite lackluster performance gain.

 

[loki_toki]

hope you're gonna disable the fast CCD to compare how the 7800x3d will bench without waiting 5 more weeks

 

[ZoneDymo]

CSGO is doing a bit worse then without hte 3DCache, probably more reliant on Clockspeed.

 

[Denver]

Still had to be having problems in managing CCDs, since there are games that the 7950x was faster.

 

[jesdals]

More interesting comparison would be faster ram on the Intel side and what it is on the AMD side. Intel can support faster ram while AMD can't. Tbh, it seems to be quite lackluster performance gain.

I second that - those numbers would look different with Intel and 8000Mhz DDR5 - but so would the price

 

[TheoneandonlyMrK]

More interesting comparison would be faster ram on the Intel side and what it is on the AMD side. Intel can support faster ram while AMD can't. Tbh, it seems to be quite lackluster performance gain.

A more interesting comparison would be with an aIo for cooling on the AMD and just a 2p coin on the Intel.
If we're biasing a test why go half way I say .

 

[Hxx]

im not impressed with these results from a processor that costs $120+ more than the current top dog at its current retail price. Realistically performance may actually be less since AMD likely cherry picked some of those results. Regardless i think the real performance gains will matter on the lower SKUs as they will be cheaper and more direct competing on price. Im looking forward to the 7800x3d in april.

 

[phanbuey]

Was expecting more to be honest. You can get +5-10% from a RPL just by bumping up TREFI -- going to wait for the 7800x3d reviews.

 

[AnotherReader]

Was expecting more to be honest. You can get +5-10% from a RPL just by bumping up TREFI -- going to wait for the 7800x3d reviews.

It's going to vary from title to title. There are some where the gains are very impressive and the difference is too great for Raptor Lake to overcome. Then there are others where there are no gains or even regressions. It isn't a surprise; not all games would be memory bound to the same degree, and many would be satisfied by a 32 MB L3.

 

[rv8000]

Was expecting more to be honest. You can get +5-10% from a RPL just by bumping up TREFI -- going to wait for the 7800x3d reviews.

There are some games that see significant gains from very tight timings at ddr5 6000 for Zen 4 just as well.

Overall it’s a nice step up from non 3D cache parts. Mixed use cases also get the benefit of a normal CCD this gen too. 7900X3D and 7950X3D are definitely a good step up from the 5000 series in terms of overall value.

 

[AnotherReader]

We will see this pattern until the stacked cache is below the CPU chiplet. AMD will do this with MI300, but we'll see what tradeoffs come with that approach.

 

[phanbuey]

There are some games that see significant gains from very tight timings at ddr5 6000 for Zen 4 just as well.

Overall it’s a nice step up from non 3D cache parts. Mixed use cases also get the benefit of a normal CCD this gen too. 7900X3D and 7950X3D are definitely a good step up from the 5000 series in terms of overall value.

Agreed - Zen 4 is extremely reliant on timings -- I think u can squeeze 10-20% just tightening timings. Compared to 5000 it's a great step up, and I do like the platform longevity also.

Tight timings giving huge boosts has been the case with all Zen forever though, it's just now starting to make it into the mainstream (because HWUB made a video about it - and also conveniently mentioned that they've been using a DDR5 6000 CL30 kit to do their reviews - their numbers make a bit more sense now).

 

[fevgatos]

A more interesting comparison would be with an aIo for cooling on the AMD and just a 2p coin on the Intel.
If we're biasing a test why go half way I say .

The only biased test is the one using similar ram. AMD's stock ram is 5200 and max oc is 6000-6400 tops. Intel stock is 5600 and oced they can easily exceed 8000. Obviously, testing both with 6000 is the very definition of a biased test. Especially considering

1) A 6000c30 costs as much as a 6600c34 kit.
2) A 7200c34 kit is just 50€ more expensive than a 6000c30 kit
3) The 7950x 3d + 6000c30 is in fact more expensive than a 13900k + 7600 kit

 

[evernessince]

I second that - those numbers would look different with Intel and 8000Mhz DDR5 - but so would the price

If you are upping the frequency on the Intel side are you tuning the timings on the AMD end? You can gain a significant amount of performance on the AMD side with tuned timings.

 

[ThrashZone]

Hi,
I'd like to see the latency numbers between the two
AMD mid 50's to intel's mid 30's is likely.

 

[trsttte]

What's that "AMD reference AM5" motherboard they speak off? Is that a "whatever you have hanging around" or is AMD making a reference board public?

 

[btk2k2]

Given that there are no super cache sensitive titles in that it's not to bad.

If it was 6% and included stuff like MSFS, ACC, Factorio etc then that would be a totally different story.

 

[Space Lynx]

The only biased test is the one using similar ram. AMD's stock ram is 5200 and max oc is 6000-6400 tops. Intel stock is 5600 and oced they can easily exceed 8000. Obviously, testing both with 6000 is the very definition of a biased test. Especially considering

1) A 6000c30 costs as much as a 6600c34 kit.
2) A 7200c34 kit is just 50€ more expensive than a 6000c30 kit
3) The 7950x 3d + 6000c30 is in fact more expensive than a 13900k + 7600 kit

imo tests should be done with 7200 cl 34 (its not that expensive surprisingly, but anything above that is) for intel, and a 6000 cl 30 top kit for AMD.

i don't know, its weird these days. lol

 

[Daven]

A more interesting comparison would be with an aIo for cooling on the AMD and just a 2p coin on the Intel.
If we're biasing a test why go half way I say .

That made me lol! Lol!

 

[AnotherReader]

I'd like to see the latency numbers between the two
AMD mid 50's to intel's mid 30's is likely.

If you're thinking of L3 latency, AMD has much lower latency than Intel.

 

[ThrashZone]

Hi,
I was going by posts on this thread here's a 5800x3d

Share your AIDA 64 cache and memory benchmark here

Playing around with a new beta bios, L3 is a bit sloppy but not horrible.. Latency is decent for the CPU..

www.techpowerup.com

 

[AnotherReader]

imo tests should be done with 7200 cl 34 (its not that expensive surprisingly, but anything above that is) for intel, and a 6000 cl 30 top kit for AMD.

i don't know, its weird these days. lol

While the tests should be done with higher speed RAM for Intel, I don't think the price differences are correct. These are the lowest prices for a 32 GB kit of DDR5 on PcPartPicker in the USA:

1. DDR5 6000 CL30: $ 146
2. DDR5 6600 CL34: $ 188
3. DDR5 7200 CL34: $ 247

Hi,
I was going by posts on this thread here's a 5800x3d

Share your AIDA 64 cache and memory benchmark here

Playing around with a new beta bios, L3 is a bit sloppy but not horrible.. Latency is decent for the CPU..

www.techpowerup.com

I see. You mean memory latency; in that case, I expect Intel to have lower latency.

 

[mb194dc]

People are going to spend 2k minimum and then use 1080p?

Let's see the 4k results with RT or higher.

 

[A Computer Guy]

Why does it still seem like we are struggling with 1080p on top tier hardware?

hope you're gonna disable the fast CCD to compare how the 7800x3d will bench without waiting 5 more weeks

I was wondering that too and if somehow it's against the rules?