AMD Ryzen 9 9950X3D Leaked PassMark Score Shows 14% Single Thread Improvement Over Predecessor

[T0@st]

Last Friday, AMD confirmed finalized price points for its upcoming Ryzen 9 9950X3D ($699) and 9900X3D ($599) gaming processors—both launching on March 12. Media outlets are very likely finalizing their evaluations of review silicon; official embargoes are due for lifting tomorrow (March 11). By Team Red decree, a drip feed of pre-launch information was restricted to teasers, a loose March launch window, and an unveiling of basic specifications (at CES 2025). A trickle of mid-January to early March leaks have painted an incomplete picture of performance expectations for the 3D V-Cache-equipped 16 and 12-core parts. A fresh NDA-busting disclosure has arrived online, courtesy of an alleged Ryzen 9 9950X3D sample's set of benchmark scores.

A pre-release candidate posted single and multi-thread ratings of 4739 and 69,701 (respectively), upon completion of PassMark tests. Based on this information, a comparison chart was assembled—pitching the Ryzen 9 9950X3D against its direct predecessor (7950X3D), a Zen 5 relative (9950X), and competition from Intel (Core Ultra 9 285K). AMD's brand-new 16-core flagship managed to outpace the previous-gen Ryzen 9 7950X3D by ~14% in single thread stakes, and roughly 11% in multithreaded scenarios. Test system build details and settings were not mentioned with this leak—we expect to absorb a more complete picture tomorrow, upon publication of widespread reviews. The sampled Ryzen 9 9950X3D CPU surpassed its 9950X sibling by ~5% with its multi-thread result, both processors are just about equal in terms of single-core performance. The Intel Core Ultra 9 285K CPU posted the highest single-core result within the comparison—5078 points—exceeding the 9950X3D's tally by about 7%. The latter pulls ahead by ~3% in terms of recorded multi-thread performance. Keep an eye on TechPowerUp's review section; where W1zzard will be delivering his verdict(s) imminently.



View at TechPowerUp Main Site | Source

 

[Scattergrunt]

Can't wait for what the 9900X3D and 9950X3D will look like. I'm all for the X3D chips!

 

[Frick]

Can't wait for what the 9900X3D and 9950X3D will look like. I'm all for the X3D chips!

I'd like one that is less than €300.

 

[Michaelgr]

My PC with Core Ultra 9 285K PassMark cpu Test without overclocking.

 

[freeagent]

My PC with Core Ultra 9 285K PassMark cpu Test without overclocking.

View attachment 388874View attachment 388875

Why did you put this here?

Anyone who is looking at X3D is not looking at Intel.

 

[Michaelgr]

Why did you put this here?

Anyone who is looking at X3D is not looking at Intel.

Βecause passmark and the article write that the 9950x3d is faster than the core ultra 9 and that is not true.

 

[Dr. Dro]

Βecause passmark and the article write that the 9950x3d is faster than the core ultra 9 and that is not true.

PassMark really needs to update their software suite. It's really not adequate for modern PCs, that struck bad when the 5090 ended up exposing that much of it still relies on crufty old 32 bit code.

 

[burntruers]

Βecause passmark and the article write that the 9950x3d is faster than the core ultra 9 and that is not true.

Sorry, are you a hardware reviewer that has been testing the 9950X3D in a comparable test bench to your 285K for a direct comparison? Or are you actually just comparing your own individual system's PassMark score to the score for your processor in the PassMark DB and then drawing your conclusion from that..

Aside from overclocking there are so many variables that could cause a 2-3% performance swing in one direction or another, individually in some cases. That's why reputable publications put so much effort into maintaining comparable test systems for comparing across different platforms.

Your statement may end up being accurate, but right now (unless you have some inside information), you don't have sufficient evidence to support it.

 

[Michaelgr]

Sorry, are you a hardware reviewer that has been testing the 9950X3D in a comparable test bench to your 285K for a direct comparison? Or are you actually just comparing your own individual system's PassMark score to the score for your processor in the PassMark DB and then drawing your conclusion from that..

Aside from overclocking there are so many variables that could cause a 2-3% performance swing in one direction or another, individually in some cases. That's why reputable publications put so much effort into maintaining comparable test systems for comparing across different platforms.

Your statement may end up being accurate, but right now (unless you have some inside information), you don't have sufficient evidence to support it.

This test is from my own computer and I posted it to show that what they write about Intel processors is not true. Most sites write that INTEL Arrow Lake processors are useless and that they are not good for games and applications, A BIG LIE!!, AMD processors are not the only fast ones. There are faster ones like the Arrow Lake Core Ultra 9 285K. In a few days when I will have the Nvidia 5080 graphics card that I am waiting for arrive I will upload a more complete test with Passmark.

 

[Michaelgr]

Sorry, are you a hardware reviewer that has been testing the 9950X3D in a comparable test bench to your 285K for a direct comparison? Or are you actually just comparing your own individual system's PassMark score to the score for your processor in the PassMark DB and then drawing your conclusion from that..

Aside from overclocking there are so many variables that could cause a 2-3% performance swing in one direction or another, individually in some cases. That's why reputable publications put so much effort into maintaining comparable test systems for comparing across different platforms.

Your statement may end up being accurate, but right now (unless you have some inside information), you don't have sufficient evidence to support it.

AMD's processors are not the fastest, there are faster ones like the Arrow Lake Core Ultra 9 285K.

 

[Dr. Dro]

AMD's processors are not the fastest, there are faster ones like the Arrow Lake Core Ultra 9 285K.

The only problem is that grunt alone doesn't always get the job done. At 16 cores the 9950X3D won't be a slug like the 8 core Ryzen 7 models at the things that Intel chips seemingly excel at, like shader compilation or video encoding. And they'll get AVX-512 too.

While it's fairly accurate that Lion Cove is indeed the strongest type of core right now, the fact that the 285K has many latency problems that it's predecessor doesn't have doesn't help it much either. There's more than a few applications my 13900KS will beat it, never mind this one.

Intel needs to execute Panther Lake flawlessly to get back in the saddle, but I don't expect them to win at least until Nova Lake. There's no use denying, AMD is indeed ahead.

 

[b1k3rdude]

Welp either the leak is misleading (highly likely) or the software isn’t compatible with this CPU as its reporting only 32MB for the L3 cache..

 

[burntruers]

This test is from my own computer and I posted it to show that what they write about Intel processors is not true. Most sites write that INTEL Arrow Lake processors are useless and that they are not good for games and applications, A BIG LIE!!, AMD processors are not the only fast ones. There are faster ones like the Arrow Lake Core Ultra 9 285K. In a few days when I will have the Nvidia 5080 graphics card that I am waiting for arrive I will upload a more complete test with Passmark.

So PassMark are lying? What about w1zzard at TechPowerUp, the site you're currently on, who just posted his 9950X3D review and showed that it is either slightly or significantly faster than the Ultra 9 285K in all but a few barely meaningful applications in which it is 1-2% behind?

What about Gamer's Nexus, who also just posted their review and showed the same result?

And Tom's Hardware, Hardware Unboxed, der8auer, they're all participating in this "BIG LIE" as well?

Considering that their results are consistent with one another, they would need to be part of a grand conspiracy in which they're agreeing on the false results they're going to publish, wouldn't they?

And what is their incentive to lie? If your explanation is just that "they're all paid shills, I don't have any evidence but trust me bro", that's just unhinged, paranoid, delusional

EDIT: Ah, just checked your post history. Literally 9 comments, all of them vigorously "debunking" negative reports about Intel. Interesting.

 

[Dr. Dro]

So PassMark are lying? What about w1zzard at TechPowerUp, the site you're currently on, who just posted his 9950X3D review and showed that it is either slightly or significantly faster than the Ultra 9 285K in all but a few barely meaningful applications in which it is 1-2% behind?

What about Gamer's Nexus, who also just posted their review and showed the same result?

And Tom's Hardware, Hardware Unboxed, der8auer, they're all participating in this "BIG LIE" as well?

Considering that their results are consistent with one another, they would need to be part of a grand conspiracy in which they're agreeing on the false results they're going to publish, wouldn't they?

And what is their incentive to lie? If your explanation is just that "they're all paid shills, I don't have any evidence but trust me bro", that's just unhinged, paranoid, delusional

EDIT: Ah, just checked your post history. Literally 9 comments, all of them vigorously "debunking" negative reports about Intel. Interesting.

I mean, there is some merit to it, the problem with the 285K is almost like the problem with Bulldozer, even though they don't share the same root. It's nice on paper, like, really nice. But in practice, well, it's a steaming heap. 285K will always be a CPU that shines in theoretics and what-ifs, just like the AMD FX processors of old.

If one opted to buy it, they should pretty much just go enjoy the thing and forget about benchmark charts lol

 

[Scrizz]

Sorry, are you a hardware reviewer that has been testing the 9950X3D in a comparable test bench to your 285K for a direct comparison?

I'd love to see you create a "comparable test bench" for an Intel/AMD CPU.
Extra points for keeping all other components the same