Tuesday, October 21st 2025

AMD Could Launch 16-Core Ryzen 9 9950X3D2 CPU with 192 MB L3 Cache and 200 W TDP

AMD seems to be preparing an update to its "Granite Ridge" Ryzen 9000 X3D series, introducing two new chips targeted at high-performance gamers and creators. Among the two, a new SKU type has appeared, carrying the "X3D2" suffix. Leaks suggest a new flagship model, the Ryzen 9 9950X3D2, featuring 16 cores and 32 threads. This chip reportedly includes 3D V-Cache on both chiplets, totaling approximately 192 MB of L3 cache. It is said to operate at a base frequency of 4.30 GHz, with a boost up to around 5.6 GHz, and has a TDP of about 200 W. The 9950X3D2, with its new naming convention, suggests a tradeoff of slightly lower peak boost speeds (-100 MHz) compared to the single-stack X3D Ryzen 9 9950X3D model, but offers significantly increased cache capacity. This combination could be a sweet spot for cache-sensitive workloads and certain games.

The expanded cache and increased thermal headroom may necessitate more attention to cooling and VRM design for system builders, potentially leading AMD to price these chips at a premium or adjust current MSRPs. A companion model, the Ryzen 7 9850X3D, is expected to follow a similar design with eight cores, 96 MB of L3 cache, and a boost up to 5.6 GHz, with a 120 W power envelope. The 9850X3D appears to be a refined, higher-clocked (+400 MHz) alternative to existing 8-core Ryzen 7 9800X3D models, offering improved single-thread performance that is highly valued by gamers, while maintaining a relatively modest TDP.
Both processors are linked to AMD's next-generation V-Cache technology, which reportedly offers better thermal performance and is more conducive to overclocking. These advancements could explain the feasibility of dual-cache retail SKUs. However, these details are still unconfirmed, so they should be considered preliminary. If accurate, AMD's strategy would maintain competitive pressure on Intel's upcoming desktop refresh by emphasizing cache-driven performance rather than solely increasing core counts. Intel's "Arrow Lake Refresh" is on the horizon, but it lacks a 3D V-Cache equivalent until "Nova Lake."
Sources: chi11eddog on X, via VideoCardz
Add your own comment

93 Comments on AMD Could Launch 16-Core Ryzen 9 9950X3D2 CPU with 192 MB L3 Cache and 200 W TDP

#1
LabRat 891
The dual-vCache 9950 was a dream and a hope; an even higher clocked 9800X3D, was not at all on the Bingo Card :eek:
Posted on Reply
#2
kondamin
That’s about time, also will there be a 2x6 core version with that cache ?
Posted on Reply
#3
ir_cow
I'll wait for Zen6 at this point. Hopefully th IPC is 20% more and a X3D2 in its lineup.
Posted on Reply
#4
BorisDG
Zen 6 is still far away ~1 year (end of 2026) + you have to wait probably more for X3D variants.
Posted on Reply
#5
evernessince
The good old Nvidia strategy of spreading out releases of top end SKUs to milk the customer. At this point, no thank you. I would have purchased this if it was available earlier but now I'll just wait with my 9800X3D until next gen (assuming it's good).
Posted on Reply
#7
Bloste
Each chiplet will have their own 3D cache, so you don't have 16 cores with 192MB of L3 cache, you'll have 8 cores with 96MB and another 8 cores with 96MB.
Didn't AMD park for gaming one of the chiplets in their Ryzen 9 chips, precisely for this reason?
Posted on Reply
#8
harm9963
I wait for reviews , most likely will pass , just keep saving my money , long as I can play at 4K with my 4090 and 5950X 4266CL16 4X16GB with ease !
Posted on Reply
#9
Mysteoa
evernessinceThe good old Nvidia strategy of spreading out releases of top end SKUs to milk the customer. At this point, no thank you. I would have purchased this if it was available earlier but now I'll just wait with my 9800X3D until next gen (assuming it's good).
I can see that, but they would need time to get enough high clocking dies. Which makes the late release reasonable.
Posted on Reply
#10
BorisDG
harm9963I wait for reviews , most likely will pass , just keep saving my money , long as I can play at 4K with my 4090 and 5950X 4266CL16 4X16GB with ease !
If it ever comes out. This "leaker" was teasing this since August and nothing...
Posted on Reply
#11
LabRat 891
kondaminThat’s about time, also will there be a 2x6 core version with that cache ?
It looks like both of these rumored SKUs are made from the exact same vCache CCDs. *Something* has to have been revised to help facilitate these clocks. Perhaps, it's as simple as production refinement allowing more better-binned CCDs?

My point: A 9900X3D2 is unlikely at this point in time.
A 6-core CCD+vCache w/ elevated clocks could/would likely be 'salvaged' out of these CCDs.
evernessinceThe good old Nvidia strategy of spreading out releases of top end SKUs to milk the customer. At this point, no thank you. I would have purchased this if it was available earlier but now I'll just wait with my 9800X3D until next gen (assuming it's good).
Fair enough to perceive it that way.
The way I see this, it was not possible/economical to build these, prior. a +400Mhz binning on the vCache CCD is notable and indicative of changes under the hood or behind-the-scenes.
Dr. DroCome to papa :)
This. The past several days, I'd been on/off hovering over the Checkout button on a 9800X3D.

TBH, I'd assumed we'd not see dual-vCache CCD Ryzen until Zen6, so I was going to 'settle' on a 9800X3D. Now, I'm stoked for the 9950X3D2; a much more fitting "Upgrade" from my 5800X3D*

*Win10-> Builds:
i7 3770k 4c8t
R5 3600 6c12t
R5 5600 6c12t
R7 5800X3D 8c16t
BlosteEach chiplet will have their own 3D cache, so you don't have 16 cores with 192MB of L3 cache, you'll have 8 cores with 96MB and another 8 cores with 96MB.
Didn't AMD park for gaming one of the chiplets in their Ryzen 9 chips, precisely for this reason?
The distinction mattered with the asymmetric 9900X3D and 9950X3D. This 9950X3D2 is symmetrical.
Core Parking was to keep gaming loads away from the non-X3D cores. That should be a non-issue w/ the 9950X3D2.



What I'm particularly curious about, is how the extra memory bandwidth of the dual-CCD dual-vCache CPU will/will not be utilized. Previously, the extra RAM bandwidth from dual-CCD Ryzen was completely overshadowed Single-CCD X3D. Even though X3D CPUs 'hit RAM' less, I wonder if there'll be any tasks where the improvements will be noticable?
Posted on Reply
#12
Wirko
kondaminThat’s about time, also will there be a 2x6 core version with that cache ?
It would be a bit strange but not extremely so. Think of the Epyc 9184X with 3D V-cache on all 8 chiplets and *two* cores per chiplet. Actually it would deserve the, uhm, "FX" suffix since "F" Epycs have Few Fast cores.
Posted on Reply
#13
Dr. Dro
LabRat 891It looks like both of these rumored SKUs are made from the exact same vCache CCDs. *Something* has to have been revised to help facilitate these clocks. Perhaps, it's as simple as production refinement allowing more better-binned CCDs?
AMD didn't release this before simply because they didn't want to. No binning required, they could build this as far back as Zen 3 without any additional effort.
Posted on Reply
#14
R-T-B
LabRat 891It looks like both of these rumored SKUs are made from the exact same vCache CCDs. *Something* has to have been revised to help facilitate these clocks. Perhaps, it's as simple as production refinement allowing more better-binned CCDs?
IIRC it's that the vcache is now in the bottom of the chipstack and not the top, or something. Why that changes everything eludes me too.
Posted on Reply
#15
Axiomatic
What is sad is the Microsoft implementation of the PPM driver tied to the stupid Windows 11 game bar. The Linux implementation is far superior and is not tied to some silly game bar, but in the kernel where it should be. We should all be collectively complaining about this to Microsoft/AMD.
Posted on Reply
#16
TheGeekn°72
imma be honest, I'd rather see 10-12 cores CCDs with V-cache, gosh I hope Zen6 will have some other leak soon...
Posted on Reply
#17
LabRat 891
WirkoIt would be a bit strange but not extremely so. Think of the Epyc 9184X with 3D V-cache on all 8 chiplets and *two* cores per chiplet. Actually it would deserve a "FX" suffix since "F" Epycs have Few Fast cores.

:roll:

Besides Ford bringing back the "Pinto" namesake, I cannot think of a more-unlikely to return branding, than AMD FX.
Dr. DroAMD didn't release this before simply because they didn't want to. No binning required, they could build this as far back as Zen 3 without any additional effort.
The dual-vCache CCD part, yes. However, AMD's been pretty open about how it was just not an economical product (at the time).

These rumored new Ryzens, are clocked higher than any vCache SKU prior, in a still-manageable TDP range. Absolutely, these would better-binner or revised 9800X3D dies.

Looking back towards AM4 and a then-prospective 5950X3D, clocks would've had to be reduced so much to stay within power/thermal envelope, it wouldn't have performed competitively.

Basically,
How/why is AMD now finding it economical to put 'Enterprise grade' silicon in a consumer product?



IMO,
Moving the vCache underneath the CCD made this rumored 9950X3D2 possible.
The 9800X3D was the top economical consumer-marketable binning for a Zen5X3D CCD.
Market pressures and/or production improvements have now allowed/encouraged an even higher-binning to be consumer-marketed.
R-T-BIIRC it's that the vcache is now in the bottom of the chipstack and not the top, or something. Why that changes everything eludes me too.
That's been a key part of Zen5 X3D, in general. The 9850X3D, is more like the 9800X3D getting the "XT" treatment, while adding a new halo-tier SKU to the lineup with the 9950X3D2.
Posted on Reply
#18
Panther_Seraphin
Dr. DroAMD didn't release this before simply because they didn't want to. No binning required, they could build this as far back as Zen 3 without any additional effort.
Combination of things.

5xxx Series IO Die would have been such a bottleneck it wouldnt have been funny

7xxx series had to give up too much clock speed due to the cache CCD being an insulator affecting thermals

9xxx series still has "issues" around the IO die and cross CCD communication but I bet yields have got to the point of "why not" and certain work loads will get a benefit. I still think games are going to be an issue in the majority of time just due to the nature of how games are extremely latency sensitive.
Posted on Reply
#19
mb194dc
They probably can but in real world use cases there's probably very little point in it. Most importantly, should use it in those real world scenarios rather than relying on benchmarks which could be misleading.
Posted on Reply
#20
GenericUsername2001
R-T-BIIRC it's that the vcache is now in the bottom of the chipstack and not the top, or something. Why that changes everything eludes me too.
The 9000 series x3d chips have the cache chip under the CPU chiplet, while prior x3d chips had it on top of the CPU chiplet. As for why that makes a difference, its pretty simple - heat dissipation. The older x3d chips could not clock as high as the non-x3d versions since having the cache chiplet (which itself does not use much power) in between the cores & heat spreader reduced how much heat they could get rid of. Having the cache chip on the bottom makes getting rid of heat easier, so they can clock higher.
Posted on Reply
#21
Wirko
Dr. DroAMD didn't release this before simply because they didn't want to. No binning required, they could build this as far back as Zen 3 without any additional effort.
Binning is always required, and it must be extremely complex at AMD. You can't just assume they have enough stock of chiplets good enough for product X but not good enough for higher-margin products Y and Z. Me thinks they use AI for binning.
Posted on Reply
#22
Space Lynx
Astronaut
BorisDGZen 6 is still far away ~1 year (end of 2026) + you have to wait probably more for X3D variants.
i think they already said early 2027 actually. they have no competition, and when you have no competition... you milk it
Posted on Reply
#23
Vayra86
Well I suppose this will definitively knock the bread out of the Intel relativists' mouths, then.

This is what scalable architectures do.
Posted on Reply
#24
Wirko
By the way, there's a nonzero probability this is going to be an asymmetrical Ryzen. 160 MB + 32 MB. Ha. Perhaps AMD's HPC customers demand even more V-cache on Epyc processors. TSMC said long ago that their hybrid bonding can be done in multiple layers. If true, AMD will have gamers test the new advanced packaging technology... again (it was the same with the first iteration of X3D).
Posted on Reply
#25
LabRat 891
WirkoBy the way, there's a nonzero probability this is going to be an asymmetrical Ryzen. 160 MB + 32 MB. Ha. Perhaps AMD's HPC customers demand even more V-cache on Epyc processors. TSMC said long ago that their hybrid bonding can be done in multiple layers. If true, AMD will have gamers test the new advanced packaging technology... again (it was the same with the first iteration of X3D).
So, you're saying there might be an "X3D Sandwich" CCD? :laugh:
Correction: I believe the stack would still have to be at the bottom. TSV'd up through each cache stack to the CCD.
Posted on Reply
Add your own comment
Nov 21st, 2026 07:28 PST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts