• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)

Page 652 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
It is interesting that Granite Ridges scores noticeably higher than Strix in ST tests per GHz. ~3400pts @ 5.7Ghz ( 596pts/Ghz) Vs ~2900pts @ 5.1Ghz ( 568pts/Ghz) -> that is ~5% IPC difference between two Zen 5 implementations.
Strix got multiple known limitations, namely:
* half max L3 per core
* half (or less?) AVX512 throughput (FP benches might care)
* heterogenous CCXes (a thread might be scheduled on the Zen5c CCX)
* energy-optimized DVFS
 
half (or less?) AVX512 throughput
AMD haven't publicly said anything about Strix Point's microarchitecture differing¹ from Granite Ridge's/ Turin's, have they? ... To be fair, we are still pre-launch, so it's somewhat natural that they aren't particularly talkative about details like that for the time being.

¹) WRT FP pipelines, if true. I thought that'll only happen to Strix Halo's low power cores.
 
AMD haven't publicly said anything about Strix Point's microarchitecture differing¹ from Granite Ridge's/ Turin's, have they? ... To be fair, we are still pre-launch, so it's somewhat natural that they aren't particularly talkative about details like that for the time being.

¹) WRT FP pipelines, if true. I thought that'll only happen to Strix Halo's low power cores.
AMD public statements are driven by marketing... David Huang benched Strix Point and found the AVX512 throughput being the same as Zen 4 and, at the same time, both CCXes being architecturally identical.
 
AMD public statements are driven by marketing... David Huang benched Strix Point and found the AVX512 throughput being the same as Zen 4 and, at the same time, both CCXes being architecturally identical.
Well to be fair, he said he tested a pre-prod. laptop with ES CPU. Still, Strix Point seems to have power optimized Zen 5 core.
 
Not sure if this has been posted but found these through notebookcheck


Also through there, there are passmark results.
 

Attachments

  • IMG_0797.png
    IMG_0797.png
    206.6 KB · Views: 57
  • IMG_0798.png
    IMG_0798.png
    91.1 KB · Views: 52
  • IMG_0799.png
    IMG_0799.png
    120.6 KB · Views: 49
The dutch Amazon had the 9600X, 9700X and 9900X online for a small amount of time. The page for the 9900X was linked on WTFTech and is still available but missing the price. I still have a screenshot of the page for the 9700X, the page was gone when reloading. Anyway, there are EUR prices incl. VAT (USD prices excl. VAT in parentheses):
  • 9600X: €307.35 (276.65 USD) (not absolutely sure if this was the price)
  • 9700X: €405.69 (365.12 USD)
  • 9900X: €508.49 (457.64 USD)
The 9950X wasn't listed.

I find these interesting because they were actually below launch MSRP of Zen4, for example around 90 USD lower for the 9900X.
 

Attachments

  • Screenshot 2024-07-22 at 12-36-18 AMD Ryzen™ 7 9700X processor 8 kernen_16 kernen_16 threads Z...png
    Screenshot 2024-07-22 at 12-36-18 AMD Ryzen™ 7 9700X processor 8 kernen_16 kernen_16 threads Z...png
    1 MB · Views: 42
One can wonder if this will be a problem in newer games that might use more than 4 cores. I mean according to AMD X3D get only one V-cache die, as cross CCD latency would be prohibitive, and here we have cross CCX latency that doesn't look too good. Could it be that Hawk Point may do better in some games than Strix Point?
 
The dutch Amazon had the 9600X, 9700X and 9900X online for a small amount of time. The page for the 9900X was linked on WTFTech and is still available but missing the price. I still have a screenshot of the page for the 9700X, the page was gone when reloading. Anyway, there are EUR prices incl. VAT (USD prices excl. VAT in parentheses):
  • 9600X: €307.35 (276.65 USD) (not absolutely sure if this was the price)
  • 9700X: €405.69 (365.12 USD)
  • 9900X: €508.49 (457.64 USD)
The 9950X wasn't listed.

I find these interesting because they were actually below launch MSRP of Zen4, for example around 90 USD lower for the 9900X.
Because no one buys the 12 cores Zen. 😛
 

That doesn't sound good.
That shop seems to practice prohibitive prices. A much larger shop in France, which is not even competitive price wise, lists 7900x at 480 Euros vs pc21.fr 620 Euros. Amazon France is at 390 Euros.

I wouldn't deduce anything from pc21 prices.
 
CPUs have gotten too expensive. I still remember buying the 6600K for 230€. Now a similar range CPU goes for 300-350€. That's a 40-50% increase
Everything got more expensive but at least we get more performance now compared to the skylake era
 
Still waiting for AMD to fumble the ball on Zen 5 pricing. The urge to milk early adopters has got to be very hard to resist for them coming from Zen 3 and Zen 4. Maybe they've finally learned their lesson. 9 days left to find out.
 
Still waiting for AMD to fumble the ball on Zen 5 pricing. The urge to milk early adopters has got to be very hard to resist for them coming from Zen 3 and Zen 4. Maybe they've finally learned their lesson. 9 days left to find out.

AMD's client profitability is very crappy, considering how much better their products are than Intel's. They gotta do something. I'd expect at least Zen 4's original MSRP.
 
AMD's client profitability is very crappy, considering how much better their products are than Intel's. They gotta do something. I'd expect at least Zen 4's original MSRP.
If we forget about other markets for a second, Ian Cutress did a BOM analysis for the 7950X when it released in 2022, estimating total BOM to be around $70. For a $700 MSRP product. I have no idea how a 90% profit margin could be considered "very crappy profitability".

Based on his breakdown, the single CCD chips would be ~$45-50 BOM at the time.
 
Back
Top