• 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 676 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
yottabit said:
a CPU that excels on unoptimized, branchy code.
Click to expand...
Instead of real 8-wide, they split it up into two 4-wides. I think they did it that way for SMT uplift so that the virtual core can boost some other branchy thread in parallel. An optimization could be that you deliberately write code that tries to stay within L2 cache and uses thread synchronization between the real and virtual core to eat its way through gobs of data.
 
Thanks for these. Is it even possible to lock frequency for an R23 run? Not sure what options are available for mobile. 4 GHz would be perfect to compare points/GHz to other chips.
 
Philste said:
Ugh. That's as fast as a 7840U. Kraken will be a total dud with 4+4. Probably will be slightly slower in both CPU and GPU compared to Phoenix.
Click to expand...

You dont know what is the power used by his device, so that s just speculations on a vaccum, it s 100% sure that a 7840U use quite more power for this score, eventually roughly 50% more power.

Edit : It should be at barely 0.5x the power for a same score as a 7840U.

IPC uplift is 17%, core count is 1.25x, so that make about 1/(1.17 x 1.25)^2.3 = 0.42x the cores power at same perf than a 7840U.
 
Last edited:
StinkyPinky said:
Silverbench P60424 Normal , X5194 Extreme
Cinebench 14242 Multicore singlepass
Cinebench 1939 Singecore singlepass

Should be noted that this is a *very* thin laptop, it's not designed to eke out every percent of performance.
Click to expand...
Can you try Cinebench 2024?
 
StinkyPinky said:
Should be noted this is a very thin laptop with what I suspect may be a quiet fan profile (I will look in the bios in a minute). This laptop is Macbook Air thin.
Click to expand...
Well it has dual fans and Asus says 28W. It could be a bit lower than that in that specific setting, but i dont believe it when its plugged in. 7840U has 13000 something at 32.5W short burst and 28W sustained in Thinkpad T14 G4 according to Notebookcheck.
 
Philste said:
Well it has dual fans and Asus says 28W. It could be a bit lower than that in that specific setting, but i dont believe it when it’s plugged in. 7840U has 13000 something at 32.5W short burst and 28W sustained in Thinkpad T14 G4 according to Notebookcheck.
Click to expand...
I agree with you Strix Point looks to be a minor improvement in CPU performance over HawkPoint.
 
StinkyPinky said:
Ok, this thing runs hot (according to the pretty basic bios). Idling in the 80s. That's a bit odd.
Click to expand...
That’s not normal. I would return it. Might be shoddy thermal paste. Even Apple fanless M chips idle at 40C.

Edit: didn’t know you were in the bios, my bad.
 
Philste said:
Well it has dual fans and Asus says 28W. It could be a bit lower than that in that specific setting, but i dont believe it when its plugged in. 7840U has 13000 something at 32.5W short burst and 28W sustained in Thinkpad T14 G4 according to Notebookcheck.
Click to expand...

I agree that I don't think it's a massive performance boost, but just pointing out this probably isn't the best device to get the full picture.

I wish the Bios was better, it's very basic. No fan profiles or anything, even in advanced mode.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top