They're comparing DT part with gamer stable memory against mysterymeat RVPs.Why does that matter at all when every sample was tested with the same RAM?
Not much to extract from an half baked bench but still an observation :Comparing this now to the 2 scores from earlier (ignoring comparisons to Intel CPUs for now since GB seems to favor them), you get:
Zen 5 = 90% higher integer PPC than Zen 4 if 1.4 GHz results are true. (Edit: and 83% in FPU)
or
Zen 5 = 16% higher integer PPC than Zen 4 if 2 GHz results are true (Edit: and 51% in FPU).
Its literally two comparisons, using both standard and OC specs. It shows that there is a 0.12 point difference between the different architectures and that their IPC is identical when you consider a margin of error.They're comparing DT part with gamer stable memory against mysterymeat RVPs.
Zen 4 - 6 wide dispatch/renameComparing this now to the 2 scores from earlier (ignoring comparisons to Intel CPUs for now since GB seems to favor them), you get:
Zen 5 = 90% higher integer PPC than Zen 4 if 1.4 GHz results are true. (Edit: and 83% in FPU)
or
Zen 5 = 16% higher integer PPC than Zen 4 if 2 GHz results are true (Edit: and 51% in FPU).
Top one seems unbelievably high and bottom one seems more believable but also too low given the breadth and magnitude of chan odges in the design.
Both are ES samples of unknown origin so I will now go back to waiting on more official numbers, lol.
Strix Point ~= RTX 3050 35W - 50WWill Strix Point iGPU be able match/exceed RTX 3050 mobile?
This might be a hot-take. In my opinion, a 128 bit APU should be able to match atleast a 2 generation old RTX xx50 dGPU.
Your Zen4 result is winning in many subtests by 25-50%. I have done some comparisons with low clocked Phoenix and got similar weird results.
I did it, lscpu shows 2.6 Base for my 7950X, but cpufreq reports 2G, I don't have GB account, so maybe someone can check the frequencies
My RAM is 6000MT/s
Bash:cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_cur_freq 400000 400000 400000 1999222 400000 400000 400000 2000130 400000 2000078 1998810 400000 400000 400000 400000 400000 400000 400000 400000 400000 400000 2000082 400000 1999932 400000 400000 400000 400000 400000 400000 1999921 1999553
STX is not recognized properly, L1 and other data seems completely off.
This is a more conservative report unlike this one @1.4G here https://browser.geekbench.com/v5/cpu/compare/22423185?baseline=22425328
If these are spoofs then well... waste of time
why pick something arbitrary like dispatch.Zen 4 - 6 wide dispatch/rename
Zen 5 - 8 wide dispatch/rename +33%
If this results in +15-20% higher IPC, it will still be very good.
I think that the first generation of Zen with 6xALU, i.e. +50%, if it translates into an IPC increase of +15-20%, will also be good.
I dare say that AMD will not add ALU for the next 2 generations of Zen6-7 and IPC increases will also be 10-20%.
To everyone. I am waiting for official benchmarks by multiple sites after its live, before I will venture any opinions.Your Zen4 result is winning in many subtests by 25-50%. I have done some comparisons with low clocked Phoenix and got similar weird results.
The Strix sample has either borked microcode or the boosting algorithm is doing something weird.
I'm sorry for asking for something dumb but what's with all the unfunny clocks?
Why do all the first tests run at such low clocks? Is there someone slowly pushing the CPUs to the max somewhere at AMD or an AIB now, when it's just about to come out in 2 months?
I'm sorry for asking for something dumb but what's with all the unfunny clocks?
Why do all the first tests run at such low clocks? Is there someone slowly pushing the CPUs to the max somewhere at AMD or an AIB now, when it's just about to come out in 2 months?
50% more ALU does not guarantee 30-40% higher IPC, 50% more L1-D does not guarantee 30-40% higher IPC, 33% more Shipping/name change does not guarantee 30-40% higher IPC.why pick something arbitrary like dispatch.
AMD ran wider retire then dispatch because it was more of a bottleneck for example.
Look at the entire core , we are generally looking at massively big ROB's , something big on the decode side ( we are still guessing ) , significantly wider int execution , more L/S , large L1D increase.
People are betting (hoping) on Bulldozer level execution. that could happen , but im not betting on it.
Looks like 2800-2900 for 5.5 GHz sample.
Second GB5 score now as well
AMD BIRMANPLUS - Geekbench
Benchmark results for an AMD BIRMANPLUS with an AMD Eng Sample: 100-000000994-14_N processor.browser.geekbench.com
Not sure if it's even real though tbh.
These results are consistent.
50% more ALU does not guarantee 30-40% higher IPC, 50% more L1-D does not guarantee 30-40% higher IPC, 33% more Shipping/name change does not guarantee 30-40% higher IPC.why pick something arbitrary like dispatch.
AMD ran wider retire then dispatch because it was more of a bottleneck for example.
Look at the entire core , we are generally looking at massively big ROB's , something big on the decode side ( we are still guessing ) , significantly wider int execution , more L/S , large L1D increase.
People are betting (hoping) on Bulldozer level execution. that could happen , but im not betting on it.
2800 - 2900 in G5 ST? That's exceptionally high wow!Looks like 2800-2900 for 5.5 GHz sample.
This result also reports more memory and seems to detect the cache a bit better (for all cores). The reported 5.1 Ghz base clock seems interesting, but is probably BS.At 5.8G it will go beyond 3K, which is pretty awesome, around 21%-24% improvement on Stepping 0, may gain a couple of percent with new steppings and microcode updates.
I'm not sure about that result you're looking at, but a 7950X at 3.1GHz on Linux scores ~2200pts (EDIT: wrong number, ~1300pts). If you use that as a basis, you get 22.3% IPC uplift.This result also reports more memory and seems to detect the cache a bit better (for all cores). The reported 5.1 Ghz base clock seems interesting, but is probably BS.
I don't get where you got that 21-25% IPC gain? Using linux a good 7840U would score a little over 2000 points at 5.0 Ghz , meaning about 1058 points at 2.6 Ghz. This makes it more of a 15% IPC gain.
But since we only have a single sample, there really isn't much to draw conlcusions on just yet.
The samples are getting 1024@2G and 1224@2.33G. My 7950X gets 2300@5.75G Linux and 840@2G.But I don't get where you got that 21-25% IPC gain? A good 7840U score would be a little over 2000 points at 5.0 Ghz in linux, meaning about 1058 points at 2.6 Ghz, this is a 15% IPC gain.
Mistake mate, but I know you meant ~5.7Gbut a 7950X at 3.1GHz on Linux scores ~2200pts
Wrong number, I actually meant to write ~1300pts at 3.1GHz.The samples are getting 1024@2G and 1224@2.33G. My 7950X gets 2300@5.75G Linux and 840@2G.
AMD BIRMANPLUS - Geekbench
Benchmark results for an AMD BIRMANPLUS with an AMD Eng Sample: 100-000000994-14_N processor.browser.geekbench.comAMD BIRMANPLUS - Geekbench
Benchmark results for an AMD BIRMANPLUS with an AMD Eng Sample: 100-000000994-14_N processor.browser.geekbench.com
I excluded the sample which got 1224@1.4G
Mistake mate, but I know you meant ~5.7G
How does that translate to GB6 ST?2800 - 2900 in G5 ST? That's exceptionally high wow!