lol G3258 gives really strange results, 0.03v and the ST number never seems to change with different GHz clocks.
Last edited:
It scales like this for me. (adds an average of 38 every 100 Mhz)lol G3258 gives really strange results, 0.03v and the ST number never seems to change with different GHz clocks.
That's very interesting as it shows nearly linear scaling. Did mem clock increase too?It scales like this for me. (adds an average of 38 every 100 Mhz)
4.0 - 1533
4.1 - 1572
4.2 - 1608
4.3 - 1645
4.4 - 1686
4.5 - 1725
4.6 - 1761
and apparently
5.014 - 1916 for Dufus which is pretty much, < 38 Score per 100 MHz.
It's an insider test mule. I built it back in September to use on the W10 insider program. It has had all of the builds auto-installed since then. They may not have wanted the software on there with the release builds, since I was a beta tester?
It scales like this for me. (adds an average of 38 every 100 Mhz)
4.0 - 1533
4.1 - 1572
4.2 - 1608
4.3 - 1645
4.4 - 1686
4.5 - 1725
4.6 - 1761
and apparently
5.014 - 1916 for Dufus which is pretty much, < 38 Score per 100 MHz.
I just got Windows 10 and don't have the media, so this is about as far as I care to go... don't want to bork anything and have to go back to Win7. I'm sure there are faster FX's, but I have to be in the 99th percentile. :awe:
5.278GHz (203 x 26) - 1251 / 7837
Just got a G3258 to experiment on but unfortunately have to go overseas for a while.
Just a quick run at ~1.47V, 2 cores at 29x
Noticed something odd with MT scores when affinity is set to threads -1 with 2C/4T CPU. Maybe someone can confirm.
Doesn't surprise me.It scales like this for me. (adds an average of 38 every 100 Mhz)
4.0 - 1533
4.1 - 1572
4.2 - 1608
4.3 - 1645
4.4 - 1686
4.5 - 1725
4.6 - 1761
and apparently
5.014 - 1916 for Dufus which is pretty much, < 38 Score per 100 MHz.
Today it's 1410/2372.
Only difference is going from 8gb of ram to 12gb of ram.
why wouldn't it? Intel has fantastic branch prediction and there's no non-frequency-scaling cache to hold it back.That's very interesting as it shows nearly linear scaling. Did mem clock increase too?
This has nothing to do with branch prediction as the benchmark code is very small.why wouldn't it? Intel has fantastic branch prediction and there's no non-frequency-scaling cache to hold it back.
I guess because of the RAM?
the branch prediction helps that.
http://www.planet3dnow.de/vbulletin...hmark-Update?p=5033059&viewfull=1#post5033059The bench is procedural landscape generator based on a Perlin noise. It is a mix of integer operations and single precision float operations. It is pure VC++, and I let Visual Studio optimizer do the job. As far as I can see, the x32 compiler uses x86 & x87, and the x64 version uses scalar SIMD (but I need to double check that point). The x64 version is much faster anyway. The result of the computation is a 3D landscape that will be displayed in the next-to-come GPU bench.
http://www.planet3dnow.de/vbulletin...sorvergleich?p=5033057&viewfull=1#post5033057there is indeed a problem with the dual core systems. The benchmark uses a thread pool to dispatch the tasks to the cores, and the management of that thread pool seems to occupy a core in a manner that the scheduler does not affect another task to it. So in a dual core system, two threads are created, and affected to the same core, so the result of the computation is the same as for the single thread test. I already work on a fix.
Wow, I have an A8-3500M just sitting around too, waiting to replace my 3400M. Nice OC! Mine reaches a smallish 1.7GHz OC'd, but slightly undervolted too.
Firstly, I want to thank you all for the wonderful software that many of us use daily. HWMonitor and CPU-Z are invaluable!
I am writing to ask a question in regards to the new Bench tab, and more specifically the Stress button which creates 100% utilization. Would you mind sharing how exactly the 100% utilization is achieved? I ask because IF it utilizes AVX instructions, overclocked 22 and 14nm CPUs can be damaged from current draw.
I understand if you'd prefer not to answer the question exactly, but if you could please confirm whether or not AVX is utilized in the manner that is damaging to overclocked CPUs?
Thanks!