cancel
Showing results for 
Search instead for 
Did you mean: 

Realbench v2 [Discussion Thread/Download Links]

X-ROG
Level 15
278 Views
522 REPLIES 522

MMikhail wrote:
What do you think about this method of getting cpu usage from window?

in console
wmic cpu get loadpercentage


you can also retrieve cpu and it's clock speed by this method


WMI uses permon counters as well. It is just another method of delivery for the same thing. So your suggestion boils down to "use the same thing but with WMI as middleware which adds processing overhead for no reason". You should really stop talking about things you have no understanding, at all, about.

Anyhow, the 2.54 release version now uses the same method as the task manager, reverse engineered from it. So now it will run in corrupted Windows installations, like yours, and you can now stop polluting this thread with misinformation and random imaginary stuff.

The downside to using undocumented APIs is that any new version of Windows or update may break this functionality without warning. This is why perfmon counters were used in the first place which are the standard practice and would always work properly. Still, now the CPU counter will work on corrupted Windows installations. Case closed.

MikeS3000 wrote:
On RB 2.44 when I would run a stress test using an AVX offset of 1, I would see in HWINFO64 the core speed would bounce around on all 4 cores (i.e. some cores doing non-avx workloads at higher clock speeds and other cores downclocking to my AVX offset under AVX workloads) On 2.54 RC2 all cores are pinned at my AVX offset frequency. I liked the ability to run a mixed workload that tests my cores at max freq non-avx and also avx downclocked frequencies. What was changed in this version that changes the entire workload to AVX code?


Handbrake has been updated. As you found out it now uses more efficient parallelization.
RAMPAGE Windows 8/7 UEFI Installation Guide - Patched OROM for TRIM in RAID - Patched UEFI GOP Updater Tool - ASUS OEM License Restorer
There are 10 types of people in the world. Those who understand binary and those who don't!

RealBench Developer.

The current solution seems to be working.
The WMI method is working fine, so the old method had flows if it used the same tech.
And for future builds, it might be good to lock the cpu counter to two decimal points (currently it can go as hight as 5 points)
and fix the alignment (right justification) so that the figures don't jump all over the place but show a steady XXX.XX counter
same for the degrees, but those are usually double digits anyway for most people.

Image

Is anyone els getting really low scores on v2.54?

Nodens wrote:
Handbrake has been updated. As you found out it now uses more efficient parallelization.


Can that be the reason why I need to re-adjust my cache multi and/or cache voltage?
With the settings I run with previous version Handbrake is crashing now...
Left the voltage and lowered the multi, running fine now - just was curious if it is caused by the changes in the software?

Arne_Saknussemm
Level 40
I had to re-tune OC for RC1 and later versions too...

Zka17
Level 16
Good then... no probs here! :cool:

DooRules
Level 10
So just to be sure guys the download link from leaderboard with version 2.54 is the latest and not version 2.54RC2???
7980XE on Asus ApexVI
Zotac 2080Ti
Intel 900P
2 x Samsung 950 Pro
EVGA 1600W
Alienware 34"
Samsung 28" 4K

DooRules wrote:
So just to be sure guys the download link from leaderboard with version 2.54 is the latest and not version 2.54RC2???


The leaderboard link RealBench V2.54 is now the latest version available - do not use RC1 and RC2 anymore.

Thanks all for your feedback, its highly appreciated and keep the benchmark improving.

Thanks Nodens for your hard work too! The benchmark is really awesome.
And don't forget, Keep pushing it !

Is 4 hours stress test good to be called stable????*
I had to re-tune my Cpu voltage for the RealBench 2.54 to be stable.......