What is favorable to OS?
Is OC of a memory effective?
Is it better to narrow the amount of consumption of a memory down?

Edited January 29, 20206 yr by GTI-R

For older platforms either XP or Vista, not sure about new hardware.

Memory OC does not really help at all.

I haven't tested reducing maximum memory either.

Traditionally thought of as a bench where mem doesn't really make any difference.

XP is still good for newer hw I think.

Mem matters a little in my testing, but basically to the extent that if your mem doesn't suck you're good. No need to crank timings or whatever like in superpi, much better to go for max possible core speed.

Isn't a good result obtained in Windows10?

30 minutes ago, GTI-R said:

Isn't a good result obtained in Windows10?

If you have a look here:https://hwbot.org/submission/4067843

It's the 8 core Global First place and it's on Windows XP.

If you look at the Global First place you can generally see what the best performing OS is.

open the bench... run a 32m quickly.. as its running... grab the window and resize it until its so small you cant see anything.

*now - run again and profit 

Place wprime.exe on desktop. Then copy it on 2-3 partitions, like C, D and so on. From my testing years ago, 3 different exe's are enough. Rename each one to wprime2.exe, wprime3.exe etc and send shortcuts to desktop to make your life easier. Have Awardfabrik tweaker or any other similar program running so that each time you open either of the exe's all instances of wprime are in real time priority. Run several times the desktop one until you see no more improvements. Close it and open the exe you have in C or D for example. Run several times until you get each best run. Continue doing the same process by using different exe's until eventually you'll get the best run with the most performing wprime.exe 

Zafiropo has a video showing the tweak in action without using the awardfabrik tweaker. We were running a Sandy Bridge 4500MHz challenge back in summer 2011 and none of us were able to do less than 6.702sec in wprime32m. In the end he posted a 6.671sec run and won the whole thing  

 

 

 

Is the environment that a bench Mate is required effective?

Why is the score of 9900KS lower than the score of 8700K?
Is there a problem just for me?
It ’s inside the benchmate.

Because your CPU is R0. Micro code it requires kills wprime performance.

I see.
It was understood.

PiFast is also the same situation.
Is 8 the influence by which this is also R0 stepping?

I did not try pifast, but possible. You can play affinity, usually core 02 or 03 best on 9900ks