Jump to content
HWBOT Community Forums

why hwbot uses wprime?


jurek

Recommended Posts

imagine that 3dmark adjusts its settings to the gpu u use to give the best possible score... everyone would say cheating... wprime does the same and it is valid hwbot benchmark. why? shouldnt benchmark give everyone same settings? i vote ban for wprime :P

Link to comment
Share on other sites

I assume he means the dual- versus single- versus quad-core logic.

 

It's a valid benchmark because benchmarks are meant as proxies for real application performance. A multi-threaded real application would be able to recongifure for the number of live cores.

 

You could get the same results by rigging wPrime's advanced settings to 1 core for all CPUs. Of course, that wouldn't accurately reflect real-world multi-threaded performance.

Link to comment
Share on other sites

I vote for proof of what you are typing.

 

as from http://www.wprime.net/

 

Threading

Our aim was to make a perfectly threaded benchmark, such that it would consistantly use 100% of the CPU while in use. This is achieved by using CPUz to detect the CPU count and use exactly that many processing threads to avoid any performance losses due to multiple threads running on any single physical thread. Each thread is designed to do 1/n of the work, where n is the number of threads. For example, if you're calculating 16 roots on 4 CPU's, each CPU will calculate 4 roots. Some might argue that this style of threading is unrealistic in real-time performance, but in fact is quite indicative of performance in several real world tasks such as F@H which allows you to run several instances of the work at any one time.

 

I assume he means the dual- versus single- versus quad-core logic.

 

It's a valid benchmark because benchmarks are meant as proxies for real application performance. A multi-threaded real application would be able to recongifure for the number of live cores.

show me such application :P

 

You could get the same results by rigging wPrime's advanced settings to 1 core for all CPUs. Of course, that wouldn't accurately reflect real-world multi-threaded performance.

 

or maybe it would :P

Link to comment
Share on other sites

your seriously asking if there are applications and games which are multithreaded? Knowing that AMD, Intel, ATI, NVIDIA and others are all developing multi-CPU/GPU solutions which require multithreaded applications to show their maximum potential? :eek:

 

3DMark06 CPU test is multithreaded; if you want to block Wprime because it can use more than 1 CPU core; you have to block PCMark, 3Dmark06 and Cinebench and other applications which will be released in the future.

 

 

In short: not going to happen, Multithreaded applications and benchmarks are here to stay. There are plenty of games which work much better with 2 cores vs 1 core (Unreal 3 engine, Crysis, SupCom, etc) and even more applications (any audio/video rendering suite).

 

u missed one point... im not against it because of multithreading (pls read more with understanding) but because of

Threading

Our aim was to make a perfectly threaded benchmark, such that it would consistantly use 100% of the CPU while in use. This is achieved by using CPUz to detect the CPU count and use exactly that many processing threads to avoid any performance losses due to multiple threads running on any single physical thread. Each thread is designed to do 1/n of the work, where n is the number of threads. For example, if you're calculating 16 roots on 4 CPU's, each CPU will calculate 4 roots. Some might argue that this style of threading is unrealistic in real-time performance, but in fact is quite indicative of performance in several real world tasks such as F@H which allows you to run several instances of the work at any one time.

i dont like repeating myself man :/

 

pls stop answering if u dont know what is discussion about

Link to comment
Share on other sites

i dont like my attitude either just wanted to point that in my opinion a benchmark should be a benchmark for every component providing the same conditions ;) i guess its why its named benchmark and if other settings are used for different hardware i guess a word "benchmark" loses its meaning ;)

 

and btw u wanted a proof and i gave it to u ;) what u want more?:P

Link to comment
Share on other sites

I see your point. It would have been better if wPrime assigned eg 4 or 2 threads per dedected core, instead of 1, I agree. But it does not make it a seriously flawed benchmark program.

 

Maybe something to talk over with the author for a next version.

 

i have a solution which isnt perfect but... lets say that now u have max 4 cores make universal wprime with 8 or more threads, for example 16 or 32, threads then for future cpus

 

after introducing more cores there can be new version of w prime made that will be for those new cpus... but changing version isnt good ;) so its better to make one version with many threads like 32 or 64 this will work for quite a time i guess and still preffer multi core cpus over those with less cores

 

you say it cheats, but how does it cheat? I still don't follow you Jurek :(

 

maybe word "cheat" isnt adequate to the situation... but still changing settings to best suit benched hardware is at least doubious for benching application ;)

Link to comment
Share on other sites

Sounds like some doesn't own a quad core to me.

 

But seriously it you think it is a cheat because it adjusts itself to use more cores because it is available then why aren't 3d benches the same, I've heard that there is something called crossfire and something else called I think SLi. Graphical benches use these setups to gain bigger scores... amazing :eek:

Link to comment
Share on other sites

By following the same logic for GPUs, we must ban 3dmark05 and 06 until it exists a patch that makes the CPU tests run in the same way as you want wprime to run, as well as running multiple "3d-threads" (1 per gpu at least!).

 

I don't see the problem with a program like this. It gives you a valid score, and it uses 100% of the CPU power available.

 

Btw, benchmarks compare one CPU to another. You can count cores, or look at the frequency, or any other relevant difference - what makes the number of cores so much more important than everything else? I don't think comparing a quad to a dual core is much worse than comparing a PIII to a Celeron-L. 3dmark compares 1vs2 as well. No way you'll get a top score in 06 with a single or dual core, and same with 05 for single cores.

 

Oooops, soon most of the benchmarks are gone:D Only 01, 03, AM3, spi1m + 32m, pifast and cpuz left:rolleyes:

Link to comment
Share on other sites

i dont see anything wrong with what wprime does. its using 100% of the cpu what more could you ask? if it didnt use 100% of the cpu then what? 95%? 5%? Random %? Looking at your logic I dont think you understand what your asking.

 

Not to mention acting like an @$$ to everyone certainly wont further your ideas, I can guarantee you that much. Lighten up buddy :D

 

p.s. u dont want to argue with jmke, he will win lol

Link to comment
Share on other sites

Running wprime with 8 threads on a quad have been faster than 4 threads for me. So I wouldn't bother if it was locked at 8 threads. The problem with wprime1024 is that on thread usually finish a lot later than the rest, but by having more threads this gap closes.

Link to comment
Share on other sites

  • 2 weeks later...
but it adjusts number of threads to cores :/ are there user applications which do this?

 

When you run F@H (or almost any other distributed processing task) on a multithreaded system - you run as many processes as you have logical threads for optimal efficiency.

 

Likewise, any well-written encoding task acts in the same way, if hdd isn't a bottleneck in such a scenario, it'll scale similarly as well.

 

The single threaded user is free to adjust it to use as many threads as he wants, but he won't benefit from it (unless his CPUs aren't all the same speed, eg. HyperThreaded CPUs or if you have background processes slowing things down).

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...