Intel GPU Validation Inconclusive

Message boards : Number crunching : Intel GPU Validation Inconclusive
Message board moderation

To post messages, you must log in.

Previous · 1 · 2

AuthorMessage
Profile Michael H.W. Weber

Send message
Joined: 13 Aug 05
Posts: 3
Credit: 3,250,984
RAC: 0
Germany
Message 1809249 - Posted: 15 Aug 2016, 8:58:16 UTC
Last modified: 15 Aug 2016, 8:59:47 UTC

Well, it is clear that a dedicated GPU will outperform Intels IGPs.
The question is rather whether usage of an OpenCL-capable IGP can compete with the usage of a CPU core of that same system (of which one usually has to be kept free to "fire" the GPU).

Michael.

P.S.: In my hands, so far neither an HD 4000 nor an HD 5500 IGP running at full load with SETI will cause any lags or crashes even when working regularly on that same machine (browsing, office, etc.).
ID: 1809249 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13742
Credit: 208,696,464
RAC: 304
Australia
Message 1809251 - Posted: 15 Aug 2016, 9:04:14 UTC - in response to Message 1809249.  

Well, it is clear that a dedicated GPU will outperform Intels IGPs.
The question is rather whether usage of an OpenCL-capable IGP can compete with the usage of a CPU core (of which one usually has to be kept free to "fire" the GPU).

It depends very much on the core, and the iGPU. Generally it can outperform a single core, but not multiple cores.
And one of the biggest issues is power/heat. The iGPU shares it's power & heat limits with the CPU- the end result is you can run the iGPU flat out, but loose the output of the CPU cores to do so; and without those cores you generally end up with less work being done than just running the CPU cores alone & no IGPU crunching.
Add to that Intel's rather poor (at best) support for OpenCL, so at this point in time iGPU crunching really isn't worth it IMHO.
Grant
Darwin NT
ID: 1809251 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6325
Credit: 106,370,077
RAC: 121
Russia
Message 1809253 - Posted: 15 Aug 2016, 9:05:29 UTC - in response to Message 1809249.  

Well, it is clear that a dedicated GPU will outperform Intels IGPs.
The question is rather whether usage of an OpenCL-capable IGP can compete with the usage of a CPU core of that same system (of which one usually has to be kept free to "fire" the GPU).

Michael.

P.S.: In my hands, so far neither an HD 4000 nor an HD 5500 IGP running at full load with SETI will cause any lags or crashes even when working regularly on that same machine (browsing, office, etc.).


It's complex topic. And strong dependence from iGPU generation here.
For introduction look this experiment: http://lunatics.kwsn.info/index.php/topic,1735.0.html
Worth to repeat it for other iGPU families.
SETI apps news
We're not gonna fight them. We're gonna transcend them.
ID: 1809253 · Report as offensive
Previous · 1 · 2

Message boards : Number crunching : Intel GPU Validation Inconclusive


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.