Welcome to MilkyWay@home

cpu and gpu

Message boards : Number crunching : cpu and gpu
Message board moderation

To post messages, you must log in.

AuthorMessage
Gregory Cotreau Sr.

Send message
Joined: 6 Apr 20
Posts: 10
Credit: 15,340,408
RAC: 0
Message 69910 - Posted: 12 Jun 2020, 9:51:01 UTC

.417 cpu + 1 nivida gpu. What does .417 cpu mean?
ID: 69910 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile mikey
Avatar

Send message
Joined: 8 May 09
Posts: 3319
Credit: 520,295,454
RAC: 20,102
Message 69912 - Posted: 12 Jun 2020, 10:30:49 UTC - in response to Message 69910.  
Last modified: 12 Jun 2020, 10:32:20 UTC

.417 cpu + 1 nivida gpu. What does .417 cpu mean?


It means it's using almost 50% of a cpu core and your gpu to run a task on your gpu
ID: 69912 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Gregory Cotreau Sr.

Send message
Joined: 6 Apr 20
Posts: 10
Credit: 15,340,408
RAC: 0
Message 69913 - Posted: 12 Jun 2020, 11:26:41 UTC - in response to Message 69912.  

I have a quad core proccesser and i am running 4 other work units for a total of 5 work units
ID: 69913 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
rjs5

Send message
Joined: 18 Nov 10
Posts: 18
Credit: 175,373,177
RAC: 60,269
Message 69919 - Posted: 13 Jun 2020, 22:14:10 UTC - in response to Message 69913.  

I have a quad core proccesser and i am running 4 other work units for a total of 5 work units


You "running 4 other work units", but those 4 work units plus the GPU 0.417 CPU need are running on 4 CPU. BOINC marks all the work units as "running" but the OS will schedule the 4 CPU among the 5 work units based on the OS scheduling algorithm. The OS overhead will increase when it has to switch between work units it is trying to run.

Interestingly, I am running the GPU work units on a Nvidia 2080 ti and MW@H says that those work units require 0.997C. It takes almost a whole CPU to keep the GPU fed.
ID: 69919 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Keith Myers
Avatar

Send message
Joined: 24 Jan 11
Posts: 708
Credit: 543,143,692
RAC: 143,263
Message 69920 - Posted: 13 Jun 2020, 23:27:29 UTC - in response to Message 69919.  

Because on Nvidia cards, the science application is not as efficient as the science application on AMD cards. Due to differences in both the OpenCL API for both platforms and the difference is gpu architectures.
ID: 69920 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote

Message boards : Number crunching : cpu and gpu

©2024 Astroinformatics Group