Same folding performance at either 50% or 100% CPU usage

Moderators: Site Moderators, FAHC Science Team

Post Reply
Mr J
Posts: 4
Joined: Thu Jun 20, 2013 11:50 pm

Same folding performance at either 50% or 100% CPU usage

Post by Mr J »

Hello dear folders,

I am new to folding, finding it fascinating and keen to contribute as much as possible so after initially running on full (100% cpu usage) all the time I realised something when playing around.

After examining how performance changes as I switch between Full/Medium/Low, I realise there are none.

I am observing the PPD, TPF, time it takes to reach 0.1% completion (if I time it manually) and it doesn't matter which speed setting I am on, it won't have an effect on any of these.

I am monitoring CPU usage with OCCT and at I am at 50% low / medium 75% / full 100% of total cpu usage and the same is shown in windows task manager for the F@H.exe. I tried monitoring for longer and longer periods of time and still noticed nothing change. I decided to post here for that reason for the hopes of somebody to enlighten me :)

This confuses me as I am wondering where the extra 50% usage is going, except added heat and power drawn at this moment I wouldn't know.

My main query is that I am assuming my CPU is only being utilized 50% max somehow, and I wish to get it up to a true 100%.

I am using the 7.3.6 client on windows 7, with a i5 4670k processor, I have just the 1 folding slot with cpu threads set to -1 letting the client automatically choose. Actually all the setting are default except for that the checkpoint is 5 minutes . I am on my second WU.

Thank you for reading, wishing you all successful and continuous folds, after recently upgrading from a 5 year old unit and only just finding out about folding, I have a lot to learn yet, so it could very well be a simple error on my side.

Any thoughts or anything to say in general is more than welcome and well appreciated.
Jesse_V
Site Moderator
Posts: 2850
Joined: Mon Jul 18, 2011 4:44 am
Hardware configuration: OS: Windows 10, Kubuntu 19.04
CPU: i7-6700k
GPU: GTX 970, GTX 1080 TI
RAM: 24 GB DDR4
Location: Western Washington

Re: Same folding performance at either 50% or 100% CPU usage

Post by Jesse_V »

Here's some relevent documentation that might be helpful to you:
http://folding.stanford.edu/English/V7- ... tion#ntoc3
http://folding.stanford.edu/English/V7- ... iate#ntoc5
Joe_H
Site Admin
Posts: 7990
Joined: Tue Apr 21, 2009 4:41 pm
Hardware configuration: Mac Studio M1 Max 32 GB smp6
Mac Hack i7-7700K 48 GB smp4
Location: W. MA

Re: Same folding performance at either 50% or 100% CPU usage

Post by Joe_H »

If you could post your log file, that would help us in figuring out what is going on with folding on your system. Directions on how to locate the log file and posting it are available here.
Image
bruce
Posts: 20824
Joined: Thu Nov 29, 2007 10:13 pm
Location: So. Cal.

Re: Same folding performance at either 50% or 100% CPU usage

Post by bruce »

To estimate the points, the client needs to estimate the Time Per Frame. To establish a meaningful average TPF, the client normally takes the mean of three 1% frames. Whatever projections you're seeing for 0.1% changes are strictly estimates and not based on brief periods of 50% or 75% or 100%. You'll need to run your measurements over a period of several hours.

FAH is designed to use UNUSED resources, so it yields to anything else running on the computer. Most foreground activity uses very little of the resources that FAH uses but since they have priority, it is possible that FAH's performance may vary.

If your computer shares floating point processing between pairs of cores (Intel HyperThreading or AMD Modules) you'll generally find that the first 50% of what your OS calls CPU time is a lot more valuable than the second 50%. You'll get more processing done as you increase the number of cores, but it's not going to be proportional. The only exception to that rule is that if something other than FAH uses one or more cores continuously for long periods of time, allocating that/those same core(s) to FAH will have a negative effect.
Post Reply