Hello everyone. I'm currently running an Intel I7-950 oc'd to 3.5ghz on air cooling with a EVGA770SC gpu. I know in the past turning Hyperthreading on or off could make a difference in games. Right now I have it turned on and I'm usually above 60fps to as high as over 100fps on very high settings. I game with a single monitor at 1080P via HDMI. Using the Logitech ARX software I've been watching my CPU usage. MWO doesn't seem to use more then the 4 physical cores with only the occasional HT thread from time to time. Would turning HT off help, hurt or make no real difference in your opinions?
Thanks
0
Intel Hyperthreading On Or Off For Better Mwo Performance?
Started by JSmith7784, Sep 27 2015 05:35 AM
7 replies to this topic
#1
Posted 27 September 2015 - 05:35 AM
#2
Posted 27 September 2015 - 08:32 AM
r_WaterUpdateThread = 2 sys_TaskThread1_CPU = 0 sys_streaming_CPU = 2 sys_physics_CPU = 3 sys_TaskThread5_CPU = 3
For your user.cfg
I wouldn't turn of HT.
#3
Posted 27 September 2015 - 10:44 AM
If I update my config file what changes should I expect to see? im assuming it's telling the game to use more cores/HT threads?
#4
Posted 27 September 2015 - 12:36 PM
I would hope, but it's moar complicated then that.
The chief problem with the engine is how no one at Crytek expected maps of this scale: None of the knock-on effects where tested. It's my experaance that one of the render threads really wants a core all to itself, and that's the main point to the set I gave you.
What's that got'a do with HT? Well, HyperThreading makes a 3.0GHz core look like two 1.6GHz cores to code, and since we are trying to make the 11 named, concrete threads fit onto three remaining cores, HT comes in handy.
The chief problem with the engine is how no one at Crytek expected maps of this scale: None of the knock-on effects where tested. It's my experaance that one of the render threads really wants a core all to itself, and that's the main point to the set I gave you.
What's that got'a do with HT? Well, HyperThreading makes a 3.0GHz core look like two 1.6GHz cores to code, and since we are trying to make the 11 named, concrete threads fit onto three remaining cores, HT comes in handy.
#5
Posted 28 September 2015 - 04:54 AM
In theory, Goose is right. In practice, I can't tell the difference between HT on and off with my 5820K. If HT on has made any improvement over having it off, it is barely measurable on Haswell. As for Nehalem, that's another story. HT saw improvements apparently in Sandy Bridge, so it's possible you will see slightly different results from what I've seen.
#7
Posted 28 September 2015 - 02:28 PM
Goose, on 28 September 2015 - 10:58 AM, said:
How about the 8GHz P4 631: http://www.engadget....y-core-2-try-t/
Or try some of these: http://valid.canardpc.com/records.php
#8
Posted 28 September 2015 - 04:54 PM
Thanks for the updates and ideas. I remember playing a few games where there was a noticeable difference with HT on or off. Usually discovered it by turning it off in the bios and then monitoring my FPS/GPU/CPU usage. I can't remember which games it helped or hurt. Never thought of trying it with MWO. Was wondering if anyone else has tried this with a 1366 socket processor? Sorry I can't be more specific on which games HT helped or hurt.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users