![]() |
Benchmark run-times Hi all, I have a few questions regarding OC'ing and benchmarks: 1. How long should I run Orthos, and have Core-temp logging my temps, to get a good average load temperature? I found after 20 minutes my temps maxed out 2. For stability testing, is Prime 95 or Orthos a better program? It appears that they run similar tests 2b. Is there a helpfile I can look at so I can setup two instances of Prime 95? I have thus far been unsuccessful at finding any information on this) 3. Are there premade game benchmark programs that will run say, Prey or Doom 3, in different areas of the game and log the FPS? Or do I have to play the game myself in various areas and log my results with fraps? 4. Thanks! I am trying to get an idea of what kind of procedure I should follow so I can make my hardware reviews as rock solid as possible. :biggrin: |
Quote:
Quote:
Quote:
Quote:
|
1. about 30 minutes, full load. 2a. they are the same. 2b. install P95 in 2 seperate folders. run both .exe. Or use Prime95.2 (Download Prime95 25.2 - Prime95 is a program which searches for very large prime numbers. It has also gained a reputation for getting errors on computers which are even slightly unstable. If there is any problem with your CPU, your RAM, or the parts ) which can run upto 4 instances at once. 3. I have no idea. 4. My pleasure. John edit: :censored:, ssilson beat me to it.:haha: |
1. Once your temps level out, that's pretty much it. I let it go for 30 minutes, just to make sure, but 15-20 usually does just fine. 2. If there's a difference between Prime and Orthos, you'd have to get pretty technical to understand it. 2b. Orthos will automatically grab 2 cores with each instance. To get Prime running on two cores, open two instances of the program. Before running each one, go to the Windows task manager, and under the list of processes running, find the two instances of Prime. Right click on each one to get a menu, and go to the 'Affinity' option. Set the affinity of one of the processes to core 0, and the other to core 1. Then run both instances like normal. 3. If you're looking for fps counts to compare with other machines, you need a reproduceable scenario. Can't speak for Doom 3 or Prey, but I know Half Life 2: Lost Coast and Lost Planet both have in-game benchmarks. Edit: Beaten twice! |
Quote:
ATi Steam Offer |
Quote:
3 = Benchmarks is basically running a game and watching the FPS with different settings, tho some games have time_demos to do this for you |
Ah thats what I was looking for, time demos. I was wondering how Tom's Hardware got all their benchmarks for Prey, doom 3, HL:Lost Coast etc to be consistent between components and playing, since you can't always look in the same direction all the time for two passes through a game... This is great information. Much appreciated. If anyone thinks it may be worth it, I am considering posting my experiences OC'ing my system. I am unsure if that kind of info is valuable or not...cheers. |
Please do come back with the results. It's always interesting, and you never know: we may point you in a direction you didn't think about and reach a higher oc. Which cooler do you have for the E2180? I still have an E2140 that did 3.0Ghz 24/7 , under air - with an AC7Pro. John |
I am going to run some benchmarks on my stock setup, and then throw on a Thermalright Ultra 120 extreme for good measure. I am hoping for 3.3ghz stable, with an optimistic goal of 3.6ghz. I also will be OC'ing my memory up from 800mhz, and I am planning on pushing my bfg gpu a little more than the factory OC as well. Hopefully I don't break something! |
| All times are GMT -7. The time now is 03:52 PM. |