Overclocked Video cards, CPU, and RAM

Message Topic Search Topic Options Post Reply Create New Topic Printable Version

   >UPDATE<

Over the next couple of days I will update this post with stock GPU's, stock CPU, slower ram speed, faster ram speed, and tighter/looser timings.

This is a very basic Overclocking series of tests, I have spent weeks of testing just one small attribute in the past, where this is a very broad test, to show what simple Overclocking can do.

I ran 2 scenarios, Avalanche, and Frost, 3 times each, took the average of the tests for each setup to come up with the F.P.S. All of the settings in Crysis warhead were the same as the screenshot bellow.

.
 This happens to be the screen shot of the HT off test. You can see the Crysis Warhead setting,  1920x1200, Enthusiast level, Direct x 10, 16AA

System sepcs.. Core i7 920 overclocked to 4.5Ghz
                         6 Gb RAM @ 1716 Mhz
                         Windows 7 OS
                         2 GTX 295 GPU's
                         Classifed MOBO
                         all water cooled

Here is the results.

==CONCLUSION==

1. Overclocking the CPU has the largest impact, nothing new here.

2. 1066 Mhz RAM chokes the i7, 1333 slows it down a bit, 1600 is decent, and for the last few frames 1866.   As a side note, I should have had more for-sight on running these tests, because I could have changed the Overclock parameters at the beginning, so I could have incorporated 2,000 Mhz RAM also.  My wifes RAM would not run 2150Mhz with the same timings, or even 1 tick loser timings, like the RAM in the Drunken Monkey machine.

3. Hyperthreading with Windows 7 is a must. With HT off, you can Overclock all you want, and a stock i7 with HT on will be more powerful. 

4. With Quad SLI, overclocking the GPU's has very little gain, (10 percent) unlike with a single card, or even SLI, where 20-30 percent gains are possible

5. I threw in a back to back CAS delay change, raising this timing alone, made a 5% change in the FPS, which may sound small, but when you think about the fact that just one timing changed the game performance that much, it really is significant.   Well to me. lol

Edited by justin.kerr - 04 Oct 2009 at 1:00pm

Author	Message Topic Search Topic Options Post Reply Create New Topic Printable Version
justin.kerr Members Profile Find Members Posts Add to Buddy List DS Veteran Joined: 06 May 2008 Online Status: Offline Posts: 5084	Quote Reply Topic: Overclocked Video cards, CPU, and RAM Posted: 02 Oct 2009 at 5:07pm
	>UPDATE< Over the next couple of days I will update this post with stock GPU's, stock CPU, slower ram speed, faster ram speed, and tighter/looser timings. This is a very basic Overclocking series of tests, I have spent weeks of testing just one small attribute in the past, where this is a very broad test, to show what simple Overclocking can do. I ran 2 scenarios, Avalanche, and Frost, 3 times each, took the average of the tests for each setup to come up with the F.P.S. All of the settings in Crysis warhead were the same as the screenshot bellow. . This happens to be the screen shot of the HT off test. You can see the Crysis Warhead setting, 1920x1200, Enthusiast level, Direct x 10, 16AA System sepcs.. Core i7 920 overclocked to 4.5Ghz 6 Gb RAM @ 1716 Mhz Windows 7 OS 2 GTX 295 GPU's Classifed MOBO all water cooled Here is the results. ==CONCLUSION== 1. Overclocking the CPU has the largest impact, nothing new here. 2. 1066 Mhz RAM chokes the i7, 1333 slows it down a bit, 1600 is decent, and for the last few frames 1866. As a side note, I should have had more for-sight on running these tests, because I could have changed the Overclock parameters at the beginning, so I could have incorporated 2,000 Mhz RAM also. My wifes RAM would not run 2150Mhz with the same timings, or even 1 tick loser timings, like the RAM in the Drunken Monkey machine. 3. Hyperthreading with Windows 7 is a must. With HT off, you can Overclock all you want, and a stock i7 with HT on will be more powerful. 4. With Quad SLI, overclocking the GPU's has very little gain, (10 percent) unlike with a single card, or even SLI, where 20-30 percent gains are possible 5. I threw in a back to back CAS delay change, raising this timing alone, made a 5% change in the FPS, which may sound small, but when you think about the fact that just one timing changed the game performance that much, it really is significant. Well to me. lol Edited by justin.kerr - 04 Oct 2009 at 1:00pm

Onkel_Ken Members Profile Find Members Posts Add to Buddy List DS Veteran Joined: 17 Dec 2008 Online Status: Offline Posts: 2961	Quote Reply Posted: 02 Oct 2009 at 7:45pm
	Good luck, we will be watching as you post your results!
	Velox X/AMD 5900X/MSI X570 Pro Carbon MB/32GB DDDR 3600 DSPS/850W PS/500GB Samsung 980 Pro/1TB Samsung 970 EVO Plus/500 GB HDD/Sound Blaster AE-7/EVGA FTW3 Ultra RTX 3080 10GB/Liquid CPU cooler AIO/

venom Members Profile Find Members Posts Add to Buddy List Newbie Joined: 08 Oct 2008 Online Status: Offline Posts: 0	Quote Reply Posted: 02 Oct 2009 at 11:05pm
	Very nice, looking forward to the future benchmark results. Edited by venom - 02 Oct 2009 at 11:06pm

DST4ME Members Profile Find Members Posts Add to Buddy List DS ELITE Joined: 14 Apr 2008 Online Status: Offline Posts: 36758	Quote Reply Posted: 03 Oct 2009 at 12:28am
	Nice, I'm expecting very little difference from ram and cpu speed at that res, as long as you keep the gpu settings as they are.
	Nvidia Drivers----AMD Drivers

justin.kerr Members Profile Find Members Posts Add to Buddy List DS Veteran Joined: 06 May 2008 Online Status: Offline Posts: 5084	Quote Reply Posted: 04 Oct 2009 at 1:11pm
	Testing done.

Overclocked Video cards, CPU, and RAM

About Us

Community

Support

Company

Contact Us

Follow Us