EVGA 790i FTW

by Rajinder Gill on November 5, 2008 9:00 AM EST

Overclocking results

The 790i FTW was truly a mixed bag for us. We did not manage to find any real consistency when running the board above stock settings. We tried three different BIOS revisions; each presented us with a series of quirks ranging from the infamous FSB holes to voltage irregularities. We also managed to lose two OS installations in the process of overclocking. If you do intend to push the board, do yourself a favor and either make a complete drive image or stick to using a "medieval" IDE hard drive.

We also had to resort to using 1GB memory modules for most of our testing. Our 2x2GB Corsair XMS3 1600MHz kit was limited by the FSB holes present in the current BIOS releases. This led to us staying under 400FSB or shooting over 460FSB to get the board to boot. The Corsair kit can run at 1800 stable on other boards and probably on capable 790i boards too. In this instance, the FTW's FSB boot up requirements pushed the modules was outside their capabilities, so we never got to run any form of stability testing near 1800MHz.

On a more positive note, we did manage to scrape out a run of Crysis with an E8500 at 545FSB at 1280X1024 resolution, but the board would not pass a test run at 1920X1200.


We tried going over to 550FSB with the memory running at 1800MHz, but once again lost our OS installation. 545FSB was more or less our benchmarking limit with water-cooling and we managed to scrape out a run of 3DMark05 without any problems.


Where's the Subzero Stuff?

Unfortunately, our board would not cooperate when we strapped a cascade to the CPU. The problem was that the board refused to boot any time we moved away from stock settings. We had to clear the CMOS by removing the battery and shorting the clear jumpers near the BIOS chip. The board would then boot back up at stock defaults, even with the cascade idling at -118C (so it's not a cold board bug).

The 790i FTW is still on the test bench and we are prepared to give it another spin if EVGA can send us an improved BIOS. It would have been interesting to run some single and dual GTX 280 cards with this board and our QX9650, but we found that getting things to work with subzero cooling proved to be impossible. This review was delayed while we tried in vain to get to the crux of what was wrong. In the end, we made a decision to go with what we have, so people who are thinking of purchasing the board know what they're getting into if they take the plunge right now.

Gaming Performance Conclusion
Comments Locked

15 Comments

View All Comments

  • cesthree - Wednesday, November 5, 2008 - link

    EDIT:

    "No FSB holes, no data corruption at STOCK SPEEDS, no unexplainable SLI anomalies like graphic corruption or freeze-ups in games or while watching videos."

    Because I don't have to deal with the underdeveloped SLI crap anymore!!

    "EVGA is brave. They'll probably have to install new servers and hire another 100 techs to answer the 1000 pages of threads with the title "FREEZEING UP AT STOCK SPEEDS, HELPXOR M3!!!" or "WILL WE EVER GET AN ANSWER FROM YOU EVGA?""

    That is, if they can get their techs out of the break room!
  • Bozo Galora - Wednesday, November 5, 2008 - link

    Where the hell is info on PCP 1200W PSU avail?
    Cant find it with google
  • abzillah - Wednesday, November 5, 2008 - link

    Why aren't the graphs the same color? Why is the EVGA green in one bar graph and orange in another? How hard is it to keep the bar graph colors consistent?
  • McFlu - Thursday, January 1, 2009 - link

    they do stay the same color, theyre just in different orders along the side. you should read a bit better.
  • Rajinder Gill - Wednesday, November 5, 2008 - link

    http://www.pcpower.com/power-supply/turbo-cool-120...">http://www.pcpower.com/power-supply/turbo-cool-120...

Log in

Don't have an account? Sign up now