My bro and I have been debating for awhile now, the debate is whether overclocking in crossfire nets the same performance increase as overclocking with a single card. He believes that you only benefit from overclocking both cards if you are running SLI. I really want to prove him wrong but I have no AMD cards in crossfire and there are no review sites that test this. So I'm wondering if you guys could help me out. What I need is this.
Single card:
Stock clock
Bench results
Max OC
Bench Results
Crossfire-
Stock clocks
Bench Results
Max OC
Bench results
*edit* seems I somehow completely misread the op. If you only overclock one card in a CrossfireX set up and leave the other card at stock then both cards will effectively only run at the lower clock speed.
If I remember correctly cards in SLI will always run with identical clock speeds. So if one card is overclocked and the other is not the the faster card will be down clocked to mach the slower. If you over clock a SLI setup you need to overclock both cards to the same speed.
Well that is not entirely accurate as CF can run wit a faster and slower card and not limit the speed of the faster card. How well this works in the real world I don't know.
I know that in crossfire/SLI that the fastest card slows down to the slowest card. What I'm uncertain of is the scaling of overclocked performance in crossfire vs SLI. My bro thinks that overclocking performance in crossfire is poor versus overclocked performance in SLI. I want definitive proof to show him that he is wrong.
IE
5850 at 900mhz gets a 25% performance boost
2x 5850's in CF at 900mhz still gets that 25% performance boost versus say a GTX 460 in SLI once overclocked.
even if there was a difference, i doubt it would be tangible. I think the results of a difference if any would be considered in the margin of error when running benches
Alright, sweet, it'd be awesome if you were to do it, since I'm currently looking for someone on the Nvidia board to do OC'ing benchies with 2x GTX 460's.
While I was overclocking my crossfire setup using heaven benchmark, I observed that every 100mhz increased net me about 10 - 15 FPS. Do not know if that will be any help.
i think what OP is getting at is the debate = hypothetically OC gpu(+15% performance) + OC GPU(+15% performance) does or does not equal +30% performance in sli/cfx (correct me if im not understand your question). But as we all know, sli/cfx does not double your performance, so in a sense your brother is right. if you take the above example, you wouldnt get 30%, maybe 20% additional from OC.
I like the question raised here. I've been wondering myself, in fact, as running quad-crossfire provides so much potential horsepower @ stock clocks..if it makes sense to further push the system considering potential heat & strain on components for what may be utterly negligible benefit. I'ma run a few tests myself here (dunno why I haven't ever before?) to see just how many "fps" are boosted in high clock relative to stock & underclocked values.
Thanks man, I use to have a 5850 myself, they were awesome cards. Hopefully he'd be satisfied with these benches. :]
Edit: Just talked with my bro, he's partially convinced but not completely, he doesn't believe the cards are OC'd enough to provide definitive data. Plus the memory is also untouched on those 5850's so that could effect part of the equation.
Well let him think what he wants its been proven that he's wrong right here. Overclocking higher isn't going to change anything and overclocking the memory yields very little extra performance so isn't really worth it and I would love for him to explain how he thinks it will have a negative impact on the performance which it would need to do if he was to be proven correct. To put it politely he's clutching at straws.
I like the question raised here. I've been wondering myself, in fact, as running quad-crossfire provides so much potential horsepower @ stock clocks..if it makes sense to further push the system considering potential heat & strain on components for what may be utterly negligible benefit. I'ma run a few tests myself here (dunno why I haven't ever before?) to see just how many "fps" are boosted in high clock relative to stock & underclocked values.
With your 5770's, How much eye candy settings were applied in
the tests.
Later
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Related Threads
?
?
?
?
?
Ask a question
Ask a question
Overclock.net
27.8M posts
541.2K members
Since 2004
A forum community dedicated to overclocking enthusiasts and testing the limits of computing. Come join the discussion about computing, builds, collections, displays, models, styles, scales, specifications, reviews, accessories, classifieds, and more!