Geno Posted June 25, 2016 Share Posted June 25, 2016 (edited) I'm having a hard time understanding whats going on between my Gigabyte G1 980Ti (Hynix) and a friends EVGA 980Ti Classified (Samsung) when running Firestrike 1.1. Both in the same PC, same drivers his graphics score is well beyond mine regardless how much I overclock the G1. Anomaly 1: He picked up 1000 graphics score going from Bios 1 to LN2 bios > Most of the difference was made up in Test 2. Anomaly 2: His card is pulling a far larger score at 1543/2078 then my card at 1595/2178. Is the Classified LN2 bios overclocking the sub-timings of his Vram to completely trump my card in test 2? In all my years overclocking/benchmarking ive never seen such confusing results when run on the same system. He picked up nearly 10 fps on graphics test 2 alone with, from what I saw, was a flick of a bios switch. On the G1 I'm experiencing no throttling what so ever. I'm also getting positive scaling the more I overclock. What on earth is going on? We've run firestrike over the cards over 30 times with close with no more then +/- 0.2 fps variance which suggest something is definately skewing the results dramatically in his favor. Edited June 25, 2016 by Geno Quote Link to comment Share on other sites More sharing options...
Lays Posted June 25, 2016 Share Posted June 25, 2016 Same CPU and rest of system as well? Quote Link to comment Share on other sites More sharing options...
Geno Posted June 25, 2016 Author Share Posted June 25, 2016 Same CPU and rest of system as well? Yep, exactly the same system, operating system, drivers and overclocking software. Quote Link to comment Share on other sites More sharing options...
Administrators websmile Posted June 25, 2016 Administrators Share Posted June 25, 2016 Apart from the fact that Samsung ICs can take much tighter timings at same volts than Hynix, which is most likely one reason for your results, LN2 BIOS might have higher power target so the card runs frequency continuously, on lower power target it might drop a few mhz during longer load. Did you monitor the frequencies during the run to check this? Quote Link to comment Share on other sites More sharing options...
Geno Posted June 25, 2016 Author Share Posted June 25, 2016 (edited) At 130% power target my G1 never got close, there was no visible throttling over any monitoring software. All clocks were held during the firestrike tests - bar physics. Even if samsung IC's can take tighter timings, will that attribute 10 fps over similar clocks on test 2? Does anyone know if LN2 bios on the classified does in fact tighten timings? Edited June 25, 2016 by Geno Quote Link to comment Share on other sites More sharing options...
crustytheclown Posted June 25, 2016 Share Posted June 25, 2016 Maybe this can help you.Evga bios is more efficient. KINGPIN 980Ti // Titan-X // NV 980Ti Performance Comparison - kingpincooling.com Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.