firstname.lastname@example.org for the results. which 3090 do you have? FTW3 air version similar to this card? If yes, it will make good apples to apples comparison!
LordredSo, I'm going through my testing gambet right now, and funny thing, 3DMark does not recognize the 3080 12gb yet! https://www.3dmark.com/spy/25661530 Oh well, I'll work through the whole thing then I'll be back with numbers, and graphs, and what have you
LordredI know it might be slightly inconsiderate to brag about getting a GPU at this time, but I was able to get my dirty little hands on a 3080 FTW3 12 GB. This card is not destined to go in my personal computer for I already have a 3090, but is actually a gift for a friend of mine going through some really tough times. (the kind of times where you can't really get out of the house and do anything due to health issues)Before I surprise him personally with the card I'm going to go ahead and toss it in my system and put it through the rigmarole to see how it performs.My plan is to put both cards through The same boring but repeatable synthetic tests that everyone knows and loves, monitor and graph out utilization performance temperature power consumption and all that fun stuff. From some of the reviews available, it looks like the 3080 12 GB should perform quite well even at 4k.Results to come tomorrow, for today I'm at work.
LordredThis is interesting, GPU-Z also fails to recognize the GPU....
fugly16LordredThis is interesting, GPU-Z also fails to recognize the GPU.... ReBAR is disabled. New drivers not enable this for this card?
fugly16 ReBAR is disabled. New drivers not enable this for this card?
CraptacularOne ReBAR is enabled in the BIOS of the motherboard. Either he didn't enable it, or GPUZ isn't reporting it. At any rate reBAR doesn't do a whole lot for Time Spy. As for his score, it looks fine to me. It's a bit better than a vanilla RTX 3080. This is my 3080 in a i9 10850k system for reference. His GPU score is a bit better as expected than what I see here. https://www.3dmark.com/spy/14483726
CraptacularOnefugly16 ReBAR is disabled. New drivers not enable this for this card?ReBAR is enabled in the BIOS of the motherboard. Either he didn't enable it, or GPUZ isn't reporting it. At any rate reBAR doesn't do a whole lot for Time Spy. As for his score, it looks fine to me. It's a bit better than a vanilla RTX 3080. This is my 3080 in a i9 10850k system for reference. His GPU score is a bit better as expected than what I see here. https://www.3dmark.com/spy/14483726
LordredAllright! Testing is complete, I've made a few graphs, and saved all the relevant data. All tests were performed in a 22°C room, using a Fractal Arc Midi tower, with 5 140mm fans, and 1 180mm fan, arranged in a diagonal flow setup. Four of the 140mm fans are induction, and the last 140 and 180 are extraction. System Specs: AMD R9 5950x on a ASRock B550 Extreme 4, filled with 4x16gb Crucial Balistix 3600 CL16. Cooled by a venerable Zalman CNPS-9900 Max (Yes it is adequate, trust me.) The system is not overclocked, but PBO is used and I spent quite some time dialing in voltage offsets to wind up with peak performance across all load situations. Starting off with Port Royal: Default operation of the 3080 and 3090, with fresh driver installs for both cards. The test was carried out with no changes to how the cards run out of the box. https://www.3dmark.com/compare/pr/1400003/pr/1400081 Here is the data plot for these cardsPeak power was measured at 431w for the 3090 FTW3 Ultra.Peak power was measured at 410w for the 3080 FTW3 Ultra 12gb.Average power was measured at 417w and 398w respectively.Peak temps were measured at 77°C for the 3090 and 74°C for the 3080. Repeating the test with both cards having modified the boost curve to end at 1800mhz and 918mv https://www.3dmark.com/co..e/pr/1400012/pr/1400113 Peak power was measured at 371w for the 3090 FTW3 Ultra.Peak power was measured at 332w for the 3080 FTW3 Ultra 12gb.Average power was measured at 344w and 319w respectively.Peak temps were measured at 70°C for the 3090 and 68°C for the 3080. (Sorry, framerate data was not captured on the 3080 for some reason, I did not double check my log before removing the card) Now for good ole Time Spy:https://www.3dmark.com/compare/spy/25663300/spy/25664200 Time Spy at 1800mhz 918mv:https://www.3dmark.com/compare/spy/25663875/spy/25664856 Time Spy Extreme:https://www.3dmark.com/co.../25663439/spy/25664356 Time Spy Extreme 1800mhz 918mvhttps://www.3dmark.com/compare/spy/25663763/spy/25664765 We can see that the 3080 12GB is basically, 9-10% slower then the 3090 in most conditions. It can close this gap to as close at 6% under 4k loads but runs wild with a 400w power target. We can also see when we reign in both cards to run at their rated speed, the 3080 is more consistently 10% slower, while only using 7% less power. Is it a powerful card? yes, is my friend going to be blown away, he damn well better be! Be well.
kraade3% for 400$ more , what a bargain
Lordredkraade3% for 400$ more , what a bargain I am not calling the gpu a bargain, it was more a matter of it was available to purchase so I did, I am fortunate enough to be in a position where I can help a friend out.