Too bad that these tests don't have the good old 280X in them to show how much amd has "improved" for a good lough. I bet that the frame rate would be almost identical.

Too bad that these tests don't have the good old 280X in them to show how much amd has "improved" for a good lough. I bet that the frame rate would be almost identical.

I asked the same question. I guess HH dont have Tahiti chip around to test.

Very nice review! Couple of things that stand out: It looks like that in games that actually use the full shader capacity of the card, all 2048 stream processors, like in your analysis of Tomb Raider @ 2560x1440, that the x version really justifies itself versus the plain 380. 57 fps for the 380 versus 75 fps for the 380x is definitely a frame-rate increase that would be directly perceivable by the player, imo. It's also interesting to note that most game engines seem not fully geared to use the total stream processor array in a given gpu--so most games are leaving a lot of performance on the table, maybe, regardless of the gpu being tested. Also, I am delighted to see that you are indeed including the AoS d3d12 bench in your testing suite...! It's not that I am so impressed with that particular game, but I think it is important for Maxwell owners to see that Maxwell is not the d3d12-compliant gpu that was originally represented by nVidia. (100% d3d12-compliant, was the original representation.) The idea that "we haven't turned it on yet" is simply silly and completely unbelievable, imo. I do wish you'd have taken it the logical step further and tested Maxwell with the feature turned on. None of this would ever have been an issue had not nVidia initially misrepresented the gpu. Keeping the IHVs honest is always a good thing, for both of them...;) And for us as consumers! AMD has its own crosses to bear. Had in the back of my mind to hand the 380 to the wife and pick up a 380x, but I'm still cogitating & collating about that...I may just hang on to the 380 for awhile.