Each of them has their own Technologies: Price vs Performance is the most important attribute to pick a graphic card...for example Nvidia has PhysX and they were the first...and do note that most game developers adopt Nvidia as they have additional physics effects...whilst Gamework Technology is present too...and G-Sync and Shadowplay not to mention... AMD also have pretty techs... AMD was more or less well recognised with bad driver support. Do note also Driver problems with Nvidia (such as Shadowplay) However there's a "brand war"... if you require any exclusive technologies[example: a FREESync Monitor but not G-Sync Compatible] , that may alter your decision. Nevertheless, both RX 480 and GTX 1060 provide value for money. If you are finding more on full HD, both cards will suffice. If looking for more grab the GTX1070 or GTX 1080.
Yeap. Preemption is not the same as asynchronous execution, it just hides the penalties from the lack of more advanced async.
I got all these results from Guru3D reviews hopefully they will give you a quick idea about hardware levels and where they stand in performance using a DX11 benchmark 3Dmark 2013 Fire Strike........STOCK.....OC REF Geforce GTX 1080...........19370 GBT GTX 980ti WATERFORCE..17916.....19178 MSI GTX 98Ti Lightning..........17853.....19011 MSI GTX 980Ti Gaming 6G.....17325.....18858 Asus GTX GTX 1070 STRIX.....16822 EVGA GTX 1070 SC Gaming...16680 Asus GTX 980ti Poseidon........16649.....18816 REF Geforce GTX1070 ...........16229 REF Geforce GTX 980Ti..........15656 Geforce GTX Titan X.............15602.....18374 Radeon R9 Fury X.................14374.....15226 Radeon R9 Fury STRIX...........13005.....14170 MSI GTX 980 Gaming............11831.....13280 MSI GTX 1060 6GT OC...........11729 MSI Radeon R9 390X GAMING..11686.....12610 MSI GTX1060 Gaming X 3GB..11368 ASUS GTX 980 Poseidon.........11872.....13722 REF Geforce GTX 1060...........11701 MSI RX 480 Gaming X.............11806 REF Geforce GTX 980.............11168 REF Radeon RX 480.................11042 MSI RX 470 8GB Gaming X......10521 Geforce Titan Black................10192.....11054 Palit GTX 970 Jetstream.........10044.....10901 Radeon RX470 (4GB).............9740 REF Geforce GTX 970..............9568
So because it favors a different way of implementing, one of that provides benefits to both AMD and Nvidia, it's not proper? That makes sense.
Yeah, for years now 3DMark hasn't been relevant to anything but 3DMark, for either AMD or Nvidia. So, in my opinion, you are correct: It should use the method that maxes the points for both Nvidia and AMD, regardless of if those points are meaningful in any way. That being said, Barry J's post is useless since 3DMark isn't supposed to reflect game performance. I don't know how many ppm of people buy an expensive video card just to get a high 3DMark score and then do nothing else with it.
I 50% disagree 3dmark is close to what you get in gaming but not exact it is a good place to start to get a rough idea of a GPU's potential performance to narrow down the GPUs you want to research more. Several games of different types and from several vendors is the only way to get a true idea but you have to start somewhere.
Well, timespy in particular isnt a very good benchmark, as its been proven thta it does not max the async workloads giving Nvidia an advantage. There are large gaps in the async work batches that give nvidias preemptive implementation time time to catch up.
I am not really interested in DX12 yet as the games are still really DX11 with DX12 added later which still requires lots of tweaking to get them working well with the new API the OS is still being tweaked. DX12 is no where near ready DX12 is in Beta when it is running well I will look at which hardware is best but at the moment DX11 still rules
1060 seems weak after all... http://www.guru3d.com/articles_page..._graphics_performance_benchmark_review,6.html Somewhat "higher texel fill rate, higher fps?" titan x 1 422 1080 2 257 fury X 3 268 1070 4 241 Fury 5 268 Nano 6 256 980ti 7 176 390x 8 184 480 9 182 390 10 160 470 11 154 1060 12 128 380x 13 137 780ti 14 205
Yeah, ive yet to read this article as well. Is Timespy not the best example of async to date? Something like a 12% increase for FuryX with it enabled. I don't think any other DX12 or Vulkan game beats that.
Has it? You'd need to do a statistical analysis. The material is all available in Guru3D's video card reviews since they all have a bunch of games and 3DMark scores. My personal gut feeling is that Nvidia's cards have always been beating their AMD market seqment equivalents in 3DMark by a good margin, but the same doesn't happen in the games. However, I'm far too lazy to go through a bunch of reviews for the sake of a single forum post.
Well, I'm part of the Fury owners club now However turns out I've had a major hardware fault, my cpu or mobo has died (infinitely rebooting at bios, not passing the post), occasionally it has managed to begin booting into windows but I've met BSODs, latest one being "Machine Code Exception" :'( Done the whole put the old card back in, test a different PSU, test the ram and the ram slots etc, nothing working so I'm at the point where it looks like an RMA job, I'm just trying to figure out if it is indeed the cpu at fault or the motherboard (I'm inclined to believe it's the CPU) as occasionally I can get into the bios however as one could imagine I'd need another cpu in order to test if it is the mobo at fault or not). It's just incredibly frustrating, I guess because of "aggitating" the hardware, it revealed the problem (I've had this bootlooping at bios a long while ago but it went away on it's own accord so didn't think much of it at the time). As for the async debate, can people leave that for the actual async debate thread, or at least tone it down, it's beginning to look like it's getting out of hand.
Hi there Did you tried to replace RAM or use spare RAM? But I think as you said,this can be CPU I would check motherboard and CPU socket and check temps in BIOS of CPU if they're not high or higher What CPU cooling are you running ? Something similar happen to me with H100 and my temp in BIOS has been unusually very high(103.5°C) and my rad has have air lock as one hose has been usually very hot etc and what I've done is just bashed with bare hand where are end tanks on radiator or where hoses goes and temps dropped to previous temps,but this does happen now more frequently which means my H100i v2 failing But congratulation there to yours Sapphire Fury there,hope you will get this sorted and good luck Thanks,Jura
