Read more.The latest release of DirectX looks like it will provide AMD a big boost in new games and apps.
Read more.The latest release of DirectX looks like it will provide AMD a big boost in new games and apps.
Wow...big step for AMD!
Does that mean that Nvidia could in theory optimise their GPUs to give 40-50% more performance?
It's hardly surprising that nVIDIA are not happy with the benchmark, as it shows *much* bugger gains for AMD hardware. On the other hand, this is just one isolated benchmark, so conclusions avout DX12 performance should probably be avoided at this point.
I've heard that AMDs performance could be due to the fact that AMD has tended to brute force their way to high performance on the hardware side, while the software side wasn't as strong, whereas nVIDIA have a pretty good software side, so have been a little more efficient on the hardware side... In which case nVIDIAs DX12 gains might not be that great, as software will matter less due to the API being more efficient than before.
Last edited by gingerninja7; 18-08-2015 at 01:27 PM. Reason: Missed the quote... D'oh!
DX12 drivers are EXTREMELY new. You cannot draw comparisons yet (unless you want to get carried away with the very first DX12 benchmarks based on a games Alpha release on non-mature drivers).
Although there goes the conspiracy theories that Hexus will only post news articles that show nVidia in the best light
Main PC: Asus Rampage IV Extreme / 3960X@4.5GHz / Antec H1200 Pro / 32GB DDR3-1866 Quad Channel / Sapphire Fury X / Areca 1680 / 850W EVGA SuperNOVA Gold 2 / Corsair 600T / 2x Dell 3007 / 4 x 250GB SSD + 2 x 80GB SSD / 4 x 1TB HDD (RAID 10) / Windows 10 Pro, Yosemite & Ubuntu
HTPC: AsRock Z77 Pro 4 / 3770K@4.2GHz / 24GB / GTX 1080 / SST-LC20 / Antec TP-550 / Hisense 65k5510 4K TV / HTC Vive / 2 x 240GB SSD + 12TB HDD Space / Race Seat / Logitech G29 / Win 10 Pro
HTPC2: Asus AM1I-A / 5150 / 4GB / Corsair Force 3 240GB / Silverstone SST-ML05B + ST30SF / Samsung UE60H6200 TV / Windows 10 Pro
Spare/Loaner: Gigabyte EX58-UD5 / i950 / 12GB / HD7870 / Corsair 300R / Silverpower 700W modular
NAS 1: HP N40L / 12GB ECC RAM / 2 x 3TB Arrays || NAS 2: Dell PowerEdge T110 II / 24GB ECC RAM / 2 x 3TB Hybrid arrays || Network:Buffalo WZR-1166DHP w/DD-WRT + HP ProCurve 1800-24G
Laptop: Dell Precision 5510 Printer: HP CP1515n || Phone: Huawei P30 || Other: Samsung Galaxy Tab 4 Pro 10.1 CM14 / Playstation 4 + G29 + 2TB Hybrid drive
Millennium (20-08-2015)
Main PC: Asus Rampage IV Extreme / 3960X@4.5GHz / Antec H1200 Pro / 32GB DDR3-1866 Quad Channel / Sapphire Fury X / Areca 1680 / 850W EVGA SuperNOVA Gold 2 / Corsair 600T / 2x Dell 3007 / 4 x 250GB SSD + 2 x 80GB SSD / 4 x 1TB HDD (RAID 10) / Windows 10 Pro, Yosemite & Ubuntu
HTPC: AsRock Z77 Pro 4 / 3770K@4.2GHz / 24GB / GTX 1080 / SST-LC20 / Antec TP-550 / Hisense 65k5510 4K TV / HTC Vive / 2 x 240GB SSD + 12TB HDD Space / Race Seat / Logitech G29 / Win 10 Pro
HTPC2: Asus AM1I-A / 5150 / 4GB / Corsair Force 3 240GB / Silverstone SST-ML05B + ST30SF / Samsung UE60H6200 TV / Windows 10 Pro
Spare/Loaner: Gigabyte EX58-UD5 / i950 / 12GB / HD7870 / Corsair 300R / Silverpower 700W modular
NAS 1: HP N40L / 12GB ECC RAM / 2 x 3TB Arrays || NAS 2: Dell PowerEdge T110 II / 24GB ECC RAM / 2 x 3TB Hybrid arrays || Network:Buffalo WZR-1166DHP w/DD-WRT + HP ProCurve 1800-24G
Laptop: Dell Precision 5510 Printer: HP CP1515n || Phone: Huawei P30 || Other: Samsung Galaxy Tab 4 Pro 10.1 CM14 / Playstation 4 + G29 + 2TB Hybrid drive
Very interesting
The more one seeks, the more one finds and so you realise there is a lot more to be found.
I don't think so. I think what these benchmarks actually show is that AMD's implementation under DX11 is very, very bad. Whereas nvidia's is quite good. So you are seeing a huge imcrease for AMD going from DX11 to DX12 because they are going from bad to good. Whereas nvidia are going from good to good so the increase is correspondingly smaller.
Put another way, it's AMD who could unlock that extra performance in DX11 if they had better drivers for it, which of course they won't bother now DX12 is on its way.
What worries me more is that I had expected the change to CPU usage (e.g. draw calls) with DX12 to really help system performance with AMD CPU's but looking at the PCPer's article results it helps but doesn't really bring them any closer to Intel CPU performance, the i3 was beating the FX8 series chip in DX12 on the same GPU!
I'm sad for AMD now.
THE ABOVE IS A THEORY: We need tests based on Total war Rome 2 & witcher 3 or other dense titles. I will not accuse 390x for being way slower than 980 coz it preforms nearly same in the current approved 4k benchmarks
Can imagine nvidia aren't too pleased! Looks like all their hard work optimising dx11 has been made redundant. Oh well, better for competition!
I think what those results actually tell us is that the game engine is very CPU intensive. The i3 and both AMD processors struggle, and per-thread performance is critical (hence the i3 outpacing the AMDs). The results with the i7s show the unconstrained performance of the GPUs, and the big performance leap for AMD confirms that DX12 is reducing the CPU graphical overhead.
The fact that the results are so flat for all the other processors demonstrates that the CPU is beibng hit hard by the game engine: again AMD pick up a decent boost as their CPU graphical overhead goes down with DX12, but the game must really hammer the CPU for physics, AI etc. to have such a clear bottleneck on an FX 8370. I also wonder whether the subsystem has a role to play - interesting that it's so much faster on the two DDR4-based platforms.... more analysis required, methinks
A lot may also be due to a bug in a particular driver version of DX11..
I wouldn't call it very very bad. It would mean the driver is doing less to go around dx11 limitations when it comes to AMD. nvidia took the route of hacking/optimizing their drivers to make the most of dx11, AMD invested in mantle and worked closely with MS during that. When you get a game that abuses the CPU you get driver differences being exaggerated. We'll have to see how things turn out in games that do not have such a dependence on the CPU.
There are currently 1 users browsing this thread. (0 members and 1 guests)