Ieldra
I Promise to RTFM
- Joined
- Mar 28, 2016
- Messages
- 3,539
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
Most important thing I say was Steam Controller support! Sweet!
Nice! Massive improvement on my 7970 with Vulkan.
Waiting for revision 2.0 to buy
Won't happen sales are awful for the steam controller and steam link, and frankly this is valve so if they do make a 2.0 expect sometime next decade.Waiting for revision 2.0 to buy
Bummer :-(
Won't happen sales are awful for the steam controller and steam link, and frankly this is valve so if they do make a 2.0 expect sometime next decade.
Solid improvement for people with amd gpus. Nvidia? Go pound sand, no improvement (because pascal still can't do concurrent graphics + compute in the same way as amd).
Nope, 1070 is 30% faster than the 480 in their Vulkan bench.More like because nvidia has the best OpenGL drivers in the industry.
Nope, 1070 is 30% faster than the 480 in their Vulkan bench.
Looking at TPU and PCGamer's 1070 averages, it should be about 50% faster.
Vulkan allows the 480 to close performance gap with Nvidia by approx 20%.
I pulled the averages from here:Should a 1070 really be 50% faster considering it likely has the same die size / transistors as the RX 480 (when considering that 25% of the 1070 chip is disabled)? Or does it have a 50% advantage because AMD's non-DX12 / Vulkan drivers are awful?
More like because nvidia has the best OpenGL drivers in the industry.
Should a 1070 really be 50% faster considering it likely has the same die size / transistors as the RX 480 (when considering that 25% of the 1070 chip is disabled)? Or does it have a 50% advantage because AMD's non-DX12 / Vulkan drivers are awful?
Yes. And now, that "advantage" will boost them less and less. More and more shackles on amd performance are being removed by the day. More nvidia engineers to throw at optimizing the cesspool known as opengl 4.5, great.
People keep saying this, but how well does it track?
How much space on a gpu die is devoted to resources that do not scale with chip size? Is there literally zero difference performance wise from an nvidia chip designed for a 300mm die size on the same architecture that was cut down by say, 25% for an "effective" die size of 225mm compared to a full nvidia chip designed from the ground up to fill out a 225mm die size?
There is zero difference between a cut down gpu vs a full chip gpu? Or are there some extra advantages to being cut down to a similar effective die size? I don't know, asking, because this keeps being parroted as if one ought to expect perfectly linear scaling between cut down chips and full chips of equal size and presumed equivalent hardware resources.
Performance is roughly proportional to the die-size raised to the 1/2 power.People keep saying this, but how well does it track?
How much space on a gpu die is devoted to resources that do not scale with chip size? Is there literally zero difference performance wise from an nvidia chip designed for a 300mm die size on the same architecture that was cut down by say, 25% for an "effective" die size of 225mm compared to a full nvidia chip designed from the ground up to fill out a 225mm die size?
There is zero difference between a cut down gpu vs a full chip gpu? Or are there some extra advantages to being cut down to a similar effective die size? I don't know, asking, because this keeps being parroted as if one ought to expect perfectly linear scaling between cut down chips and full chips of equal size and presumed equivalent hardware resources.
youtube.com/watch?v=cjgFNkNW8zI … DOOM GL vs Vulkan on an awesome AMD 480. Heads up benchmarkers, use TSSAA or no AA (else Async Compute is disabled)
Your loss. It's a fun game. I just-about-maxed it out (no Nightmare textures or AA) on a Fury at 4K so one 980Ti should be fine.I ain't buying it then considering I have a 3x 980Ti setup for 4k.
Sorry but try as you might you can't deny what Nvidia lacks. Now whether it impacts them greatly or not is up for discussion.My GTX 1080 hit 200fps without Vulcan but dropped frequently to 120 ish with Vulcan it consistently stays higher with some frame drops. Considering the open gl implementation on AMD cards for Doom was already broken we can not assume the performance gain is indicative of a nvidia weakness, but instead should just be happy AMD is improving their pipeline.
Most important thing I say was Steam Controller support! Sweet!
They say they support async compute their has been no conclusive evidence to the contrary. AMD sees such large gains because in this game the version of opengl they used on AMD cards was sub-par, who's fault that is I do not know, but we literally do not have enough evidence anywhere to suggest nvidia suffers from low level api's.Sorry but try as you might you can't deny what Nvidia lacks. Now whether it impacts them greatly or not is up for discussion.
Saying one supports a feature does not mean they support it at a hardware level. They mean, in the case of Nvidias current lineup, that they emulate the feature via software. There is plenty of proof all over the internet that proves Nvidia lacks true hardware support of asynchronous compute+graphics.They say they support async compute their has been no conclusive evidence to the contrary. AMD sees such large gains because in this game the version of opengl they used on AMD cards was sub-par, who's fault that is I do not know, but we literally do not have enough evidence anywhere to suggest nvidia suffers from low level api's.