[computerbase.de] DOOM + Vulkan Benchmarked. - Page 26 - Overclock.net

Forum Jump: 
Reply
 
Thread Tools
post #251 of 632 Old 07-16-2016, 11:05 AM
Graphics Card Aficionado
 
Mahigan's Avatar
 
Join Date: Aug 2015
Location: Ottawa, Canada
Posts: 1,745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 874
Quote:
Originally Posted by Glottis View Post

I hope you are joking. Pretty much every time new AMD drivers is released all big sites re-run benchmarks and publish gains. Nvidia hardly ever get same treatment. I don't know why do I even bother replying to you. When I posted some benchmark which forgot to enabled TSSAA which enables Async for AMD you gone mental instantly screaming fault. But when Nvidia doesn't even get to use Async at all it's all good in your eyes. There is no reasoning with someone like you.

I am being sarcastic. Dark humor.

As for me going mental... do you have a quote which indicates me going mental? Or is it more like my pointing out that the Async was not enabled in a test which supported Async and was meant to showcase that ability? nVIDIA cannot do Async in this particular test therefore we have to go with what is available just as we did with the initial Doom results. At first AMD was in the lead during the Alpha stages... then nVIDIA as they gained better OpenGL optimizations while AMD concentrated on Vulkan optimizations. Now that AMDs results are out... we go with those and if nVIDIA and Bethesda release the nVIDIA Vulkan optimizations we will have to also consider those.

I am not the one who has issues reasoning bud. All I do is reason and post reasoned explanations and tech observations. I have to deal with people like you... I am not talking about nVIDIA fans but just FANS. I am either insulted by AMD fans (over at Anandtech from my Pascal coverage for daring to suggest that Pascal could do Asynchronous Compute but not Asynchronous Compute + Graphics) or I have to deal with NV fans who claim everything I say is BS when for the most part it is true.

I was right about the MAxwell caching issues (now fixed with Pascal) and I was right about Maxwell lacking Async support as well. At some point in time I hope that I will not have to deal with such hostilities when commenting on the inner happenings/behind the scenes happenings of GPU architectures.

I have been pissed off at NV at times and recently AMD annoyed me with their 150W TDP claim for Polaris (110W for the GPU is not a defense). That being said... I am still waiting for the Maxwell Async drivers folks shoved in my face the last time around. I still have not gotten over that one. That was a mighty big lie on nVIDIAs part.
bluezone likes this.

"Once you eliminate the impossible, whatever remains, no matter how improbable, must be the truth." - Arthur Conan Doyle (Sherlock Holmes)
Mahigan is offline  
Sponsored Links
Advertisement
 
post #252 of 632 Old 07-16-2016, 11:05 AM
 
Join Date: Oct 2014
Location: UK
Posts: 304
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 19
Quote:
Originally Posted by ChevChelios View Post

just shows how bad OpenGL on AMD was in Doom

AMDs drivers in openGL have been laughably bad for as long as they have been making GPUs. The sad part is they never bothered to improve them for their customers sake.

BrightCandle is offline  
post #253 of 632 Old 07-16-2016, 11:08 AM
Graphics Card Aficionado
 
Mahigan's Avatar
 
Join Date: Aug 2015
Location: Ottawa, Canada
Posts: 1,745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 874
Quote:
Originally Posted by BrightCandle View Post

AMDs drivers in openGL have been laughably bad for as long as they have been making GPUs. The sad part is they never bothered to improve them for their customers sake.

100% true. They likely did not see the value in updating those drivers seeing as few games utilize the OpenGL API. In a way.. AMD were also late to the game with Vulkan drivers as they concentrated on DX12. Now AMD seems to have pretty robust DX and Vulkan support so those past issues from AMD are likely just that... the past.

"Once you eliminate the impossible, whatever remains, no matter how improbable, must be the truth." - Arthur Conan Doyle (Sherlock Holmes)
Mahigan is offline  
Sponsored Links
Advertisement
 
post #254 of 632 Old 07-16-2016, 11:30 AM
 
Glottis's Avatar
 
Join Date: Feb 2012
Posts: 1,389
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 88
Quote:
Originally Posted by Mahigan View Post

I am being sarcastic. Dark humor.

As for me going mental... do you have a quote which indicates me going mental? Or is it more like my pointing out that the Async was not enabled in a test which supported Async and was meant to showcase that ability? nVIDIA cannot do Async in this particular test therefore we have to go with what is available just as we did with the initial Doom results. At first AMD was in the lead during the Alpha stages... then nVIDIA as they gained better OpenGL optimizations while AMD concentrated on Vulkan optimizations. Now that AMDs results are out... we go with those and if nVIDIA and Bethesda release the nVIDIA Vulkan optimizations we will have to also consider those.

I am not the one who has issues reasoning bud. All I do is reason and post reasoned explanations and tech observations. I have to deal with people like you... I am not talking about nVIDIA fans but just FANS. I am either insulted by AMD fans (over at Anandtech from my Pascal coverage for daring to suggest that Pascal could do Asynchronous Compute but not Asynchronous Compute + Graphics) or I have to deal with NV fans who claim everything I say is BS when for the most part it is true.

I was right about the MAxwell caching issues (now fixed with Pascal) and I was right about Maxwell lacking Async support as well. At some point in time I hope that I will not have to deal with such hostilities when commenting on the inner happenings/behind the scenes happenings of GPU architectures.

I have been pissed off at NV at times and recently AMD annoyed me with their 150W TDP claim for Polaris (110W for the GPU is not a defense). That being said... I am still waiting for the Maxwell Async drivers folks shoved in my face the last time around. I still have not gotten over that one. That was a mighty big lie on nVIDIAs part.
didn't mean to offend you, merely used word mental to say that you are passionate in your posts. i agree you were right on a lot of things, but there's a lot inconsistency in DOOM Vulcan benchmarks and there's no denying that. i don't know why that is, but for example TimeSpy benchmark is very consistent across all websites.
Glottis is offline  
post #255 of 632 Old 07-16-2016, 11:39 AM
PC Gamer
 
FLCLimax's Avatar
 
Join Date: Mar 2011
Location: New York
Posts: 3,914
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 271
Couldn't use video capture program...didn't work. crappy hand held camera time, sorry. So nice on the Fury. It does dip into the 50's when doing melee finishers and when the room is full of enemies and explosions, but i believe this is higher than all nvidia cards @ 4K or equal to the 1080?.

Warning: Spoiler! (Click to show)

Warning: Spoiler! (Click to show)
FLCLimax is online now  
post #256 of 632 Old 07-16-2016, 11:43 AM
 
airfathaaaaa's Avatar
 
Join Date: Aug 2015
Posts: 825
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 62
Quote:
Originally Posted by Mahigan View Post

What we do know about Doom Vulkan at 1080p is that a
FuryX can range between around 154 - 161 FPS varying on the scene being rendered.
GTX 980 Ti can range between around 135 - 148 FPS varying on the scene being rendered.
GTX 1070 can range between around 136 - 143 FPS varying on the scene being rendered.


What we do know about Doom Vulkan at 1440p is that a
FuryX can range between around 97 - 111 FPS varying on the scene being rendered.
GTX 980 Ti can range between around 85 - 101 FPS varying on the scene being rendered.
GTX 1070 can range between around 88 - 98 FPS varying on the scene being rendered.

So yeah... the FuryX beats out the GTX 980 Ti and GTX 1070. If Bethesda does include nVIDIAs support for concurrent execution of tasks (Asynchronous Compute but not Asynchronous Compute + Graphics) then we may see the GTX 1070 gaining ground on the FuryX.
someone took your posts and made a reddit post for it
https://www.reddit.com/r/Amd/comments/4t5ckj/apparently_3dmark_doesnt_really_use_any/
airfathaaaaa is offline  
post #257 of 632 Old 07-16-2016, 11:43 AM
 
kfxsti's Avatar
 
Join Date: Jun 2009
Location: NC
Posts: 701
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 25
Woohoo the high end cards run it above 60fps. Sweet lol


Originally Posted by deafboy
Go on, tell me why I have no idea what I am talking about. This should be interesting.
kfxsti is offline  
post #258 of 632 Old 07-16-2016, 12:12 PM
 
DaaQ's Avatar
 
Join Date: Mar 2013
Location: South Central Kentucky
Posts: 1,031
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 125
Quote:
Originally Posted by Glottis View Post

Async isn't yet working in Vulcan DOOM for Geforce graphics cards, and Pascal does have gains from Async. That's why I was more interested in 980Ti vs Fury X. Too bad all tech sites seem to ignore or aren't aware of this fact. We need to wait for Async to be enabled for Geforce before we can properly compare Fury X to Pascal (1070/1080).

Oh the irony.
The waiting game ensues.

Quote:
Originally Posted by Tsumi go_quote.gif

You're just nitpicking words now. I was talking about in terms of performance, those are the only measurements that matter. Not IPC or clock speed.
Quote:
Originally Posted by Tsumi go_quote.gif

If Bulldozer came out of the box at 5 ghz and overclocked to 6 ghz on air, no one would have said it was a failure.
DaaQ is online now  
post #259 of 632 Old 07-16-2016, 12:19 PM
PC Gamer
 
FLCLimax's Avatar
 
Join Date: Mar 2011
Location: New York
Posts: 3,914
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 271
Haha, yea. And what's worse is that NVIDIA can't just put some software out and say "Hey, come get your Async Compute for GeForce!". They have to make a driver on their end(and the devs have to make special optimizations) in each and every individual game that uses the feature.

NVIDIA, soon*™.
FLCLimax is online now  
post #260 of 632 Old 07-16-2016, 12:39 PM
 
Pereb's Avatar
 
Join Date: Oct 2015
Posts: 478
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 18
Quote:
Originally Posted by Mahigan View Post

We do not see Maxwell losing performance under 3D Mark Time Spy. We see a tiny performance boost. Thus 3D Mark Time Spy is not running Asynchronous Compute + graphics

This seems wrong to me... async is supposed to be disabled for Maxwell on the driver level. It should be normal for the performance to be the same with async on and off since it's never actually enabled. The 0,1% difference is just margin of error.
Pereb is offline  
Reply

Quick Reply

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off