[WCCF] HITMAN To Feature Best Implementation Of DX12 Async Compute Yet, Says AMD - Page 39 - Overclock.net - An Overclocking Community
Forum Jump: 

[WCCF] HITMAN To Feature Best Implementation Of DX12 Async Compute Yet, Says AMD

Reply
 
Thread Tools
post #381 of 799 (permalink) Old 02-15-2016, 08:00 PM
 
Join Date: Oct 2012
Location: Florida
Posts: 1,490
Quote:
Originally Posted by spyshagg View Post

Hi

"AMD is in a fragile position without there being a good reason for it. Their graphics cards performance should warrant a much stronger market position. Yet its not happening. Why?
Market perception perhaps. And this market is easily swayed by reviews opinions etc, as it should! but it seems that every time AMD market perception is brought down, nvidia as finger on it. And they are willing to go dirty big time. Gameworks was a stroke of genius. Its nothing but a cancer, but its being spread as a cure for games for the benefit of gamers. And by gamers I mean everyone who buys their currently available hardware, not their 2 year old hardware. And now, we have async. A dx12 spec feature in its own right. Capable of providing double digits performance gains when supported. Maxwell cant do it natively, so what do they do? fix the game with drivers! as it turns out, image quality was brought down and nobody noticed when it mattered. But that is OK!!! Its OK because performance reviews and articles were already written! the market was already swayed when it mattered. The resulting perception? = nvidia only needed drivers! its async performance is on PAR with AMD. Job done nvidia.

Only our money can keep companies honest. spend it on fallacies tricks and dishonesty and soon this will be the only choice on the market.

I said it before and I'll quote my self


AMD have countless time stated they can't optimize anything for Gameworks titles during the development process, Nvidia is the only one who can and Nvidia must approve developers's decision to add AMD's own code in a case to case basis to optimize the game for Radeon Cards, AMD have to make these optimizations with their drivers instead of being released with the game because most of the times the approval comes late in the development process or not at all, this means the game have to launch greatly favoring Nvidia's architecture.

In all Gameworks titles AMD always loses miserably the performance war on early game but always catch up at late game, the thing is, for a marketing standpoint... Late game doesn't matter.

When a game comes out every Web Site will benchmark the soul out of it, once the early benchmarks for Gameworks titles are out, AMD Reputations gets tainted for the remaining of that generation and people always perceives them as the cheap offering, even when they are in fact the best performing offering at the time (Late game)

also when casual players are looking to upgrade they only check benchmarks from Techspot, Anandtech or the like who only benchmark games when they're freshly out in the market and never revisit them with new benchmarks after AMD is done optimizing the game with their drivers thanks to Nvidia Gameworks and their shady business model.

The tessellation Scandal, Hairworks Effects, Godrays, PhysX, Gameworks... They all have something in common, they hurt AMD more than they hurt nVidia, and they grant nVidia the performance crown when games are releasing, which in the end is all that matters for newcomers and people looking to upgrade who just do a few benchmark searches when deciding for a new video card.
This right here is what I'm talking about."


We GAMERS have to stop Gameworks, if we don't I promise you, we'll end up with games being available only in one Company's hardware and locked for the competition, I absolutely understand AMD Getting hammered down by a superior product, one that destroys AMD's offering, a solid choice in the market coming from Nvidia that wins solely for its performance and supreme features and not carried by shady business decisions that hurt gamers to earn more money and make their sometimes inferior products look better, I loathe what they are doing and if we don't stop them we will face the dark ages of Gaming, Boycotts have greater power when combined with a restrictive wallet.

All those people wishing AMD to be gone, none of them understand how this industry, any industry would be changed when there's a monopoly, history has proven that this change is never for the better.
Quote:
Originally Posted by Kollock View Post

Async compute is currently forcibly disabled on public builds of Ashes for NV hardware. Whatever performance changes you are seeing driver to driver doesn't have anything to do with async compute.

The Performance Difference you're seeing with new drivers is caused by "Nvidia's game optimizations" which is a funny way to name lowering the graphic settings through drivers. mad.gif
Dargonplay is offline  
Sponsored Links
Advertisement
 
post #382 of 799 (permalink) Old 02-15-2016, 08:03 PM
New to Overclock.net
 
3DVagabond's Avatar
 
Join Date: Mar 2011
Location: Christchurch, NZ
Posts: 4
Rep: 1 (Unique: 1)
Quote:
Originally Posted by Dargonplay View Post

I said it before and I'll quote my self
We GAMERS have to stop Gameworks, if we don't I promise you, we'll end up with games being available only in one Company's hardware and locked for the competition, I absolutely understand AMD Getting hammered down by a superior product, one that destroys AMD's offering, a solid choice in the market coming from Nvidia that wins solely for its performance and supreme features and not carried by shady business decisions that hurt gamers to earn more money and make their sometimes inferior products look better, I loathe what they are doing and if we don't stop them we will face the dark ages of Gaming, Boycotts have greater power when combined with a restrictive wallet.

All those people wishing AMD to be gone, none of them understand how this industry, any industry would be changed when there's a monopoly, history has proven that this change is never for the better.

All people have to do is, 1) stop preordering and 2) don't buy the first week of release. That would be enough of a boycott to get their attention. People won't wait a week though.
3DVagabond is offline  
post #383 of 799 (permalink) Old 02-15-2016, 08:08 PM
New to Overclock.net
 
christoph's Avatar
 
Join Date: Jan 2009
Posts: 1,670
Rep: 30 (Unique: 26)
Quote:
Originally Posted by spyshagg View Post

Hi

AMD is in a fragile position without there being a good reason for it. Their graphics cards performance should warrant a much stronger market position. Yet its not happening. Why?
Market perception perhaps. And this market is easily swayed by reviews opinions etc, as it should! but it seems that every time AMD market perception is brought down, nvidia as finger on it. And they are willing to go dirty big time. Gameworks was a stroke of genius. Its nothing but a cancer, but its being spread as a cure for games for the benefit of gamers. And by gamers I mean everyone who buys their currently available hardware, not their 2 year old hardware. And now, we have async. A dx12 spec feature in its own right. Capable of providing double digits performance gains when supported. Maxwell cant do it natively, so what do they do? fix the game with drivers! as it turns out, image quality was brought down and nobody noticed when it mattered. But that is OK!!! Its OK because performance reviews and articles were already written! the market was already swayed when it mattered. The resulting perception? = nvidia only needed drivers! its async performance is on PAR with AMD. Job done nvidia.

Only our money can keep companies honest. spend it on fallacies tricks and dishonesty and soon this will be the only choice on the market.


this^

christoph is offline  
Sponsored Links
Advertisement
 
post #384 of 799 (permalink) Old 02-15-2016, 08:25 PM
New to Overclock.net
 
Join Date: Jul 2013
Location: Purgatory
Posts: 2,280
Rep: 125 (Unique: 82)
Quote:
Originally Posted by 3DVagabond View Post

All people have to do is, 1) stop preordering and 2) don't buy the first week of release. That would be enough of a boycott to get their attention. People won't wait a week though.

Hmmm, well, I went a step further and ended up getting FO4, just cause3 (before I bought this new fury card) on the console.. Having a blast in my spare time, and may soon purchaser rise of the tomb raider on a console too or may be on PC, provided I am absolutely sure that any performance degradation related to gimpworks has been completely mitigated.... Lol.. and don't really believe in pre-orders anyway...

Simplicity
provost is offline  
post #385 of 799 (permalink) Old 02-15-2016, 10:46 PM
New to Overclock.net
 
Mahigan's Avatar
 
Join Date: Aug 2015
Location: Ottawa, Canada
Posts: 1,781
Rep: 877 (Unique: 233)
Quote:
Originally Posted by Kollock View Post

Async compute is currently forcibly disabled on public builds of Ashes for NV hardware. Whatever performance changes you are seeing driver to driver doesn't have anything to do with async compute.
I can confirm that the latest shipping DX12 drivers from NV do support async compute. You'd have to ask NV how specifically it is implemented.

Now this I've got to see... Will Maxwell execute Graphics + Compute commands concurrently or will Asynchronous Compute simply mean that there is no defined order by which compute commands are executed.

AMD appear to stress that performance gains are best achieved through concurrent execution of Graphics and Compute commands whereas asynchronous compute doesn't really mean that in the computer science world.

Currently the prevailing conclusiom has been (sourced from across the web)

Nvidia executes Asynchronous compute + graphics code synchronously under DX12. Nvidia supports Async Compute through Hyper-Q, CUDA, but Hyper-Q doesn't support the additional wait conditions of barriers (a DX12 requirement). So no, there is no Async Compute + Graphics for Fermi, Kepler or Maxwell under DX12 currently.

Let me explain, Microsoft have introduced additional compute queues into 3D apps with their DX12 API:

Graphics queue for primary rendering tasks
Compute queue for supporting GPU tasks (lighting, post processing, physics etc)
Copy queue for simple data transfers

Command lists, from a specific queue, are still executed synchronously, while those in different queues can execute asynchronously (ex: concurrently and in parallel). What does asynchronous mean? Asynchronous means that the order of execution of each queue in relation to another is not defined. Work loads submitted to these queues may start or complete in a different order than they were issued. In terms of Fences and barriers, they only apply to each respective queue. When the work load in a queue is blocked by a fence, the other queues can still be running and submitting work for execution. If Synchronisation points between two or more queues are required, they can be defined and enforced by using fences.

Similar features have been available under OpenCL and CUDA for some time. The fences and signals, under DX12 map directly to a subset of the event system under OpenCL and CUDA. Under DX12, however, Barriers have additional wait conditions. These wait conditions are not supported by either OpenCL or CUDA. Instead, a write through of dirty buffers needs to be explicitly requested. Therefore Asynchronous compute + Graphics under DX12, though similar to Asynchronous compute under OpenCL and CUDA, requires explicit feature support for compatibility with the Asynchronous Compute + Graphics feature.

These new queues are also different than the classic Graphics queue. While the classic Graphics queue can be fed with compute commands, copy commands and graphics commands (draw calls), the new compute and copy queues can only accept compute and copy commands respectively. Hence their names.

For Maxwell, Compute and Graphics can't be active at the same time under DX12, currently, it is theorized that this is due to the fact that there is only a single function unit (Command Processor) rather than having access to ACEs as well. Copy commands, however, can run concurrently to Graphics and Compute commands due to the inclusion of more than one DMA engine in Maxwell. We see this when looking at how Fable Legends executes the various queues. What nvidia would need, in order to execute graphics and compute commands asynchronously, is to add support for additional barrier wait times for their Hyper-Q implementation. Why? This would expose the additional execution unit under Hyper-Q. The Hyper-Q interface used for CUDAs concurrent executions supports Asynchronous compute as we see in DX11 + Physx titles (Batman Arkham series for example). Hyper-Q is, however, not compatible with the DX12 API as of the time of writting this (for reasons mentioned above). If it was compatible, there would be a hardware limit of 31 asynchronous compute queues and 1 Graphics queue (as Anandtech reported).

So all that to say that if you fence often, you can get nvidia hardware to run the Asynchronous + Graphics code synchronously. You also have to make sure you use large batches of short running shaders, long running shaders would complicate scheduling on nvidia hardware and introduce latency. Oxide, because they were using AMD supplied code, ran into this problem in Ashes of the Singularity (according to posts over at overclock.net).

Since AMD are working with IO for the Hitman DX12 path, then you can be sure that the DX12 path will be AMD optimized. That means less fencing and longer running shaders.

For Hitman, Nvidia basically have to work with IO as well, in order to add a vendor ID specific DX12 path (like we saw Oxide do). It's probably not worth it seeing as nvidia have little to gain from DX12 over DX11. AMD, however, will likely suffer from a CPU bottle neck under Hitman DX11 (as they do under Rise of the Tomb Raider DX11). AMD have a lot to gain from working with developers on coding and optimizing a DX12 path.

So to summarize,

Nvidia do not support Async compute + Graphics under DX12 at this time or perhaps ever. Hitman's DX12 path may run like crap on nvidia hardware unless nvidia convince IO Interactive to code a vendor ID specific path and supply IO with optimized short running shaders. Basically, same thing that nvidia did with Oxide for Ashes of the Singularity (if memory serves me right). Since nvidia have little to gain from moving from DX11 to DX12, best for them to not waste time and money helping IO code a vendor ID specific path.

AMD will suffer performance issues due to a CPU bottleneck, brought on by the lack of support for DX11 multi-threaded command listing, when running the Hitman DX11 path. AMD has everything to gain in assisting IO Interactive in the implementation of a DX12 path. Asynchronous compute is just an added bonus on top of the removal of the CPU bottle neck which plagues AMD GCN under DX11 titles.

"Once you eliminate the impossible, whatever remains, no matter how improbable, must be the truth." - Arthur Conan Doyle (Sherlock Holmes)
Zen
(16 items)
Kn0wledge
(18 items)
Ryzen 5 2600
(12 items)
CPU
AMD Ryzen 9 3900x
Motherboard
Gigabyte X470 AORUS Gaming 7 Wifi
GPU
Sapphire Radeon 5700XT Pulse 8GB
RAM
32GB G.SKILL Trident Z Neo (For AMD Ryzen) Series
Hard Drive
XPG GAMMIX 512GB S11 Pro
Hard Drive
Samsung 840 Pro
Hard Drive
WD Caviar Black
Hard Drive
WD Caviar Black
Hard Drive
WD Caviar Red
Power Supply
OCZ ZX Series 1250W PSU
Case
Lian Li PC011 Dynamic XL Full-Tower Case
Operating System
Microsoft Windows 10 Pro
Monitor
ViewSonic VX3258-2KC-MHD
Keyboard
CORSAIR Gaming K95 RGB Mechanical Gaming Keyboard
Mouse
Corsair M65 PRO RGB Gaming Mouse
Audio
Asus Xonar Essence STX
CPU
Intel® Xeon E5-1680 V2 @ 4.5GHz
Motherboard
ASUS ROG Rampage IV Extreme
GPU
ASUS ROG Strix GeForce RTX 2080Ti
RAM
32GB Corsair Dominator Platinum PC3-17066
Hard Drive
WD Black NVMe M.2 2280
Hard Drive
256GB Samsung 840 Pro SSD
Hard Drive
4TB Western Digital Caviar Black
Optical Drive
ASUS internal BW-12B1ST Blu-ray writer
Power Supply
XFX PRO1050W Black Edition
Case
Cosmos II Ultra Tower
Operating System
Windows 10
Operating System
Ubuntu 14.04.3 LTS
Monitor
Dell UltraSharp 2408WFP
Monitor
Dell UltraSharp 2408WFP
Monitor
Dell UltraSharp 2408WFP
Keyboard
Corsair Vengeance® K90 Performance
Mouse
Logitech G9x Laser Mouse
Audio
ASUS Xonar Essence STX II
CPU
Ryzen 5 2600
Motherboard
Gigabyte X470 AORUS Ultra Gaming
GPU
Sapphire Radeon Fury Nitro OC+
RAM
Corsair Vengeance RGB Pro
Hard Drive
XPG SX8200 Pro
Hard Drive
OCZ Vertex 3
Hard Drive
Seagate 7200.10
Hard Drive
Seagate Scorpio
Hard Drive
Seagate Barracuda
Power Supply
XFX XxX Black Edition 750W
Case
AZZA CSAZ-211 Gaming Case
Operating System
Windows 10 Pro
▲ hide details ▲
Mahigan is offline  
post #386 of 799 (permalink) Old 02-15-2016, 10:52 PM
 
Join Date: Nov 2009
Location: Turkiye
Posts: 6,565
This came up on NX Gamer recently which I expected RedTechGaming to have brought up earlier, but anyway: Asynchronous Shaders are better for gpu tetris.
mtcn77 is offline  
post #387 of 799 (permalink) Old 02-16-2016, 01:14 AM
New to Overclock.net
 
ValSidalv21's Avatar
 
Join Date: Aug 2011
Posts: 253
Rep: 22 (Unique: 18)
Quote:
Originally Posted by Dargonplay View Post

The Performance Difference you're seeing with new drivers is caused by "Nvidia's game optimizations" which is a funny way to name lowering the graphic settings through drivers. mad.gif

I think I'll take the Oxide's developer word over yours on that one. If he says there are no noticeable differences between AMD and Nvidia then I'm inclined to believe him. He should know best, right?

On the other hand you don't seem to care that AMD optimizes tessellation if it feels it's to much, do you? Yup, because tessellation is evil, just like GameWorks rolleyes.gif
ValSidalv21 is offline  
post #388 of 799 (permalink) Old 02-16-2016, 06:11 AM
New to Overclock.net
 
Serios's Avatar
 
Join Date: Apr 2013
Posts: 1,112
Rep: 48 (Unique: 32)
Quote:
Originally Posted by ValSidalv21 View Post

I think I'll take the Oxide's developer word over yours on that one. If he says there are no noticeable differences between AMD and Nvidia then I'm inclined to believe him. He should know best, right?

On the other hand you don't seem to care that AMD optimizes tessellation if it feels it's to much, do you? Yup, because tessellation is evil, just like GameWorks rolleyes.gif
That tessellation thing is like lowering settings from ultra to high whit minimal visual impact or no visual impact at all while gaining a few fps.
Sometimes more tessellation than needed is used in Nvidia sponsored games and it affects performance without improving visual quality.
I bet Nvidia users would like a tessellation slider in Geforce Drivers.
Serios is offline  
post #389 of 799 (permalink) Old 02-16-2016, 06:42 AM
Retired but not dead!
 
mcg75's Avatar
 
Join Date: Jun 2011
Posts: 7,435
Rep: 311 (Unique: 231)
Quote:
Originally Posted by Serios View Post

That tessellation thing is like lowering settings from ultra to high whit minimal visual impact or no visual impact at all while gaining a few fps.
Sometimes more tessellation than needed is used in Nvidia sponsored games and it affects performance without improving visual quality.
I bet Nvidia users would like a tessellation slider in Geforce Drivers.

I'd like to have it in the game itself rather than drivers. Fallout 4 is a good example of how to do it properly. It has off, low, medium, high and ultra for Godrays.

Low provides 90% of the visuals that ultra does with only an 8% drop in fps. Ultra almost cuts fps in half when enabled.
mcg75 is offline  
post #390 of 799 (permalink) Old 02-16-2016, 07:28 AM
New to Overclock.net
 
airfathaaaaa's Avatar
 
Join Date: Aug 2015
Posts: 825
Rep: 62 (Unique: 31)
Quote:
Originally Posted by mcg75 View Post

I'd like to have it in the game itself rather than drivers. Fallout 4 is a good example of how to do it properly. It has off, low, medium, high and ultra for Godrays.

Low provides 90% of the visuals that ultra does with only an 8% drop in fps. Ultra almost cuts fps in half when enabled.
i wonder how it would look like on 100% probably like a game developed by jj abrams
airfathaaaaa is offline  
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
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