Overclock.net › Forums › Industry News › Software News › [PCPER] NVIDIA Publishes DirectX 12 Tips for Developers
New Posts  All Forums:Forum Nav:

[PCPER] NVIDIA Publishes DirectX 12 Tips for Developers - Page 7

post #61 of 127
Quote:
Originally Posted by Klocek001 View Post

The thing that thrills me is how a combination of 980ti with Fury X would do in multi adapter mode. Can't wait for that.

Well be prepared to wait forever, because Nvidia will block you from doing that, the same way they do with Physx if you have an AMD card in the same system.
post #62 of 127
Oh come on nVidia! You don't want developers using half the resources given to them?

nVidia should change it's motto to, "We tell you how to experience a game, because that's how we intend you to enjoy it!".
Power Tower
(22 items)
 
SteamBox
(9 items)
 
Doge Miner
(7 items)
 
CPUMotherboardGraphicsRAM
Ryzen 1700X AX370-Gaming 5 AMD Radeon R9 200 Series G.Skill DDR4-2400 
RAMRAMRAMHard Drive
G.Skill DDR4-2400 G.Skill DDR4-2400 G.Skill DDR4-2400 Samsung 840 Pro 
Hard DriveHard DriveHard DriveHard Drive
CX300 Crucial 480GB Toshiba 4TB Toshbia 4TB Western Digital Black 1TB 
CoolingOSMonitorMonitor
h110i Windows 10 42" LG TV 20" Digitizer ASUS 
KeyboardPowerCaseMouse
Corsair Vengeance Mechanical Keyboard  850watt Vampire Gold Rated NZXT S340 Elite Corsair RGB FPS Mouse 
Mouse PadAudio
Borderlands Mousepad Realtek HD 
  hide details  
Reply
Power Tower
(22 items)
 
SteamBox
(9 items)
 
Doge Miner
(7 items)
 
CPUMotherboardGraphicsRAM
Ryzen 1700X AX370-Gaming 5 AMD Radeon R9 200 Series G.Skill DDR4-2400 
RAMRAMRAMHard Drive
G.Skill DDR4-2400 G.Skill DDR4-2400 G.Skill DDR4-2400 Samsung 840 Pro 
Hard DriveHard DriveHard DriveHard Drive
CX300 Crucial 480GB Toshiba 4TB Toshbia 4TB Western Digital Black 1TB 
CoolingOSMonitorMonitor
h110i Windows 10 42" LG TV 20" Digitizer ASUS 
KeyboardPowerCaseMouse
Corsair Vengeance Mechanical Keyboard  850watt Vampire Gold Rated NZXT S340 Elite Corsair RGB FPS Mouse 
Mouse PadAudio
Borderlands Mousepad Realtek HD 
  hide details  
Reply
post #63 of 127
Quote:
Originally Posted by Kollock View Post

No, we aren't switching between compute to graphics jobs rapidly. The async compute pipe has only a handful of very big jobs. Usually there is no reason to swap between graphics and compute like this, however on GCN I'm told (by console guys) that you can use this as poor-mans async compute. That is, if you do Draw(), Dispatch(), Draw(), Dispatch() the scheduler will start executing things in parallel very efficiently, but I've never tried it. Due to the fact that the fences in D3D12 require a GPU pipeline flush, it's possible that this is a good method to do things on GCN vs async compute in some cases where you'd want fine grained dependencies. Another piece of sound advice they should mention is not to use too many fences per frame. I'm hoping Vulkan doesn't have this issue.

This Do and Don't list is pretty sound advice, particularly the DXGI swapchain information. I wish we had this info 6 months ago, as we had to figure out alot of this information the hard way. The Do's list could almost be a description of how Nitrous is implemented.
So nvidia was right about the fact that to much context switching is unnecessary and i think thats why AMD said that 8 ACE are overkill.
I have two questions :
Are oxide rewriting their engine now they know what to do with maxwell and dx12? And do you have any news about the DX12 nvidia driver that is supposed to perform software sheduling?
Thnx in ad advance smile.gif
Number 3
(16 items)
 
Dust Catcher
(9 items)
 
CPUMotherboardGraphicsRAM
i7 6700k 4.5 ghz MSI Gaming M7  1080ti aorus xtreme 2x8 go Gskill 3200mhz C15 
Hard DriveHard DriveHard DriveHard Drive
crucial ssd 128 Go 2x Samsung 840 evo 250 Go Hitachi HDD 2To Toshiba HDD 2To 
Optical DriveCoolingOSMonitor
Old DVD-RW KRAKEN X61 Windows 10 x64 Swift 2k 165hz 
KeyboardPowerCaseMouse
g910 CM V1000 CM Trooper modded g502 proteus 
CPUMotherboardGraphicsRAM
Phenom ii x6 1075t Asus Sabertooth 990fx CF 2x Asus 6950 directcu ii  2x4Gb Gskill 1333 
Hard DriveCoolingOSPower
80go Intel x25 ssd series + 500 go Samsung  cooler master V6 gt Win 7 64bit ultimate edition cooler master silent pro 850w 
Case
HAF 912 advanced 
  hide details  
Reply
Number 3
(16 items)
 
Dust Catcher
(9 items)
 
CPUMotherboardGraphicsRAM
i7 6700k 4.5 ghz MSI Gaming M7  1080ti aorus xtreme 2x8 go Gskill 3200mhz C15 
Hard DriveHard DriveHard DriveHard Drive
crucial ssd 128 Go 2x Samsung 840 evo 250 Go Hitachi HDD 2To Toshiba HDD 2To 
Optical DriveCoolingOSMonitor
Old DVD-RW KRAKEN X61 Windows 10 x64 Swift 2k 165hz 
KeyboardPowerCaseMouse
g910 CM V1000 CM Trooper modded g502 proteus 
CPUMotherboardGraphicsRAM
Phenom ii x6 1075t Asus Sabertooth 990fx CF 2x Asus 6950 directcu ii  2x4Gb Gskill 1333 
Hard DriveCoolingOSPower
80go Intel x25 ssd series + 500 go Samsung  cooler master V6 gt Win 7 64bit ultimate edition cooler master silent pro 850w 
Case
HAF 912 advanced 
  hide details  
Reply
post #64 of 127
Quote:
Originally Posted by 47 Knucklehead View Post

Let's face it, the whole Async Compute thing is just way overblown tripe.

Async Compute will change the way games are written just about as much as Bulldozer was supposed to change how CPU's were developed.

rolleyes.gif

I couldn't disagree more. Asynchronous computing allows more parallelism and lowers latency because a process does not block and is executed in parallel(like SMT), it can do some computation while the message is in transit. GCN's approach is to pair a series of execution engines (ACEs) to a large array of computational processors. The computational processors are broken into groups of four or five units which share a local cache. This is akin to Intel's implementation in its processors. GCN's ACE approach "feeds" under utilized computational processors with new work every cycle. Since these computational processors can queue up a series of "tasks to do", ordered based on the priority assigned to them, then you can keep latency down all the while obtaining an optimum usage of available compute resources. In other words... you gain efficient use of your Graphics Processing Unit. This allows a developer to do more, to add more cinematic effects, in order to obtain a degree of realism which would otherwise be impossible to achieve. This translates into better and more realistic looking games.

Bulldozer was based on a CMT approach, where you pair several execution engines to few Computational Engines, per cluster, but with all of the clusters communicating by way of a single shared cache. Cache misses, saturation and latency hampers efficiency. The available compute resources are over saturated leading to a loss in performance.

The two approaches are not at all like on another.

As for parallelism, it is inevitable that developers will seek to maximize the use of available compute and graphical capabilities of an architecture due to the rather limited SOCs available in current consoles. If you would simply take the time to read up on and listen to talks on these topics, from developers (GDC, SIGGRAPH etc) you would likely understand just what developers are working on at this point in time. Developers are looking to maximize the use of available compute and graphical resources (across the board) and that's why we see Asynchronous Computing being adopted across a wide range of titles (from Console to PC titles).

For any gamer, who wants improvements in terms of the Cinematic effects and physics effects (overall realism) of games, then Asynchronous Compute would be something you would champion as this is exactly what it does.

Of course if you're a partisan of one particular brand which does not yet offer this technology, then of course you would downplay its importance.

You mentioned that AMD is holding the gaming market back, I can't see how this is true in this instance. In this instance Team Green, however, is holding gaming back wink.gif
Edited by Mahigan - 9/27/15 at 12:31pm
Kn0wledge
(20 items)
 
Pati3nce
(14 items)
 
Wisd0m
(10 items)
 
Reply
Kn0wledge
(20 items)
 
Pati3nce
(14 items)
 
Wisd0m
(10 items)
 
Reply
post #65 of 127
Quote:
Originally Posted by 47 Knucklehead View Post

Let's face it, the whole Async Compute thing is just way overblown tripe.

Async Compute will change the way games are written just about as much as Bulldozer was supposed to change how CPU's were developed.

rolleyes.gif

Except it's already being utilized on the GCN-based consoles as a NEEDED technique to increase performance. Guess which platforms studios target for AAA games?
AMD Box
(15 items)
 
  
CPUMotherboardGraphicsRAM
AMD FX-8320E @ 4.6 GHz +0.356250v offset Asus Sabertooth 990FX Rev1 eVGA GTX 970 SC ACX2.0 Patriot Viper Xtreme 2x4 GB 1600LL 8-9-8-24 1T 
Hard DriveHard DriveHard DriveCooling
Samsung 840 EVO WD Black 1 TB 32MB cache FALS WD Blue 1 TB 7200rpm EZEX Corsair H80i 
OSMonitorKeyboardPower
Windows 10 x64 HP LP2475w Logitech Illuminated Corsair TX750  
CaseMouseAudio
You don't want to know Logitech G9x Creative Sound Blaster Z 
  hide details  
Reply
AMD Box
(15 items)
 
  
CPUMotherboardGraphicsRAM
AMD FX-8320E @ 4.6 GHz +0.356250v offset Asus Sabertooth 990FX Rev1 eVGA GTX 970 SC ACX2.0 Patriot Viper Xtreme 2x4 GB 1600LL 8-9-8-24 1T 
Hard DriveHard DriveHard DriveCooling
Samsung 840 EVO WD Black 1 TB 32MB cache FALS WD Blue 1 TB 7200rpm EZEX Corsair H80i 
OSMonitorKeyboardPower
Windows 10 x64 HP LP2475w Logitech Illuminated Corsair TX750  
CaseMouseAudio
You don't want to know Logitech G9x Creative Sound Blaster Z 
  hide details  
Reply
post #66 of 127
Quote:
Originally Posted by umeng2002 View Post

Except it's already being utilized on the GCN-based consoles as a NEEDED technique to increase performance. Guess which platforms studios target for AAA games?

Incorrect. Only in exclusive titles do we or will we ever see this exploitation of hardware. Not in any cookie cut titles. Cookie cut titles are usually games that are placed across all platforms.
Edited by SpeedyVT - 9/27/15 at 2:14pm
Power Tower
(22 items)
 
SteamBox
(9 items)
 
Doge Miner
(7 items)
 
CPUMotherboardGraphicsRAM
Ryzen 1700X AX370-Gaming 5 AMD Radeon R9 200 Series G.Skill DDR4-2400 
RAMRAMRAMHard Drive
G.Skill DDR4-2400 G.Skill DDR4-2400 G.Skill DDR4-2400 Samsung 840 Pro 
Hard DriveHard DriveHard DriveHard Drive
CX300 Crucial 480GB Toshiba 4TB Toshbia 4TB Western Digital Black 1TB 
CoolingOSMonitorMonitor
h110i Windows 10 42" LG TV 20" Digitizer ASUS 
KeyboardPowerCaseMouse
Corsair Vengeance Mechanical Keyboard  850watt Vampire Gold Rated NZXT S340 Elite Corsair RGB FPS Mouse 
Mouse PadAudio
Borderlands Mousepad Realtek HD 
  hide details  
Reply
Power Tower
(22 items)
 
SteamBox
(9 items)
 
Doge Miner
(7 items)
 
CPUMotherboardGraphicsRAM
Ryzen 1700X AX370-Gaming 5 AMD Radeon R9 200 Series G.Skill DDR4-2400 
RAMRAMRAMHard Drive
G.Skill DDR4-2400 G.Skill DDR4-2400 G.Skill DDR4-2400 Samsung 840 Pro 
Hard DriveHard DriveHard DriveHard Drive
CX300 Crucial 480GB Toshiba 4TB Toshbia 4TB Western Digital Black 1TB 
CoolingOSMonitorMonitor
h110i Windows 10 42" LG TV 20" Digitizer ASUS 
KeyboardPowerCaseMouse
Corsair Vengeance Mechanical Keyboard  850watt Vampire Gold Rated NZXT S340 Elite Corsair RGB FPS Mouse 
Mouse PadAudio
Borderlands Mousepad Realtek HD 
  hide details  
Reply
post #67 of 127
Quote:
Originally Posted by ZealotKi11er View Post

Yes because they are the position to do so. They can tell the developers what to do and developers have to listen.

There are more GCN systems than Maxwell for supporting these API designs, between console and desktop combined. It's hard to feel that nVidia will have much say on game development when they are in the minority. Even if AMD only has 25-30% of the desktop sales for graphics total.
Power Tower
(22 items)
 
SteamBox
(9 items)
 
Doge Miner
(7 items)
 
CPUMotherboardGraphicsRAM
Ryzen 1700X AX370-Gaming 5 AMD Radeon R9 200 Series G.Skill DDR4-2400 
RAMRAMRAMHard Drive
G.Skill DDR4-2400 G.Skill DDR4-2400 G.Skill DDR4-2400 Samsung 840 Pro 
Hard DriveHard DriveHard DriveHard Drive
CX300 Crucial 480GB Toshiba 4TB Toshbia 4TB Western Digital Black 1TB 
CoolingOSMonitorMonitor
h110i Windows 10 42" LG TV 20" Digitizer ASUS 
KeyboardPowerCaseMouse
Corsair Vengeance Mechanical Keyboard  850watt Vampire Gold Rated NZXT S340 Elite Corsair RGB FPS Mouse 
Mouse PadAudio
Borderlands Mousepad Realtek HD 
  hide details  
Reply
Power Tower
(22 items)
 
SteamBox
(9 items)
 
Doge Miner
(7 items)
 
CPUMotherboardGraphicsRAM
Ryzen 1700X AX370-Gaming 5 AMD Radeon R9 200 Series G.Skill DDR4-2400 
RAMRAMRAMHard Drive
G.Skill DDR4-2400 G.Skill DDR4-2400 G.Skill DDR4-2400 Samsung 840 Pro 
Hard DriveHard DriveHard DriveHard Drive
CX300 Crucial 480GB Toshiba 4TB Toshbia 4TB Western Digital Black 1TB 
CoolingOSMonitorMonitor
h110i Windows 10 42" LG TV 20" Digitizer ASUS 
KeyboardPowerCaseMouse
Corsair Vengeance Mechanical Keyboard  850watt Vampire Gold Rated NZXT S340 Elite Corsair RGB FPS Mouse 
Mouse PadAudio
Borderlands Mousepad Realtek HD 
  hide details  
Reply
post #68 of 127
Quote:
Originally Posted by Noufel View Post

So nvidia was right about the fact that to much context switching is unnecessary and i think thats why AMD said that 8 ACE are overkill.
I have two questions :
Are oxide rewriting their engine now they know what to do with maxwell and dx12? And do you have any news about the DX12 nvidia driver that is supposed to perform software sheduling?
Thnx in ad advance smile.gif
Quote:
Originally Posted by Mahigan View Post

As always Kollock,

Thank you smile.gif

Are we to think that the DXGI swapchain information is part of the cause for some of the issues you ran into when optimizing for Maxwell? I think you've also indirectly answered another question I would have posed. The question would have been "Is it an nVIDIA driver issue which caused the Maxwell issues?" but it would appear that your work with nVIDIA has revealed that the Nitrous engine was perhaps doing to many of the "don't" in the list? Is this a correct assertion?


Take Care!
Source

No, DXGI swap chain has nothing to do with any vendor. It has to do with being more directly exposed to the swap buffer, as well as some big changes MS made with windows 10. The big issue is that you really don't have the equivalent of VSYNC being disabled in D3D12 that you did in D3D11. the compositor gives you a buffer to write to that is used directly. It makes things a little quirky. To top it off, the DXGI interface date back to Vista and there are some dragons burrowed deep at an OS level. It's proved pretty tricky to navigate. This becomes even more apparent when you start dealing with MGPU stuff.

I would say that Oxide does almost everything in the recommended list and virtually nothing on the don't list. The list is very good advice and pretty vendor independent; It's just good general advice for using D3D12. Many of them are really good to do with D3D11 as well. Ironically, if you refactor your engine for D3D12, you will typically end up getting much better perf in D3D11. Our D3D11 performance is very good for many of the same reasons.
post #69 of 127
Quote:
Originally Posted by lacrossewacker View Post

Nvidia releases tips on how to maximize its hardware performance.

Dear god what a scandal.
This
Us and Them
(19 items)
 
Mini Me
(11 items)
 
The Traveller
(12 items)
 
CPUMotherboardGraphicsGraphics
i7 4930K Overclocked  Rampage IV Extreme Black Edition EVGA Titan X Superclocked EVGA Titan X Superclocked 
RAMHard DriveHard DriveOptical Drive
24GB Crucial Ballistix Elite Lots of SSD's WD Caviar Black 2TB x 2 Logitech BD-RW 
CoolingOSMonitorKeyboard
Corsair H80i Windows 7 Professional Dell 34" Ultrawide 1440p x 2  Ducky Shine 2 MX Red 
PowerCaseMouseMouse Pad
Enermax MaxRevo 1500w Cooler Master Cosmos 2 R.A.T. 9 Wireless Razor Aluminum 
AudioAudioOther
Xonar Essence SXT  Sennheiser HD598 Headphones Mini Fridge full of German Beer 
CPUMotherboardGraphicsRAM
Intel i5 4790k MSI Z97I MSI GTX 970 Million Edition 8gb Crucial Ballistix Elite 
Hard DriveOptical DriveOSMonitor
1TB WD Blue 7200rpm LG USB External Windows 7 Professional 55" Lg Tv 
KeyboardPowerCase
Microsoft Wireless Bluetooth EVGA 500W EVGA Hadron Air (modified) 
CPUMotherboardGraphicsRAM
Intel 6700K Asus Maximus VIII Impact  MSI Gaming GTX 1070 16GB DDR4 Corsair Vengence 
Hard DriveOptical DriveCoolingOS
HGST 1TB x 2 2.5" HDD External Blu-Ray Cooler Master Gemini Microsoft Windows 7 64 Bit 
MonitorKeyboardCaseOther
4K LG TV Cheap Portable Fractal Design Node 202 256GB SD Card 
  hide details  
Reply
Us and Them
(19 items)
 
Mini Me
(11 items)
 
The Traveller
(12 items)
 
CPUMotherboardGraphicsGraphics
i7 4930K Overclocked  Rampage IV Extreme Black Edition EVGA Titan X Superclocked EVGA Titan X Superclocked 
RAMHard DriveHard DriveOptical Drive
24GB Crucial Ballistix Elite Lots of SSD's WD Caviar Black 2TB x 2 Logitech BD-RW 
CoolingOSMonitorKeyboard
Corsair H80i Windows 7 Professional Dell 34" Ultrawide 1440p x 2  Ducky Shine 2 MX Red 
PowerCaseMouseMouse Pad
Enermax MaxRevo 1500w Cooler Master Cosmos 2 R.A.T. 9 Wireless Razor Aluminum 
AudioAudioOther
Xonar Essence SXT  Sennheiser HD598 Headphones Mini Fridge full of German Beer 
CPUMotherboardGraphicsRAM
Intel i5 4790k MSI Z97I MSI GTX 970 Million Edition 8gb Crucial Ballistix Elite 
Hard DriveOptical DriveOSMonitor
1TB WD Blue 7200rpm LG USB External Windows 7 Professional 55" Lg Tv 
KeyboardPowerCase
Microsoft Wireless Bluetooth EVGA 500W EVGA Hadron Air (modified) 
CPUMotherboardGraphicsRAM
Intel 6700K Asus Maximus VIII Impact  MSI Gaming GTX 1070 16GB DDR4 Corsair Vengence 
Hard DriveOptical DriveCoolingOS
HGST 1TB x 2 2.5" HDD External Blu-Ray Cooler Master Gemini Microsoft Windows 7 64 Bit 
MonitorKeyboardCaseOther
4K LG TV Cheap Portable Fractal Design Node 202 256GB SD Card 
  hide details  
Reply
post #70 of 127
Quote:
Originally Posted by 47 Knucklehead View Post

Let's face it, the whole Async Compute thing is just way overblown tripe.

Async Compute will change the way games are written just about as much as Bulldozer was supposed to change how CPU's were developed.

rolleyes.gif

+1

By the time async is a big deal, GM200 and Fury X will be old news.
Glorious 4K
(23 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7-5930k @ 4.625 GHz ASUS X99-A USB 3.1 EVGA GTX 1080 Ti @ 1950/12200 Corsair Vengeance LPX 32GB DDR4-2666 C15 
Hard DriveHard DriveHard DriveCooling
OCZ Agility 3 240GB Seagate 600 Pro 240 GB WD Black 4TB Corsair H110 
CoolingCoolingCoolingCooling
Noctua NF-A14 x2 Corsair SP140 x2 Corsair AF120 x1 Corsair SP120 x4 
CoolingOSMonitorKeyboard
EVGA Titan X Hybrid AIO Windows 10 Professional x64 Sharp Aquos 4K LC-60UD27U Logitech K520 
PowerCaseMouseAudio
Corsair AX1200 Corsair Obsidian 450D Logitech Wireless Gaming Mouse G700 Sound Blaster Z Sound Card 
AudioAudioOther
Logitech Z906 Speakers Razer Chimaera 5.1 Headphones Sunbeam Rheosmart 6 fan controller 
  hide details  
Reply
Glorious 4K
(23 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7-5930k @ 4.625 GHz ASUS X99-A USB 3.1 EVGA GTX 1080 Ti @ 1950/12200 Corsair Vengeance LPX 32GB DDR4-2666 C15 
Hard DriveHard DriveHard DriveCooling
OCZ Agility 3 240GB Seagate 600 Pro 240 GB WD Black 4TB Corsair H110 
CoolingCoolingCoolingCooling
Noctua NF-A14 x2 Corsair SP140 x2 Corsair AF120 x1 Corsair SP120 x4 
CoolingOSMonitorKeyboard
EVGA Titan X Hybrid AIO Windows 10 Professional x64 Sharp Aquos 4K LC-60UD27U Logitech K520 
PowerCaseMouseAudio
Corsair AX1200 Corsair Obsidian 450D Logitech Wireless Gaming Mouse G700 Sound Blaster Z Sound Card 
AudioAudioOther
Logitech Z906 Speakers Razer Chimaera 5.1 Headphones Sunbeam Rheosmart 6 fan controller 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Software News
Overclock.net › Forums › Industry News › Software News › [PCPER] NVIDIA Publishes DirectX 12 Tips for Developers