Overclock.net › Forums › Industry News › Video Game News › [computerbase.de] DOOM + Vulkan Benchmarked.
New Posts  All Forums:Forum Nav:

[computerbase.de] DOOM + Vulkan Benchmarked. - Page 34

post #331 of 632
Quote:
Originally Posted by Eroticus View Post

Conclusion:
3DMARK is right, Game developers wrong.

That is a perfectly valid conclusion to draw from the available data. rolleyes.gif

















lachen.gif

Wondering how long it's going to take to sink in that DX11 --> DX12 or OpenGL --> Vulcan doesn't equate to no async compute --> async compute
Tsar Bomba
(12 items)
 
Fat Man
(13 items)
 
Little Boy
(29 items)
 
CPUMotherboardRAMHard Drive
Intel Core i7-6700K ASRock Fatal1ty Z170 Gaming-ITX/ac Crucial Ballistix Sport LT 32GB (2 x 16GB) DDR4... Samsung 850 EVO 500GB M.2-2280 SSD 
Hard DriveCoolingCoolingOS
Sandisk Ultra II 960GB 2.5" SSD Zalman CNPS8900 Quiet Noctua NF-F12 PWM Microsoft Windows 10 Pro 
KeyboardPowerCaseMouse
Corsair K95 Corsair SF 600W Silverstone SG05BB-LITE Corsair M65 
CPUMotherboardGraphicsGraphics
AMD A10-7850K GIGABYTE GA-F2A88X-UP4 XFX R9-290A-ENBC Radeon R9 290 XFX R9-290A-ENBC Radeon R9 290 
Hard DriveHard DriveCoolingCooling
Samsung 830 128GB Western Digital Blue 1TB XSPC Raystorm CPU/APU Waterblock - AMD XSPC D5 Photon 170 Reservoir/Pump Combo 
CoolingCoolingCoolingPower
XSPC AX360 Radiator XSPC AX240 Radiator 2x XSPC Razor R9 290X / 290 - Full Cover Water ... EVGA Supernova 1000 P2 
Case
Corsair Carbide Series Air 540 Black  
CPUMotherboardGraphicsGraphics
Intel Core i7-4770K  Asus Maximus VI Extreme Sapphire R9 290 unlocked to 290X Sapphire R9 290 unlocked to 290X 
GraphicsGraphicsRAMHard Drive
Sapphire R9 290 unlocked to 290X Sapphire R9 290 unlocked to 290X G.SKILL Sniper Series 32GB (4 x 8GB) DDR3 2400 2x SAMSUNG 840 EVO 250GB 
Hard DriveOptical DriveCoolingCooling
2x TOSHIBA PH3300U-1I72 3TB Asus BW-12B1ST/BLK/G/AS Blu-Ray/DVD/CD Writer 29x Noctua NF-F12 PWM 2x Swiftech MCP655 + Bitspower Dual D5 Mod Top 
CoolingCoolingCoolingCooling
Phobya Balancer 450 Silver Nickel Reservoir 2x Alphacool NexXxoS XT45 480mm Radiator Alphacool NexXxoS XT45 360mm Radiator Alphacool NexXxoS XT45 240mm Radiator 
CoolingCoolingCoolingCooling
Alphacool NexXxoS ST30 240mm Radiator Alphacool NexXxoS XT45 120mm Radiator 4x Koolance VID-AR290X Radeon VGA Liquid Coolin... EK Supreme LTX Intel CPU Liquid Cooling Block -... 
OSMonitorMonitorMonitor
Microsoft Windows 8.1 Pro Asus VS238H-P 23.0" Asus VS238H-P 23.0" Asus VS238H-P 23.0" 
KeyboardPowerCaseMouse
Corsair Vengeance K95 ENERMAX Maxrevo EMR1500EWT 1500W Corsair Obsidian Series 900D Razer DeathAdder 2013 
Audio
Astro Gaming A50 Circumaural Headset - Black 
  hide details  
Reply
Tsar Bomba
(12 items)
 
Fat Man
(13 items)
 
Little Boy
(29 items)
 
CPUMotherboardRAMHard Drive
Intel Core i7-6700K ASRock Fatal1ty Z170 Gaming-ITX/ac Crucial Ballistix Sport LT 32GB (2 x 16GB) DDR4... Samsung 850 EVO 500GB M.2-2280 SSD 
Hard DriveCoolingCoolingOS
Sandisk Ultra II 960GB 2.5" SSD Zalman CNPS8900 Quiet Noctua NF-F12 PWM Microsoft Windows 10 Pro 
KeyboardPowerCaseMouse
Corsair K95 Corsair SF 600W Silverstone SG05BB-LITE Corsair M65 
CPUMotherboardGraphicsGraphics
AMD A10-7850K GIGABYTE GA-F2A88X-UP4 XFX R9-290A-ENBC Radeon R9 290 XFX R9-290A-ENBC Radeon R9 290 
Hard DriveHard DriveCoolingCooling
Samsung 830 128GB Western Digital Blue 1TB XSPC Raystorm CPU/APU Waterblock - AMD XSPC D5 Photon 170 Reservoir/Pump Combo 
CoolingCoolingCoolingPower
XSPC AX360 Radiator XSPC AX240 Radiator 2x XSPC Razor R9 290X / 290 - Full Cover Water ... EVGA Supernova 1000 P2 
Case
Corsair Carbide Series Air 540 Black  
CPUMotherboardGraphicsGraphics
Intel Core i7-4770K  Asus Maximus VI Extreme Sapphire R9 290 unlocked to 290X Sapphire R9 290 unlocked to 290X 
GraphicsGraphicsRAMHard Drive
Sapphire R9 290 unlocked to 290X Sapphire R9 290 unlocked to 290X G.SKILL Sniper Series 32GB (4 x 8GB) DDR3 2400 2x SAMSUNG 840 EVO 250GB 
Hard DriveOptical DriveCoolingCooling
2x TOSHIBA PH3300U-1I72 3TB Asus BW-12B1ST/BLK/G/AS Blu-Ray/DVD/CD Writer 29x Noctua NF-F12 PWM 2x Swiftech MCP655 + Bitspower Dual D5 Mod Top 
CoolingCoolingCoolingCooling
Phobya Balancer 450 Silver Nickel Reservoir 2x Alphacool NexXxoS XT45 480mm Radiator Alphacool NexXxoS XT45 360mm Radiator Alphacool NexXxoS XT45 240mm Radiator 
CoolingCoolingCoolingCooling
Alphacool NexXxoS ST30 240mm Radiator Alphacool NexXxoS XT45 120mm Radiator 4x Koolance VID-AR290X Radeon VGA Liquid Coolin... EK Supreme LTX Intel CPU Liquid Cooling Block -... 
OSMonitorMonitorMonitor
Microsoft Windows 8.1 Pro Asus VS238H-P 23.0" Asus VS238H-P 23.0" Asus VS238H-P 23.0" 
KeyboardPowerCaseMouse
Corsair Vengeance K95 ENERMAX Maxrevo EMR1500EWT 1500W Corsair Obsidian Series 900D Razer DeathAdder 2013 
Audio
Astro Gaming A50 Circumaural Headset - Black 
  hide details  
Reply
post #332 of 632
Quote:
Originally Posted by Mahigan View Post

It does not mean that 3D Mark cheated. Some people love to take what I say out of context and that includes AMD fans. What it means is that 3D Mark chose nVIDIAs implementation over AMDs. This will likely not be the case for games because games are using AMDs implementation for the consoles.

That is too early to judge atm since we do not know precisely how it runs on AMD gpus and how it runs on Nvidia gpus. AMD gpus still gains more from turning on Asyc compute than off.
For Maxwell, they will probably have to release a statement since async doesn't function on Maxwell even with Async on because it is disabled on Maxwell driver side. Since it gives out inaccurate result.
My home PC
(16 items)
 
  
CPUMotherboardGraphicsRAM
AMD Threadripper 1950x Gigabyte Aorus X399 Gaming 7  MSI Geforce GTX 1080ti Gaming X G.Skill DDR4 3600 CL16 
Hard DriveHard DriveCoolingOS
Samsung Evo 840 500GB Samsung 960 Pro 500GB Noctua NH-U14S TR4 Windows 10 Pro 
MonitorMonitorKeyboardPower
Dell U2711 Samsung 55" 4k Corsair K70  EVGA SuperNova G2 1300W 
CaseMouseAudio
Corsair Carbide Air 540 Logitech G502 Denon AVR-X3300W 
  hide details  
Reply
My home PC
(16 items)
 
  
CPUMotherboardGraphicsRAM
AMD Threadripper 1950x Gigabyte Aorus X399 Gaming 7  MSI Geforce GTX 1080ti Gaming X G.Skill DDR4 3600 CL16 
Hard DriveHard DriveCoolingOS
Samsung Evo 840 500GB Samsung 960 Pro 500GB Noctua NH-U14S TR4 Windows 10 Pro 
MonitorMonitorKeyboardPower
Dell U2711 Samsung 55" 4k Corsair K70  EVGA SuperNova G2 1300W 
CaseMouseAudio
Corsair Carbide Air 540 Logitech G502 Denon AVR-X3300W 
  hide details  
Reply
post #333 of 632
Quote:
Originally Posted by Slomo4shO View Post

Wondering how long it's going to take to sink in that DX11 --> DX12 or OpenGL --> Vulcan doesn't equate to no async compute --> async compute

Something about Brawndo and electrolytes
post #334 of 632
Doom support intrinsic shader functions. It is a bigger feature than async support for GCN gpus. I bet intrinsic shader functions + async compute is the big reason that gcn performs so well on Doom.

We cant compare 3dmarks and Doom no matter what.
Workstation
(4 items)
 
  
CPUMotherboardGraphicsMonitor
Xeon E5-2690 Supermicro 2011 Nvidia GP100/ Vega FE Dell ultrasharp 4k 
  hide details  
Reply
Workstation
(4 items)
 
  
CPUMotherboardGraphicsMonitor
Xeon E5-2690 Supermicro 2011 Nvidia GP100/ Vega FE Dell ultrasharp 4k 
  hide details  
Reply
post #335 of 632
Quote:
Originally Posted by Kpjoslee View Post

That is too early to judge atm since we do not know precisely how it runs on AMD gpus and how it runs on Nvidia gpus. AMD gpus still gains more from turning on Asyc compute than off.
For Maxwell, they will probably have to release a statement since async doesn't function on Maxwell even with Async on because it is disabled on Maxwell driver side. Since it gives out inaccurate result.

AMD would still gain from using nVIDIAs implementation but it simply would not enjoy the same potential gains. We still have Asynchronous Compute (execution of tasks without a defined order) happening in a concurrent manner. What we do not appear to have is the parallel execution of Graphics and Compute tasks making use of a Synchronization point (fence). If a Fence were used... we would see a pretty hefty performance loss on Maxwell generation GPUs. Of course this pretty hefty performance loss would be determined by the amount of parallel tasks being submitted as each task requires a fence and the more fences... the more GPU idle time (stalls) are introduced.

3DMark claim that they make heavy use of Async Compute (20-30%)... so this sort of does not play into there being only a few parallel tasks per frame.

If you look at the Pascal gains under AotS (especially under the crazy preset) and you compare them to the gains under 3DMark Time Spy you start to see that 3DMark Time Spy is not very realistic because... well... AotS is an actual game. The same can be said for Doom under Vulkan. The Async gains are quite hefty for GCN yet tiny in comparison when running 3DMark Time Spy. Of course Vulkan does not have this OS level pre-emption that Kollock was talking about. Vulkan is a better gaming API than DX12 overall.

We saw similar behavior under Total War Warhammer (lower pascal gains) and we will likely keep seeing this behavior in titles which incorporate Async Compute. 3DMark Timespy just stands out as being out of the norm.

This can be explained by the type of implementation being used by 3DMark... you are right though... only time will tell what is going on. I could be wrong and 3DMark were magically able to enable Async Compute without a performance loss for Maxwell. That would be an industry first.
Kn0wledge
(20 items)
 
Pati3nce
(14 items)
 
Wisd0m
(10 items)
 
Reply
Kn0wledge
(20 items)
 
Pati3nce
(14 items)
 
Wisd0m
(10 items)
 
Reply
post #336 of 632
Quote:
Originally Posted by Mahigan View Post

AMD would still gain from using nVIDIAs implementation but it simply would not enjoy the same potential gains. We still have Asynchronous Compute (execution of tasks without a defined order) happening in a concurrent manner. What we do not appear to have is the parallel execution of Graphics and Compute tasks making use of a Synchronization point (fence). If a Fence were used... we would see a pretty hefty performance loss on Maxwell generation GPUs. Of course this pretty hefty performance loss would be determined by the amount of parallel tasks being submitted as each task requires a fence and the more fences... the more GPU idle time (stalls) are introduced.

3DMark claim that they make heavy use of Async Compute (20-30%)... so this sort of does not play into there being only a few parallel tasks per frame.

If you look at the Pascal gains under AotS (especially under the crazy preset) and you compare them to the gains under 3DMark Time Spy you start to see that 3DMark Time Spy is not very realistic because... well... AotS is an actual game. The same can be said for Doom under Vulkan. The Async gains are quite hefty for GCN yet tiny in comparison when running 3DMark Time Spy. Of course Vulkan does not have this OS level pre-emption that Kollock was talking about. Vulkan is a better gaming API than DX12 overall.

We saw similar behavior under Total War Warhammer (lower pascal gains) and we will likely keep seeing this behavior in titles which incorporate Async Compute. 3DMark Timespy just stands out as being out of the norm.

This can be explained by the type of implementation being used by 3DMark... you are right though... only time will tell what is going on. I could be wrong and 3DMark were magically able to enable Async Compute without a performance loss for Maxwell. That would be an industry first.

I think you are little overestimating how much Async compute gains on AoTS n Doom. AMD already benefited a lot from going from DX11 to DX12 in AoTS, without Async, and far more going from OpenGL to Vulkan.
However, within DirectX12, turning on Async only resulted in gains of 10-15% on AoTS. Under Doom, it gained about similar percentage (From MSAA to Async enabled TSSAA). Lot of people are trying to use Doom as an example of async doing massive performance boost but most of the gains came from poor OpenGL driver to great Vulkan driver lol, not from utilizing Async.


But yeah, I think it is better to wait until we can get some analysis about how it runs on either GPUs.
My home PC
(16 items)
 
  
CPUMotherboardGraphicsRAM
AMD Threadripper 1950x Gigabyte Aorus X399 Gaming 7  MSI Geforce GTX 1080ti Gaming X G.Skill DDR4 3600 CL16 
Hard DriveHard DriveCoolingOS
Samsung Evo 840 500GB Samsung 960 Pro 500GB Noctua NH-U14S TR4 Windows 10 Pro 
MonitorMonitorKeyboardPower
Dell U2711 Samsung 55" 4k Corsair K70  EVGA SuperNova G2 1300W 
CaseMouseAudio
Corsair Carbide Air 540 Logitech G502 Denon AVR-X3300W 
  hide details  
Reply
My home PC
(16 items)
 
  
CPUMotherboardGraphicsRAM
AMD Threadripper 1950x Gigabyte Aorus X399 Gaming 7  MSI Geforce GTX 1080ti Gaming X G.Skill DDR4 3600 CL16 
Hard DriveHard DriveCoolingOS
Samsung Evo 840 500GB Samsung 960 Pro 500GB Noctua NH-U14S TR4 Windows 10 Pro 
MonitorMonitorKeyboardPower
Dell U2711 Samsung 55" 4k Corsair K70  EVGA SuperNova G2 1300W 
CaseMouseAudio
Corsair Carbide Air 540 Logitech G502 Denon AVR-X3300W 
  hide details  
Reply
post #337 of 632
Quote:
Originally Posted by Eroticus View Post




You just said ASYNC = 5~15%.

So why did Nvidia lost performance ?

While Fury X gained that much ?


Why did i gained that much ? why nvidia lost permanence everywhere ... even with next gen gpus .... ?

Yes... when running the crazy preset... Pascal gains no performance from using Async Compute in AotS. Why? Because the GPU is overloaded and there are no idling resources in each GPC in order to balance the load too and speed up the processing of a long running compute task.

Like i have said before... Pascal can only handle minute amounts of Asynchronous Compute before it becomes overloaded. The architecture is not as Parallel as GCN in that respect.

  • We have to remember that a GTX 1080 is made up of 4 GPCs.
  • Each GPC can only handle one type of Context (Graphics or Compute) and not both at the same time.
  • A single Async Compute Task occupies resources in two separate GPCs.
  • These resources (when running Asynchronous Compute + Graphics) operate under a fence.
  • A Fence is a synchronization point
  • Under Maxwell.. If a Graphics task is done processing in one GPC but the Compute task is running longer in another GPC then the GPC with the Graphics task stalls (goes idle) and waits for the Compute task in the other GPC to complete before taking on new work (this is due to the Fence which forces a synchronization point between the two). This is why Oxide disabled Asynch Compute on Maxwell and why nVIDIA removed the exposed feature from their Maxwell drivers when AotS was in Alpha stages of development (see Async Compute controversy).
  • Pascal improves upon this by allowing more resources in the GPC containing the Compute task to be allocated (dynamically) in order to complte the long running compute task quicker. This reduces the amount of idle time.
  • If there are no extra resources available in the GPC containing the long running compute task due to the GPU being under heavy load then you end up with no performance benefits from enabling Async Compute.
  • You can even lose performance due to the GPU stalls introduced by a fence/synchronization point which is the case for Maxwell.
  • This is what happens under AotS at higher resolutions under the Extreme Preset (1440p and higher) or even at 1080p under the Crazy Preset.

This should help many people understand what Pascal has as it pertains to Asynchronous Compute Capabilities (Dynamic Load Balancing).
Edited by Mahigan - 7/16/16 at 6:51pm
Kn0wledge
(20 items)
 
Pati3nce
(14 items)
 
Wisd0m
(10 items)
 
Reply
Kn0wledge
(20 items)
 
Pati3nce
(14 items)
 
Wisd0m
(10 items)
 
Reply
post #338 of 632
Just as support to the above, https://developer.nvidia.com/dx12-dos-and-donts hints pretty much at a lot of what Mahigan has been saying in these threads. They're using a lot of the same jargon as AMD but read the full text and it's taking a different tack.
     
CPUMotherboardGraphicsRAM
AMD Ryzen 7 1700X @ 4.0 GHz 1.37v o/s  Biostar X370 GT7 Sapphire Radeon RX Vega 64 8GB HBM2 G.SKILL TridentZ RGB Series 16GB (2 x 8GB) @ 32... 
Hard DriveHard DriveHard DriveHard Drive
Biostar M200-240 240GB m.2 SATA SSD (Win) Samsung 850 Pro 256GB SSD (Linux) SanDisk Ultra II 480GB SSD (Games) Seagate 2TB SSHD (Storage) 
Optical DriveCoolingCoolingCooling
LG DVD RW Swiftech H220-X2 Swiftech Komodo RX-ECO Vega Water Block XSPC EX240 Rad 
CoolingOSOSMonitor
Gentle Typhoon 120mm x4 Ubuntu 17.10 w/ custom kernel Windows 10 Pro Pixio PX277 27" 1440p 144Hz DP-Adaptive Sync w... 
KeyboardPowerCaseMouse
JAR-OWL J1 Rainbow LED Mechanical SeaSonic X-Series 1050w Gold+ Corsair 760T Zowie FK1 
Mouse PadAudioAudioAudio
Razer Firefly Plantronics Rig 500e Gaming Headset Marantz Pro MPH-1 Monitoring Headphones S.M.S.L SD-793II Audio DAC + Headphone Amplifier 
Other
Xbox One Controller 
CPUMotherboardGraphicsRAM
Intel Core i7-4790k @ 4.6GHz Asrock Z97 Extreme6 EVGA GTX 1070 SC Corsair Vengeance 16GB (2x8GB) @ 2133 10-10-10-... 
Hard DriveHard DriveOptical DriveCooling
Intel 530 Series 480GB SSD Seagate 2TB SSHD LG DVD RW Cooler Master Hyper 212 Plus 
OSMonitorKeyboardPower
Windows 10 40" Vizio HDTV HP Generic EVGA SuperNOVA NEX 650w 
CaseMouseAudio
Corsair Carbide 300R Logitech Generic HyperX Cloud 
CPUMotherboardGraphicsRAM
AMD FX-6300 @ 4.3GHz 1.35v Biostar TA970 v5.3 MSI HD 6950 2GB G.Skill Ripjaws Z 16GB (4x4GB) 1600 
CoolingCoolingCoolingOS
Cryorig H5 Ultimate Corsair AF120 (White LED) x3 Corsair AF140 (White LED) x2 Windows 10 Pro 
MonitorKeyboardPowerCase
Dell S2440L Microsoft Ergonomic 4000 EVGA B2 850w Corsair 300R 
Mouse
SteelSeries Rival 300 
  hide details  
Reply
     
CPUMotherboardGraphicsRAM
AMD Ryzen 7 1700X @ 4.0 GHz 1.37v o/s  Biostar X370 GT7 Sapphire Radeon RX Vega 64 8GB HBM2 G.SKILL TridentZ RGB Series 16GB (2 x 8GB) @ 32... 
Hard DriveHard DriveHard DriveHard Drive
Biostar M200-240 240GB m.2 SATA SSD (Win) Samsung 850 Pro 256GB SSD (Linux) SanDisk Ultra II 480GB SSD (Games) Seagate 2TB SSHD (Storage) 
Optical DriveCoolingCoolingCooling
LG DVD RW Swiftech H220-X2 Swiftech Komodo RX-ECO Vega Water Block XSPC EX240 Rad 
CoolingOSOSMonitor
Gentle Typhoon 120mm x4 Ubuntu 17.10 w/ custom kernel Windows 10 Pro Pixio PX277 27" 1440p 144Hz DP-Adaptive Sync w... 
KeyboardPowerCaseMouse
JAR-OWL J1 Rainbow LED Mechanical SeaSonic X-Series 1050w Gold+ Corsair 760T Zowie FK1 
Mouse PadAudioAudioAudio
Razer Firefly Plantronics Rig 500e Gaming Headset Marantz Pro MPH-1 Monitoring Headphones S.M.S.L SD-793II Audio DAC + Headphone Amplifier 
Other
Xbox One Controller 
CPUMotherboardGraphicsRAM
Intel Core i7-4790k @ 4.6GHz Asrock Z97 Extreme6 EVGA GTX 1070 SC Corsair Vengeance 16GB (2x8GB) @ 2133 10-10-10-... 
Hard DriveHard DriveOptical DriveCooling
Intel 530 Series 480GB SSD Seagate 2TB SSHD LG DVD RW Cooler Master Hyper 212 Plus 
OSMonitorKeyboardPower
Windows 10 40" Vizio HDTV HP Generic EVGA SuperNOVA NEX 650w 
CaseMouseAudio
Corsair Carbide 300R Logitech Generic HyperX Cloud 
CPUMotherboardGraphicsRAM
AMD FX-6300 @ 4.3GHz 1.35v Biostar TA970 v5.3 MSI HD 6950 2GB G.Skill Ripjaws Z 16GB (4x4GB) 1600 
CoolingCoolingCoolingOS
Cryorig H5 Ultimate Corsair AF120 (White LED) x3 Corsair AF140 (White LED) x2 Windows 10 Pro 
MonitorKeyboardPowerCase
Dell S2440L Microsoft Ergonomic 4000 EVGA B2 850w Corsair 300R 
Mouse
SteelSeries Rival 300 
  hide details  
Reply
post #339 of 632
Quote:
Originally Posted by Mahigan View Post

Actually.. 3DMArk are wrong and we should all know this by now from our interactions with Kollock. Async Compute + Graphics is not enabled in the nVIDIA driver for their Maxwell cards... yet if you enable it in AotS... what happens? You get a performance loss attributed to the Synchronization points implemented in the AotS path. Try it yourself. This is because AotS is making use of Asynchronous Compute + Graphics which is both parallel and concurrent. Each time a parallel workload is requested... there is a synch point between the Graphics and Compute contexts involved in that workload.

Since we do not see this loss in performance for a GTX 980 Ti under 3DMark Time spy then obviously it is not doing Parallel executions because if it were there would be Synch points adverly affecting Maxwell performance whn Async is turned on (even if the driver does not support the feature).

This was my argument. It stands.

There are two other possibilities one is if the Graphics and Compute workloads were specifically tailored to take the same amount of time to complete and this would make 3DMark unreliable as a Gaming performance metric tool due to the fact that games do not behave in this manner.

The other is if a special Maxwell path was created within the benchmark which did not include the synch points but this would be misleading as the benchmark allows you to "turn on" asynch compute for the Maxwell GPUs. In any other game in which we can turn on Asynch compute or disable it... Maxwell loses performance due to these synch points. Kollock had a post explaining it all.

couldn't it just be nvidia has it enabled for AoTS in the drivers but doesn't have it enabled for time spy or other things?
strix
(14 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7 8700k Gigabyte Z370 Aorus Gaming 7 XFX Vega 56 Corsair Vegeance LED 16gb ddr4 3000 
Hard DriveHard DriveHard DriveCooling
2x Seagate Ironwolf 4TB 2x 512GB Samsung 850 Pro 1x 512GB Samsung 950 Pro Cooler Master MasterAir Pro 4  
OSMonitorKeyboardPower
Windows 10 Education Asus MG279Q Ducky One TKL RGB Evga Supernova G3 750 
CaseMouse
Cooler Master Haf XB Evo Logitech 403 
  hide details  
Reply
strix
(14 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7 8700k Gigabyte Z370 Aorus Gaming 7 XFX Vega 56 Corsair Vegeance LED 16gb ddr4 3000 
Hard DriveHard DriveHard DriveCooling
2x Seagate Ironwolf 4TB 2x 512GB Samsung 850 Pro 1x 512GB Samsung 950 Pro Cooler Master MasterAir Pro 4  
OSMonitorKeyboardPower
Windows 10 Education Asus MG279Q Ducky One TKL RGB Evga Supernova G3 750 
CaseMouse
Cooler Master Haf XB Evo Logitech 403 
  hide details  
Reply
post #340 of 632
Quote:
Originally Posted by Mahigan View Post

3DMark claim that they make heavy use of Async Compute (20-30%)... so this sort of does not play into there being only a few parallel tasks per frame.
Didn't they say 10-20%?

Anyway would it be realistic to code a game in such way that would get the most out of GCN and Pascal? Implementing both ways?
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Video Game News
Overclock.net › Forums › Industry News › Video Game News › [computerbase.de] DOOM + Vulkan Benchmarked.