Overclock.net › Forums › Industry News › Hardware News › [Various] Futuremark Releases 3DMark Time Spy DirectX 12 Benchmark
New Posts  All Forums:Forum Nav:

[Various] Futuremark Releases 3DMark Time Spy DirectX 12 Benchmark - Page 57

post #561 of 772
Quote:
Originally Posted by kfxsti View Post

But if it was done correctly to begin with, there wouldn't be a team working to answer questions lol

If we used that analogy, wouldn't we be stuck on just the release drivers?

There are few things that are perfect, especially when it come to hardware. Too many variables....just look at what we call the "silicon lottery". If the chips were made correctly, as you propose, then every chip would run the same, use the same voltage, etc.

I understand what you are saying though...so don't take my comment the wrong way. We are talking about software. The developers do have to account for the massive array of hardware that is being used.

I also think it's great that a rep from the developer is on here trying to explain what is going on.
Kyanite
(20 items)
 
Kyanite Photos
(43 photos)
 
CPUMotherboardGraphicsRAM
AMD Ryzen 5 1600 YD1600BBAEBOX MSI X370 KRAIT GAMING(MS-7A33) Gigabyte GTX1080 G1 Gaming GV-N1080G1 G.SKILL Flare X Series F4-3200C14D-16GFX 
Hard DriveHard DriveHard DriveHard Drive
Corsair Force MP500 M.2 2280 120GB OS) PNY XLR8 Gaming SSD  Silicon Power S55 480gb SSD Secondary Programs Micron SSD C400-MTFDDAC512MAM Origin/Steam Games 
Hard DriveOptical DriveCoolingOS
WD Caviar Blue WD10EZEX-75ZF5A0 Internal Back Up Samsung SH-224BB/RSBS Deepcool Captain EX 240 (White) Windows 10 
MonitorMonitorPowerCase
Acer GN246HL 24" Main Monitor 144mhz Acer S271HL 27" Secondary Monitor EVGA Supernova G2 750w Phanteks Enthoo Pro 
MouseOtherOtherOther
Kensington Pro Fit Sunbeam Rheobus Extreme 6-ch 5.25" Bay Fan Con... [2]Yate Loon 140mm fans~Front Intake Red LED  [3]APEVIA 312L-DWH 120mm White LED 
CPUMotherboardGraphicsRAM
Intel i7-4770k 4.0ghz Z87-G45 GAMING (MS-7821) Radeon RX470 Powercolor Red Devil Kingston Hyper-X Predator KHX18C9T2K2/16x 
Hard DriveHard DriveHard DriveHard Drive
Corsair LS Force 120gb SSD Kingston 300v 120gb SSD Western Digital 1tb  Seagate 250gb 
CoolingOSMonitorMonitor
Lepa AquaChanger 240 Windows 10 Acer 24" AOC 24" 
PowerCase
Rosewill 750M Rosewill Rise  
  hide details  
Reply
Kyanite
(20 items)
 
Kyanite Photos
(43 photos)
 
CPUMotherboardGraphicsRAM
AMD Ryzen 5 1600 YD1600BBAEBOX MSI X370 KRAIT GAMING(MS-7A33) Gigabyte GTX1080 G1 Gaming GV-N1080G1 G.SKILL Flare X Series F4-3200C14D-16GFX 
Hard DriveHard DriveHard DriveHard Drive
Corsair Force MP500 M.2 2280 120GB OS) PNY XLR8 Gaming SSD  Silicon Power S55 480gb SSD Secondary Programs Micron SSD C400-MTFDDAC512MAM Origin/Steam Games 
Hard DriveOptical DriveCoolingOS
WD Caviar Blue WD10EZEX-75ZF5A0 Internal Back Up Samsung SH-224BB/RSBS Deepcool Captain EX 240 (White) Windows 10 
MonitorMonitorPowerCase
Acer GN246HL 24" Main Monitor 144mhz Acer S271HL 27" Secondary Monitor EVGA Supernova G2 750w Phanteks Enthoo Pro 
MouseOtherOtherOther
Kensington Pro Fit Sunbeam Rheobus Extreme 6-ch 5.25" Bay Fan Con... [2]Yate Loon 140mm fans~Front Intake Red LED  [3]APEVIA 312L-DWH 120mm White LED 
CPUMotherboardGraphicsRAM
Intel i7-4770k 4.0ghz Z87-G45 GAMING (MS-7821) Radeon RX470 Powercolor Red Devil Kingston Hyper-X Predator KHX18C9T2K2/16x 
Hard DriveHard DriveHard DriveHard Drive
Corsair LS Force 120gb SSD Kingston 300v 120gb SSD Western Digital 1tb  Seagate 250gb 
CoolingOSMonitorMonitor
Lepa AquaChanger 240 Windows 10 Acer 24" AOC 24" 
PowerCase
Rosewill 750M Rosewill Rise  
  hide details  
Reply
post #562 of 772
You guys is a broad term. Lol
Secondly. If your having to gather a team to try and butter up the press there's another problem. And until some see proof that AMD indeed had a hand in this.. lol no one is going to believe it.
I own many many cards from both sides. Doesn't hurt me either way who's the fastest on a botched bench mark. What does bother me is you are trying to defend or stand by a product that used to brand Nvidia all within its software to begin with. Or maybe I got the wrong version. Lol then make the statements about about how timespy attempted to be the benchmark that tells you how games, on average, perform on the hardware when you average out all vendor specific bits. I just can't do anything but giggle
post #563 of 772
Quote:
Originally Posted by Hequaqua View Post

If we used that analogy, wouldn't we be stuck on just the release drivers?

There are few things that are perfect, especially when it come to hardware. Too many variables....just look at what we call the "silicon lottery". If the chips were made correctly, as you propose, then every chip would run the same, use the same voltage, etc.

I understand what you are saying though...so don't take my comment the wrong way. We are talking about software. The developers do have to account for the massive array of hardware that is being used.

I also think it's great that a rep from the developer is on here trying to explain what is going on.

I understand what your saying. I really do. But the ONLY reason this has came to show is because of what it is. There is a difference between hardware variables. And a bench mark that half uses a feature lol . I never once said be perfect. Lol I only said that if AMD truly had a hand in this.. Then i seriously doubt there would be all the pages of wanting to know what is going with this benchmark
post #564 of 772
Quote:
Originally Posted by FMJarnis View Post

The goal is to have a benchmark that gives an accurate indication how DX12 games, on average, perform on various graphics cards. To help people make educated purchasing decisions and to serve as an unbiased, neutral "yardstick" on gaming performance of various systems.

There are plenty of games out there that have various colored teams doing super special optimizations that favor one or the other architecture that may or may not show "what hardware is truly capable of". With 3DMark you know that it gives you the real deal *without* those bits that may influence hardware comparisons considerably. This also means it stays valid when a new generation of hardware arrives while those super-special-optimized games may suddenly perform much worse on the latest hardware when their optimizations no longer fit the new architecture.

"educated purchasing decisions".. Thanks Time Spy, based on your guidance i have come to the conclusion (based on the benchmark scores) that the 970 will essentially be equal to the 300 series cards in DX12/Vulkan titles. thumb.gif

If it turns out that Maxwell cards end up getting slaughtered in the future when real DX12 titles drop, I'm sure 3DMark will accept accountability? rolleyes.gif

We heard it here, 3DMark is the real deal, not those actual games. Sorry, but i find it hard to believe AMD is happy for you guys to market yourselves as the DX12 performance standard, when you aren't taking advantage of the things they've been marketing for over a year..
Edited by GorillaSceptre - 7/19/16 at 9:57am
post #565 of 772
Quote:
Originally Posted by GorillaSceptre View Post

"educated purchasing decisions".. Thanks Time Spy, based on your guidance i have come to the conclusion (based on the benchmark scores) that the 970 will essentially be equal to the 300 series cards in DX12/Vulkan titles. thumb.gif

If it turns out that Maxwell cards end up getting slaughtered in the future when real DX12 titles drop, I'm sure 3DMark will accept accountability? rolleyes.gif

We heard it here, 3DMark is the real deal, not those actual games. Sorry, but i find it hard to believe AMD is happy for you guys to market yourselves as the DX12 performance standard, when you aren't taking advantage of the things they've been marketing for over a year..

This ! Thank you! Lol
post #566 of 772
Honestly, it's all pretty clear from this post.
Quote:
Originally Posted by FMJarnis View Post

3DMark does work in parallel. Read the clarification when it is released (hopefully in a few hours, team is working on it). Graphics and Compute queues are used. Rest is up to the driver and hardware and cannot be controlled by the app.

We are naming it DX12 because it uses DX12 API. Do not confuse APIs and Feature Levels. Async compute is a feature of DX12 not tied to any feature level. However, DX12 spec does not specify how the execution of multiple queues is implemented by driver and hardware.

All current games that use DX12, use DX12 FL11. Now I think some games offer alternative paths for higher FL cards for some minor features, but all have a DX12 FL11 path for compatibility. The vast majority of future DX12 games will target DX12 FL11, because anything higher would immediately drop a massive number of older cards from the list of compatible hardware, including all Intel chips before Skylake.

And could you please stop accusing Futuremark of lying without some actual proof. AMD, NVIDIA and Intel all have access to our source code and have participated in the development. Ask them if you don't believe me.


Today Futuremark is pulling the covers off of their new Time Spy benchmark, which is being released today for all Windows editions of 3DMark. A showcase of sorts of the last decade or so of 3DMark benchmarks, Time Spy is a modern DirectX 12 benchmark implementing a number of the API's important features. All of this comes together in a demanding test for those who think their GPU hasn’t earned its keep yet.

DirectX 12 support for game engines has been coming along for a few months now. To join in the fray Futuremark has written the Time Spy benchmark on top of a pure DirectX 12 engine. This brings features such as asynchronous compute, explicit multi-adapter, and of course multi-threading/multi-core work submission improvements. All of this comes together into what I think is not only visually interesting, but also borrows a large number of gaming assets from benchmarks of 3DMarks past.





Just reading this I already made my own conclusions about your program. Whatever you try to do in damage control now, it's pretty much clear that this is a compromised benchmark.

Here - http://dictionary.cambridge.org/dictionary/english/compromise
-an agreement in an argument in which the people involved reduce their demands or change their opinion in order to agree
Here - because anything higher would immediately drop a massive number of older cards.

This is where you compromised with your DX 12 benchmark, making it half a DX12 benchmark, and now the uninformed think it's a, how were the marketing words... a showcase modern benchmark built on a pure DX12 engine encompassing various important features such as async compute. Which kind of leaves out the fact that because you think that during the Pascal days the vast majority of games will be built around FL11 (it's a benchmark, you're not making a game by the way. It kind of sets the bar for your hardware, that's where you don't kinda decide to compromise because of X reason that you think will happen in the future - oddly enough, your reason falls perfectly in line with nVidia's agenda) then you made a benchmark to do what in that case, show people their Pascal performance? Sorry, I just don't buy it, not one bit =)


It's a DX12 FL11 program that will roughly tell you how nVidia's GPU's will perform on DX12 FL11 games coded without proper async during Pascal's lifetime. Honorable mention - do not confuse DX12 and Feature Levels, that's just marketing doing its job making you think this is a full DX12 bench, no sweat . Also async is a feature, doesn't got much to do with DX12, so we coded for nVidia's implementation of the feature. But don't forget that Pascal is full DX12 with async compute, that's why we are naming it DX12. That's what you should've advertised it as.

Guess we're waiting for Volta.
Snappy
(10 items)
 
  
CPUMotherboardGraphicsRAM
i7-4790K 1.428V 4900MHz GA-Z97X-Gaming G1 WIFI-BK Sapphire R9 Fury X Corsair CMD16GX3M4A2133C9 Dominator Platinum 16... 
Hard DriveCoolingPowerCase
Sandisk Extreme Pro 240GB Corsair H110i GT Seasonic Platinum SS-760XP2 Cooler Master Storm Stryker 
MouseMouse Pad
Razer Mamba 4G 2012 Edition Genesis Soft Pad 
  hide details  
Reply
Snappy
(10 items)
 
  
CPUMotherboardGraphicsRAM
i7-4790K 1.428V 4900MHz GA-Z97X-Gaming G1 WIFI-BK Sapphire R9 Fury X Corsair CMD16GX3M4A2133C9 Dominator Platinum 16... 
Hard DriveCoolingPowerCase
Sandisk Extreme Pro 240GB Corsair H110i GT Seasonic Platinum SS-760XP2 Cooler Master Storm Stryker 
MouseMouse Pad
Razer Mamba 4G 2012 Edition Genesis Soft Pad 
  hide details  
Reply
post #567 of 772
Quote:
Originally Posted by GorillaSceptre View Post

"educated purchasing decisions".. Thanks Time Spy, based on your guidance i have come to the conclusion (based on the benchmark scores) that the 970 will essentially be equal to the 300 series cards in DX12/Vulkan titles. thumb.gif

If it turns out that Maxwell cards end up getting slaughtered in the future when real DX12 titles drop, I'm sure 3DMark will accept accountability? rolleyes.gif

We heard it here, 3DMark is the real deal, not those actual games. Sorry, but i find it hard to believe AMD is happy for you guys to market yourselves as the DX12 performance standard, when you aren't taking advantage of the things they've been marketing for over a year..

+1
Snappy
(10 items)
 
  
CPUMotherboardGraphicsRAM
i7-4790K 1.428V 4900MHz GA-Z97X-Gaming G1 WIFI-BK Sapphire R9 Fury X Corsair CMD16GX3M4A2133C9 Dominator Platinum 16... 
Hard DriveCoolingPowerCase
Sandisk Extreme Pro 240GB Corsair H110i GT Seasonic Platinum SS-760XP2 Cooler Master Storm Stryker 
MouseMouse Pad
Razer Mamba 4G 2012 Edition Genesis Soft Pad 
  hide details  
Reply
Snappy
(10 items)
 
  
CPUMotherboardGraphicsRAM
i7-4790K 1.428V 4900MHz GA-Z97X-Gaming G1 WIFI-BK Sapphire R9 Fury X Corsair CMD16GX3M4A2133C9 Dominator Platinum 16... 
Hard DriveCoolingPowerCase
Sandisk Extreme Pro 240GB Corsair H110i GT Seasonic Platinum SS-760XP2 Cooler Master Storm Stryker 
MouseMouse Pad
Razer Mamba 4G 2012 Edition Genesis Soft Pad 
  hide details  
Reply
post #568 of 772
Quote:
And until some see proof that AMD indeed had a hand in this.. lol no one is going to believe it.

What do you mean by this? Futuremark has said it builds these tests to be what they are because AMD, NV, and Intel prefer they not use vendor-specific paths. As the person who discovered that PCMark 05 used an application-specific paths that disadvantaged the VIA Nano, I'd like to note that FM has changed its overall practices since we found evidence of these kinds of problems.

I am not currently aware of any objections or issues AMD has raised about Time Spy -- and I've asked.
post #569 of 772
Quote:
Originally Posted by DigiHound View Post

What do you mean by this? Futuremark has said it builds these tests to be what they are because AMD, NV, and Intel prefer they not use vendor-specific paths. As the person who discovered that PCMark 05 used an application-specific paths that disadvantaged the VIA Nano, I'd like to note that FM has changed its overall practices since we found evidence of these kinds of problems.

I am not currently aware of any objections or issues AMD has raised about Time Spy -- and I've asked.

Yes, AMD prefers not to have one of its most toted features used so it can lose in the most famous benchmark of all. I'm completely sure of it.
Big Timmah
(13 items)
 
  
CPUMotherboardGraphicsRAM
Ryzen 5 1600 Asrock x370 Killer SLI/AC Sapphire Radeon Nitro Fury CORSAIR Vengeance LPX 16GB 3200mhz 
Hard DriveCoolingOSMonitor
PNY 480GB SSD PH-TC12DX Black Windows 10 Pro LG 29inch Ultrawide 
KeyboardPowerCaseMouse
Corsair K70 Thermaltake SMART M Series 850W NZXT S340 White Steel ATX Mid Tower Case Wireless Logitech thing 
Mouse Pad
With a supple pad  
  hide details  
Reply
Big Timmah
(13 items)
 
  
CPUMotherboardGraphicsRAM
Ryzen 5 1600 Asrock x370 Killer SLI/AC Sapphire Radeon Nitro Fury CORSAIR Vengeance LPX 16GB 3200mhz 
Hard DriveCoolingOSMonitor
PNY 480GB SSD PH-TC12DX Black Windows 10 Pro LG 29inch Ultrawide 
KeyboardPowerCaseMouse
Corsair K70 Thermaltake SMART M Series 850W NZXT S340 White Steel ATX Mid Tower Case Wireless Logitech thing 
Mouse Pad
With a supple pad  
  hide details  
Reply
post #570 of 772
Quote:
Just reading this I already made my own conclusions about your program. Whatever you try to do in damage control now, it's pretty much clear that this is a compromised benchmark.

Asynchronous compute is not a mandatory component of the DirectX 12 specification. It's ONE type of computational engine that DX12 supports. It's not the only type and it's not the sole qualifier that determines whether or not a benchmark is "really" DX12. In fact, you're committing the same error that led people to insist that DX11 benchmarks weren't "real" benchmarks if they didn't include huge amounts of tessellation (which played to the strengths of Nvidia's Fermi).

AMD has full access to Futuremark's source code. They have not raised the alarm. They have not issued press releases or whitepapers arguing against the way Futuremark implemented Time Spy. You can trust that, because AMD has a long and illustrious history of discussing why certain benchmarks aren't a fair way to evaluate the performance of its hardware. From Sysmark 2002 to the present day, when AMD has a problem with a major industry test, it makes that problem public.

Futuremark wrote a benchmark in line with the DX12 specification that's in-use today. Given the way that its past tests have evolved, it will probably patch that benchmark in the future to take advantage of higher feature levels when those feature levels are accurately reflected in other shipping titles. Choosing to emphasize compatibility over maximum feature level presentation expands the test base for DX12 graphics testing -- and yes, as a company that builds benchmarks that Intel, AMD, and NV use internally for product profiling, they have an interest in ensuring that their programs can be used to qualify current-generation products.

Also, I'd like to point out that AMD GPUs can't run a true Feature Level 12_1 at this point without falling back to 12_0 features in some areas. Each of the three companies has a different degree of support for DX12 features right now, and the fighting over feature levels reflects a continued lack of understanding among users regarding what it means to support DX12. That's understandable -- this **** is incredibly complex and often poorly understood, even by the press. But a lack of understanding by users, however understandable, doesn't translate into ethical violations by Futuremark.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [Various] Futuremark Releases 3DMark Time Spy DirectX 12 Benchmark