Overclock.net › Forums › Benchmarks › Benchmarking Software and Discussion › [Various] Futuremark's Time Spy DirectX 12 "Benchmark" Compromised. Less Compute/Parallelism than Doom/Aots. Also...
New Posts  All Forums:Forum Nav:

[Various] Futuremark's Time Spy DirectX 12 "Benchmark" Compromised. Less Compute/Parallelism than Doom/Aots. Also... - Page 20

post #191 of 253
Are these Time Spy results really a surprise? Nvidia has a 70+% market share and has a good deal of influence to make sure the benchmark makes its products look great. Heavy use of Async compute like in AotS would only happen if AMD had a equal or greater market share then Nvidia. In the end the only benchmarks that will show what brand/cards have better DX12 performance are future games built on a DX12 game engine, not just a port from dx11.
post #192 of 253
I think this whole deal is comparable to having a CPU benchmark that ignores entire instruction sets from one vendor, when they should support both.
post #193 of 253
People should stop giving the "AMD/Nvidia could do better/worse" argument on this, the main point is this is not a DX12 benchmark that predicts game performance as advertised because it does not follow what DX12 games are meant to do, their "neutrality" is just a joke because they already failed at "predicting", timespy dx12 async scores vs real dx12 games for gtx1060 vs rx480 are the ultimate proof.
post #194 of 253
Fact is FM is not going to release a bench that makes Nvidia look bad. Period. They know where their market share is... rolleyes.gif
post #195 of 253
PC per interview with a 3D mark rep

https://youtu.be/8OrHZPYYY9g
post #196 of 253
That's like, iunno man, Lisa Su interviewing Raja Koduri about Freesync? devil.gif
post #197 of 253
I'll leave this resume with proof that TimeSpy fails as a benchmark by not meeting Futuremark own objectives of being a DX12 performance predictor, and slowly leave rolleyes.gif

(Click to enlarge if needed?)
post #198 of 253
Quote:
Originally Posted by Remij View Post

You know, it's weird to me. I agree with you, except that what I see is a fairly well balanced benchmark. AMD cards see gains with async compute. And Nvidia drivers disable async on Maxwell gpus and perform the same instructions except serially so that they aren't losing performance.

They're doing the same amount of work to render the scene, and Maxwell has to do it serially and thus doesn't see any gains. But that's not what people care about. You're wanting Maxwell cards to be gimped and to show how they stall when trying to run async and you guys want that to be represented in the score, right?
Huh no?? They should implement both methods, so that we can see how both cards perform under the best case scenario when conditions are optimal, instead of it being biased one way or the other.
If nvidia gain less with their async method, then that is a nvidia problem and AMD should not be punished for that by being forced to run something that does not properly utilize the hardware features.
If you have a cpu benchmark that is supposed to compare two different architectures/brands, would it be okay to ignore the instructions for just one of them leading to a skew in the result?
What is the point in benchmark software if it is not going to give a clue about actual real life performance and pros and cons of the hardware being tested?
Might as well then just run the benchmarks in a gameworks ready game and be done with it.
Edited by boxman222 - 7/20/16 at 9:02pm
post #199 of 253
Quote:
Originally Posted by magnek View Post

That's like, iunno man, Lisa Su interviewing Raja Koduri about Freesync? devil.gif

What a rubbish PR interview.. They never addressed any of the questions that have been raised.. They acted like they need to clarify to the uneducated masses what is going on, instead of answering some the questions a few (obviously) very knowledgeable users brought up.

Also funny after all of PCper's damage control for Maxwell they just ignore the fact that it's still not doing async, nearly a year later, and are even going so far as to use it as an example of what non-async looks like.. lachen.gif

As for "Firestrike shows what DX11 performance is like so Time Spy will do the same" well... the 780 and Titan compete with the 290X/390X in Firestrike, but all know what actual real-world performance is like..

Here's a hint..
Warning: Spoiler! (Click to show)

Edited by GorillaSceptre - 7/20/16 at 9:10pm
post #200 of 253
Quote:
Originally Posted by boxman222 View Post

Huh no?? They should implement both methods, so that we can see how both cards perform under the best case scenario when conditions are optimal, instead of it being biased one way or the other.
If nvidia gain less with their async method, then that is a nvidia problem and AMD should not be punished for that by being forced to run something that does not properly utilize the hardware features.
If you have a cpu benchmark that is supposed to compare two different architectures/brands, would it be okay to ignore the instructions for just one of them leading to a skew in the result?
What is the point in benchmark software if it is not going to give a clue about actual real life performance and pros and cons of the hardware being tested?
Might as well then just run the benchmarks in a gameworks ready game and be done with it.

It's like you simply choose to ignore what the Futuremark developer said.
My main PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7 6700k Asus ROG Maximus VIII Gene Nvidia GTX 1080Ti G.Skill Ripjaws 
Hard DriveOSKeyboardPower
Samsung 850 EVO  Windows 10 Razer Blackwidow Chroma EVGA Supernova 1300w 
  hide details  
Reply
My main PC
(8 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7 6700k Asus ROG Maximus VIII Gene Nvidia GTX 1080Ti G.Skill Ripjaws 
Hard DriveOSKeyboardPower
Samsung 850 EVO  Windows 10 Razer Blackwidow Chroma EVGA Supernova 1300w 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
Overclock.net › Forums › Benchmarks › Benchmarking Software and Discussion › [Various] Futuremark's Time Spy DirectX 12 "Benchmark" Compromised. Less Compute/Parallelism than Doom/Aots. Also...