Overclock.net › Forums › Industry News › Video Game News › [Anand] Fable Legends DX12 Benchmark Analysis
New Posts  All Forums:Forum Nav:

[Anand] Fable Legends DX12 Benchmark Analysis - Page 19  

post #181 of 443
Quote:
Originally Posted by Defoler View Post

So it's a canned benchmark? That makes it irrelevant? What about a pre-alpha-pre-drivers-benchmark which only concentrates on async? Is that ok? Why doesn't it makes that benchmark also irrelevant?

You are falling (intentionally or not) to the same age old traps.
The only thing that will matter, at all, is how released games with updated drivers for those games, perform. Not alpha, beta, gamma, pre, canned, artificial or what ever benchmark is being released before the release of the game.

AoTS is not something that concentrates on async. It just happens to use it. What he meant by canned is that it's a very simple level run. It's like unigine valley and heaven actually so it has some relevance. The reason people don't like that though is that this is a game benchmark, not just a benchmark, so they expected more. I think it's valid, though disappointing.

AMD is winning in most areas so I am not sure why you are acting like they aren't.
Quote:
Originally Posted by Olivon View Post

GCN = 20 % Market Share
Maxwell 2 = 80% Market Share

No need to be Einstein to understand what arch will obtain more efforts from devs.
It will be totally dumb for a studio to privilege AMD and see your sales going down.
If AMD want it, they will have to pay exclusivity and they don't have the money for that.
DX12 supremacy is an AMD technical PR trick for now and time will tell. Acts are better than speeches.

It's going to depend on what the game engine supports. For this game they only had async support in the xbox one version because the unreal engine does not support it natively. Other engines won't be like that I think.

I really wish nvidia didn't get involved in unreal engine but it seems like those 2 were destined for each other.
Quote:
Originally Posted by Dudewitbow View Post

total market share != architecture market share. Nvidia owns ~80% of the dgpu market, but not 100% of that 80 is maxwell 2(the majority is most likely still fermi and kepler)

market share is also relative if you decide to include console as well, as async is already implemented/planned in AAA titles on the console front which already runs on GCN for sure

nvidia doesn't even own 80% of the dGPU market.

I have been saying forever that most do not OC or at least keep their hardware in the overclocked state. Some do minor OCs and are happy. When I had the 970 I pushed it to 1508 before artifacts. Then I set it to default and left it. Unless you really need it why take the extra heat, power consumption etc.

When I had a 5770 I had it stock for a really long time. Then it was really showing it's age and I left it with an overclock. So I would say slower hardware is more likely to see that OC left in place.
Edited by semitope - 9/25/15 at 5:56am
post #182 of 443
Quote:
Originally Posted by Mahigan View Post

You cannot simply disable Async Compute for nVIDIA hardware. You would have to write a vendor ID specific path. This requires a lot more work for the developers and I'm not sure they'll be likely to pursue such a thing unless AMD is fronting part of the marketing costs (Gaming Evolved titles).
What about Kepler then or Maxwell 1? from what I know those GPUs can't run Async Compute at all, not even emulated.
Developers will still have to create an Nvidia specific path in the end.
post #183 of 443
Quote:
Originally Posted by Olivon View Post

GCN = 20 % Market Share
Maxwell 2 = 80% Market Share

No need to be Einstein to understand what arch will obtain more efforts from devs.
It will be totally dumb for a studio to privilege AMD and see your sales going down.
If AMD want it, they will have to pay exclusivity and they don't have the money for that.
DX12 supremacy is an AMD technical PR trick for now and time will tell. Acts are better than speeches.
Actualy GCN has a bigger market share.
post #184 of 443
Quote:
Originally Posted by Defoler View Post

If I'm not mistaken they are also using different nvidia drives. AT are using the qualified one, and TR are using the dev one?
Also since there will be more aftermarket versions fury x, I think it will always be a torn in AMD's arse when it comes to comparisons.

TR continue to label it as vanilla cards without specifying that they're using the aftermarket cooler OCed cards, that's the problem.
    
CPUMotherboardGraphicsRAM
e2140@3.2Ghz abit IP35-E HIS IceQ4 4850 4GB 667@800 5-5-5-15 
OSPower
win xp 32 bit Corsair 450VX@stock 
  hide details  
    
CPUMotherboardGraphicsRAM
e2140@3.2Ghz abit IP35-E HIS IceQ4 4850 4GB 667@800 5-5-5-15 
OSPower
win xp 32 bit Corsair 450VX@stock 
  hide details  
post #185 of 443
Thread Starter 
Quote:
Originally Posted by gamervivek View Post

TR continue to label it as vanilla cards without specifying that they're using the aftermarket cooler OCed cards, that's the problem.

TR labels things just fine. The problem is that people don't read. rolleyes.gif

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  
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  
post #186 of 443
Quote:
Originally Posted by kingduqc View Post

Released games these days are broken 6 months after release. Is a benchmark of an unfinished game really that great to conclude on performence? Considering we saw many benchmark in the past not represent well the game performance, so to for exemple.

I would go even further, games as benchmarks are only good for showing what THAT game can do, they are not useful for showing true card performance. To many games today are optimized for one card or another. Gaming benchmarks are great to tell you what card A will do in the specific game vs card B but do not give any kind of meaningful information when looking at performance of a card.

Synthetic benchmarks are not a great solution either as they can show potential "horsepower" of a card but this seldom directly translates into actual gaming experience.

What is the right way to tell? I am not surer there is one anymore other than to play games. I mean we can give a quantifiable number that says Card A can throw more frames around than Card B but in the end the difference makes little to no difference in real gaming experience.
post #187 of 443
Quote:
Originally Posted by GorillaSceptre View Post


Congratulations to both of you for completely missing my point..

The AMD argument is: 90% of people don't overclock so overclocking results don't matter.rolleyes.gif

The Nvidia argument is: Most people who buy high-end cards overclock them, so the overclocked numbers are the only ones that matter.

My underlying point was that it's personal preference, it's the individuals choice to overclock or not. So the AMD loyal's can put their fingers in their ears and pretend overclocking doesn't exist to make the Fury X look better, and the Nvidia loyal's can pretend that everyone overclocks to make the Ti look better.

At the end of the day, the best card for you, is the one that meets your requirements. There are many reasons someone may not want to overclock, so they can ignore the overclocked numbers, overclocking applies to people like me, so i'll include them.

Well, at least regarding Fiji the card doesn´t give a good overclock headroom, which makes me thing it is binned to run at as high factory clocks as possible. And all those custom NVIDIA 980 TIs have a factory overclock and also a much higher headroom for overclocking. So, the way I see it, we have two different factory policies: AMD doesn´t give much room to overclock while NVIDIA does.

Considering that, I don´t see why it wouldn´t be fair to compare a factory overclocked 980 TI to a Fury X, considering those factory overclocked custom 980 TIs can also be greatly overclocked.

Why would anyone purchase a factory overclocked 980 TI to downclock it to stock reference clocks? And seriously, almost NO ONE purchases reference 980 TIs without the intention to overclock them to at least 20%.
post #188 of 443
I am just going to quote the text which I wrote to HardOCP:
Quote:
It feels like everybody is missing the point with this test. It is a Nvidia sponsored Unreal Engine 4 game where Nvidia traditionally has been running circles with AMD cards in DX11.
Like this:
http--www.gamegpu.ru-images-stories-Test_GPU-Action-ARK_Survival_Evolved-test-arc_1920m.jpg
http--www.gamegpu.ru-images-stories-Test_GPU-Action-Unreal_Tournament_-cach-UE4_2560.jpg
http--www.gamegpu.ru-images-stories-Test_GPU-Action-Kholat-test-Kholat_2560.jpg
http--www.gamegpu.ru-images-stories-Test_GPU-Action-The_Vanishing_of_Ethan_Carter_Redux-test-EthanCarter_2560.jpg

Nvidia cards are roflstomping AMD counterparts in DX11. Now lets add in DX12 to that still nvidia sponsored UE4 Engine and we suddenly get this:

Untitled.jpg

Btw, those are all custom cards with the exception of Fury, Nano and TX

Now the tables have evened out considerably and AMD card are actually beating their counterparts from Nvidia like 290 vs 780Ti and 970, 390X vs 980, 280X vs 770, 380 vs 960 etc. how the hell did this happen confused.gif

Sure the 980Ti is still the king of the hill but that doesn't downplay the fact that AMD got considerable boosts with their older cards. But because Fury X didn't win 980Ti, people conclude that DX12 is not doing anything for the AMD lineup even tough they can just look at the data. My gues is that "most" people just looks at the top of the charts and see 980Ti sitting there so they conclude that there's nothing to see here and then come to forums to say that AMD loses even in DX12 but the reality is that AMD only looses the Top of the line battle and wins in every other category.

It feels like the whole driver overhead bottleneck which haunts AMD in DX11 is gone from AMD drivers when using DX12. Of course there's only two data point so we still can't make any final conclusions but I for one actually like these turn of events. It evens the playing field when comparing nvidia and amd cards. Now you can't just flat out say that 970 is the better card when compared to 390 as 390 might actually get quite big performance boost in DX12 games when looking at these initial DX12 results.

I don't know if this still holds any value when the actual DX12 games come out but for now, AMD cards are doing mighty fine in an engine where Nvidia has usually swiped the floor with AMD offerings. Let's wait and see what the future is bringing to us.
    
CPUMotherboardGraphicsRAM
Intel Core i7 3930K @ 4.7Ghz Asus Rampage IV GENE AMD Radeon R9 Fury X 16GB (4x4) G.Skill TridentX F3-2400C10Q 
Hard DriveCoolingCoolingCooling
Samsung SSD 850 EVO 500GB EK-Supremacy Acetal CPU Block EK-Mosfet R4F Block 2x XSPC D5 PWM Pump + EK-D5 Dual Top 
CoolingCoolingCoolingCooling
Phobya EXTREME 480 QUAD Radiator Alphacool NexXxoS ST30 240mm EK-Multioption RES X2 - 150 Advanced Reservoir Aquacomputer Aquaero 5 LT Fan Controller 
CoolingOSMonitorMonitor
Fittings, Sleeves, Fans, Sensors, Cables, etc. ... Windows 7 x64 Ultimate HP ZR30W 30" 2x Dell 2007fp 20" 
MonitorKeyboardPowerCase
Oculus Rift DK2 Razer Lycosa Mirror Edition Silverstone Strider 1200w Gold Evo Anidees AI7 Cube 
MouseMouse PadAudioAudio
Razer Mamba 2012 "White Edition" SteelSeries Experience I-2 Glass Pad Sennheiser HD800 Headphones Schiit Vali Headphone Amplifier 
  hide details  
    
CPUMotherboardGraphicsRAM
Intel Core i7 3930K @ 4.7Ghz Asus Rampage IV GENE AMD Radeon R9 Fury X 16GB (4x4) G.Skill TridentX F3-2400C10Q 
Hard DriveCoolingCoolingCooling
Samsung SSD 850 EVO 500GB EK-Supremacy Acetal CPU Block EK-Mosfet R4F Block 2x XSPC D5 PWM Pump + EK-D5 Dual Top 
CoolingCoolingCoolingCooling
Phobya EXTREME 480 QUAD Radiator Alphacool NexXxoS ST30 240mm EK-Multioption RES X2 - 150 Advanced Reservoir Aquacomputer Aquaero 5 LT Fan Controller 
CoolingOSMonitorMonitor
Fittings, Sleeves, Fans, Sensors, Cables, etc. ... Windows 7 x64 Ultimate HP ZR30W 30" 2x Dell 2007fp 20" 
MonitorKeyboardPowerCase
Oculus Rift DK2 Razer Lycosa Mirror Edition Silverstone Strider 1200w Gold Evo Anidees AI7 Cube 
MouseMouse PadAudioAudio
Razer Mamba 2012 "White Edition" SteelSeries Experience I-2 Glass Pad Sennheiser HD800 Headphones Schiit Vali Headphone Amplifier 
  hide details  
post #189 of 443
I love how now the whole AMD/NVDIA sponsored the game thing comes up yet when AoS (an AMD one) benchmarks were a thing they were considered the whole basis of the industry for a week to the point where people were encouraged to sell their Nvidia cards because "AMD is the future".
post #190 of 443
Quote:
Originally Posted by Osjur View Post

I am just going to quote the text which I wrote to HardOCP:
I don't know if this still holds any value when the actual DX12 games come out but for now, AMD cards are doing mighty fine in an engine where Nvidia has usually swiped the floor with AMD offerings. Let's wait and see what the future is bringing to us.
Holy crap, this is some improvement, coming from DX11 to DX12. And lol, 295x2 getting lower FPS than the 290x... I guess the UE was really made in favor of nVidia.
Edited by ku4eto - 9/25/15 at 9:34am
Desktop PC
(22 items)
 
ASUS ROG G72GX
(6 items)
 
 
CPUMotherboardGraphicsGraphics
AMD FX-8320 ASRock Fatal1ty 990FX Killer Sapphire Nitro+ RX480 Sapphire R9 290 Tri-X 
GraphicsRAMHard DriveHard Drive
XFX RX470 Singlefan Mushkin Redline 996996 2x4GB 2133Mhz Maxtor 6Y080L0 80GB 7200 RPM 8MB Western Digital 160GB 7200RPM 8MB 
Hard DriveHard DriveHard DriveOptical Drive
Maxtor 250GB 7200RPM 8MB Corsair Force LS WesternDigital Blue 500GB 7200RPM 16MB ASUS DVD-RW 
CoolingOSOSOS
ThermalTake Frio Silent 14 Windows 10 Enterprise Linux Mint 17.3 Rosa OphCrack 
MonitorKeyboardPowerCase
ASUS VS228HR Logitech K120 Corsair VS650 ThermalTake View 27 
MouseAudio
Bloody V5 Corsair HS30 Raptor 
CPUMotherboardGraphicsRAM
Intel Mobile Core 2 Duo P8700 G72GX NVIDIA GeForce GTX 260M  Hyundai Electronics  
RAMRAM
Hyundai Electronics  Hyundai Electronics  
  hide details  
Desktop PC
(22 items)
 
ASUS ROG G72GX
(6 items)
 
 
CPUMotherboardGraphicsGraphics
AMD FX-8320 ASRock Fatal1ty 990FX Killer Sapphire Nitro+ RX480 Sapphire R9 290 Tri-X 
GraphicsRAMHard DriveHard Drive
XFX RX470 Singlefan Mushkin Redline 996996 2x4GB 2133Mhz Maxtor 6Y080L0 80GB 7200 RPM 8MB Western Digital 160GB 7200RPM 8MB 
Hard DriveHard DriveHard DriveOptical Drive
Maxtor 250GB 7200RPM 8MB Corsair Force LS WesternDigital Blue 500GB 7200RPM 16MB ASUS DVD-RW 
CoolingOSOSOS
ThermalTake Frio Silent 14 Windows 10 Enterprise Linux Mint 17.3 Rosa OphCrack 
MonitorKeyboardPowerCase
ASUS VS228HR Logitech K120 Corsair VS650 ThermalTake View 27 
MouseAudio
Bloody V5 Corsair HS30 Raptor 
CPUMotherboardGraphicsRAM
Intel Mobile Core 2 Duo P8700 G72GX NVIDIA GeForce GTX 260M  Hyundai Electronics  
RAMRAM
Hyundai Electronics  Hyundai Electronics  
  hide details  
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Video Game News
This thread is locked  
Overclock.net › Forums › Industry News › Video Game News › [Anand] Fable Legends DX12 Benchmark Analysis