Overclock.net › Forums › Graphics Cards › AMD/ATI › AMD RX 480 Review Thread
New Posts  All Forums:Forum Nav:

AMD RX 480 Review Thread - Page 265

post #2641 of 3674
There was definitely something running in the background on that test with the nvidia cards. Even gaming you wouldn't have a single core running at 100% with just the game playing. My gtx970 strix boosted to 1266mhz but it also had a 56%asic so that's why. I could easily get 1450mhz with no voltage bump. Sounds to me like the 480 battle is up against a 980 at overclocked settings and is pretty even with a 970 at stock.
post #2642 of 3674
Quote:
Originally Posted by lolerk52 View Post

Yes, the scheduling is done in hardware on AMD, and in software for NVIDIA.
That means AMD requires less CPU power, but has issues with multithreaded rendering, since the hardware scheduler itself can't do it.

This is the main reason AMD cards age better. NVIDIA had to constantly create profiles and adjust drivers for their cards, while AMD needs relatively minimal work to get near full performance. So as NVIDIA drops support, their cards no longer recieve the software optimization they NEED, while AMD cards don't really care either way.

That's not to say there isn't improvement headroom in drivers, but it's far less necessary than for NVIDIA.

That's also the main cause for the power efficiency, clocks, and die size difference. The hardware schedulers are hard to clock high, they eat around 1/3rd of the chip's power, and they take up valuable die space.
That's the main reason for the jump between Kepler and Maxwell, in addition to the removal of DP.

http://images.anandtech.com/graphs/graph10446/82412.png

Looking at total system power consumption, despite the 1070 by itself taking less system power (as can be seen in Techpowerup's review), the total system power is higher on the 1070, since the CPU is working harder. And if you think that's just because the 1070 is stronger, hence stressing the CPU more, you can look at the 970 as well. Clearly taking more total system power than the 480, and yet it's still below the 480 in power consumption from the TPU review.

Thx for clearing that for me. Now as you say that it makes me wonder what kind of chip would be when nvidia would add hardware sheluder for pascal. Fail or not. Who knows XD
post #2643 of 3674
Quote:
Originally Posted by FLCLimax View Post

Warning: Spoiler! (Click to show)

Ahem.

They're doing something through drivers alright.
can't see no more than two results being marginally worse than before, and that's within 1-2 fps. it's no news nvidia cards don't gain that much over time, they're pretty much squeezed out at launch, that and a lot of attention paid to power/thermal efficiency figures,which in turn allows for things like turbo boost, makes nvidia cards nail pretty much every new launch of a GPU or a release of a latest game benchmark.
some ppl prefer that if they're on a yearly upgrade cycle, others will take poor launch performance as a basis for claiming huge benefits over time. well, 290X gains a lot more performance over those three years that this chart presents, but still most of those go 780Ti way (8 won by 780Ti to 5 won by 290X).
The question is: shouldn't 290X have performed like that from the beginning if it was clearly capable of doing so ? Are you gonna hold on playing 2012/13 titles for three or four years to get that optimal performance or are those gains just irrelevant cause those games are pretty much forgotten now.
Edited by Klocek001 - 7/10/16 at 7:55am
post #2644 of 3674
Quote:
Originally Posted by Klocek001 View Post

can't see no more than two results being marginally worse than before, and that's within 1-2 fps. it's no news nvidia cards don't gain that much over time, they're pretty much squeezed out at launch, that and a lot of attention paid to power/thermal efficiency figures,which in turn allows for things like turbo boost, makes nvidia cards nail pretty much every new launch of a GPU or a release of a latest game benchmark.
some ppl prefer that if they're on a yearly upgrade cycle, others will take poor launch performance as a basis for claiming huge benefits over time. well, 290X gains a lot more performance over those three years that this chart presents, but still most of those go 780Ti way (8 won by 780Ti to 5 won by 290X).
The question is: shouldn't 290X have performed like that from the beginning if it was clearly capable of doing so ? Are you gonna hold on playing 2012/13 titles for three or four years to get that optimal performance or are those gains just irrelevant cause those games are pretty much forgotten now.
If that were valid then the 780ti would be considered less than acceptable by comparison.
post #2645 of 3674
Quote:
Originally Posted by Klocek001 View Post

can't see no more than two results being marginally worse than before, and that's within 1-2 fps. it's no news nvidia cards don't gain that much over time, they're pretty much squeezed out at launch, that and a lot of attention paid to power/thermal efficiency figures,which in turn allows for things like turbo boost, makes nvidia cards nail pretty much every new launch of a GPU or a release of a latest game benchmark.
some ppl prefer that if they're on a yearly upgrade cycle, others will take poor launch performance as a basis for claiming huge benefits over time. well, 290X gains a lot more performance over those three years that this chart presents, but still most of those go 780Ti way (8 won by 780Ti to 5 won by 290X).
The question is: shouldn't 290X have performed like that from the beginning if it was clearly capable of doing so ? Are you gonna hold on playing 2012/13 titles for three or four years to get that optimal performance or are those gains just irrelevant cause those games are pretty much forgotten now.
I had both a gtx 470 and 480. They sure paid a lot of attention to the ability to double as a heater.
post #2646 of 3674
Quote:
Originally Posted by Klocek001 View Post

can't see no more than two results being marginally worse than before, and that's within 1-2 fps. it's no news nvidia cards don't gain that much over time, they're pretty much squeezed out at launch, that and a lot of attention paid to power/thermal efficiency figures,which in turn allows for things like turbo boost, makes nvidia cards nail pretty much every new launch of a GPU or a release of a latest game benchmark.
some ppl prefer that if they're on a yearly upgrade cycle, others will take poor launch performance as a basis for claiming huge benefits over time. well, 290X gains a lot more performance over those three years that this chart presents, but still most of those go 780Ti way (8 won by 780Ti to 5 won by 290X).
The question is: shouldn't 290X have performed like that from the beginning if it was clearly capable of doing so ? Are you gonna hold on playing 2012/13 titles for three or four years to get that optimal performance or are those gains just irrelevant cause those games are pretty much forgotten now.

Your argument assumes that AMD cards aren't competitive in their price bracket at launch. Hawaii was a competitive card for the price in Q4 2013.

The fact that AMD cards improve over time don't need to factor into a day one purchase. Product improvement over time is a consumer-positive attitude. It doesn't matter who it is; long-term support makes me want to stick with a brand. And graphics APIs haven't been moving any faster than GCN or Maxwell since the platforms define the API.

Maybe a long-term customer support attitude is the wrong call for the market it's in, but I don't think so, nor do most of my gamer friends and family. GPUs are expensive.
Parasite
(18 items)
 
  
CPUMotherboardGraphicsGraphics
i7 4770K @ 4.7GHz Z87 MPOWER (MS-7818) Sapphire Radeon 290x @1100/1500 EVGA 1080Ti SC2 Hybrid 
RAMHard DriveHard DriveCooling
G.SKILL 2133 Samsung 850 Pro Caviar Black Corsair H100 
CoolingCoolingOSMonitor
Corsair HG10 Corsair H60 Windows 7 x64 Sony XBR65X850B 
KeyboardPowerCaseMouse
CMSTORM Quickfire XT Corsair AX1200i Antec P280 Logitec G700 
Mouse PadAudio
Black, came with my NeXTcube 25 years ago. Sound Blaster Recon 3D PCIe 
  hide details  
Reply
Parasite
(18 items)
 
  
CPUMotherboardGraphicsGraphics
i7 4770K @ 4.7GHz Z87 MPOWER (MS-7818) Sapphire Radeon 290x @1100/1500 EVGA 1080Ti SC2 Hybrid 
RAMHard DriveHard DriveCooling
G.SKILL 2133 Samsung 850 Pro Caviar Black Corsair H100 
CoolingCoolingOSMonitor
Corsair HG10 Corsair H60 Windows 7 x64 Sony XBR65X850B 
KeyboardPowerCaseMouse
CMSTORM Quickfire XT Corsair AX1200i Antec P280 Logitec G700 
Mouse PadAudio
Black, came with my NeXTcube 25 years ago. Sound Blaster Recon 3D PCIe 
  hide details  
Reply
post #2647 of 3674
Quote:
Originally Posted by Drake87 View Post

I had both a gtx 470 and 480. They sure paid a lot of attention to the ability to double as a heater.
I wasn't really referring to six years ago, maxwell v1 and up
post #2648 of 3674
Goal posts
post #2649 of 3674
Quote:
Originally Posted by FLCLimax View Post

Warning: Spoiler! (Click to show)

Ahem.

They're doing something through drivers alright.
awesoke find. This easly proves somethig i was saying smile.gif proves that gimpvidia is gimping performance their cards.
post #2650 of 3674
Quote:
Originally Posted by prznar1 View Post

awesoke find. This easly proves somethig i was saying smile.gif proves that gimpvidia is gimping performance their cards.

How have you come to that conclusion?
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD/ATI
Overclock.net › Forums › Graphics Cards › AMD/ATI › AMD RX 480 Review Thread