AMD: Speculation, Rumors, and Discussion (Archive)

Status
Not open for further replies.
GDDR5 7750 variant. Rated 75W, PCIe only, same for R250. R265/270 (150W with 1 6pin). GTX 640 (GK107 GDDR5) was PCIe at 75W.
And then when we look the actual consumption:
7750 41W average gaming, 43W peak gaming, 43W max (furmark)
R7 265 95W average gaming, 103W peak gaming, 132W max (furmark)
I'm sure the rest follow same suite, you see where I'm going with this?
 
It is even dangerous to make a VGA that uses 100% of the power available. specially in the not-super-high-end market.
 
And then when we look the actual consumption:
7750 41W average gaming, 43W peak gaming, 43W max (furmark)
R7 265 95W average gaming, 103W peak gaming, 132W max (furmark)
I'm sure the rest follow same suite, you see where I'm going with this?
Ah, I misunderstand you. I am talking about AMD rating the card 150W TBP. While only having 150W rated power imput. This has happened. If it is rated 150W I do not expect it to draw 150W. Maybe 120-130 in games. Like the R9 270, rated 150W, has 150W imputs and draws about 120W in games and 140W in GPGU/stress.
 
As far as I know most cards draw(excluding ones that don't have any power connectors) at most about 30w from pcie. Even cards like the 295x2. For cards that wants to stay in spec, they will draw only about 30w + max external power. Every card AMD has made with only a 6 pin has only drawn around 110w or lower at load.

I seriously doubt the 480 will use 150w gaming.
 
Two things:
First, what we don't know yet is the actual power consumption of the RX 480. One 6-pin connector only means >75 watts. It still could be VERY low.
Second, AotS is a showcase benchmark for AMD, yes. And here, we will probably see the least performance increase (uplift? ;)), because honestly, we expect to see improvements to the weak points of the architecture and AotS mostly shows off the strong ones. So, perf slightly above Hawaii XT could be the lower end of what to expect.
IF that was the case I would had expected them to present say Rise of the Tomb Raider using PureHair (better visuals for AMD) :)
Or show the improved tessallation behaviour-performance of Polaris with say latest Witcher 3 game.

I think AoTS was carefully chosen for the dual cheap GPU beating expensive single card narrative, but if there substantial gains anywhere else I would had expected that game to be showcased as well.
It may be the GPU does have extra surprises in other games, and AMD missed a trick with presenting that at same time as AoTS.

Cheers
 
I think AMD just has a partnership with stardock to show that game as much as possible to be honest. The game does play to their strengths generally but they could have used any other of the newer AMD partnered games and it would have probably been the same.
 
As far as I know most cards draw(excluding ones that don't have any power connectors) at most about 30w from pcie. Even cards like the 295x2. For cards that wants to stay in spec, they will draw only about 30w + max external power. Every card AMD has made with only a 6 pin has only drawn around 110w or lower at load.

I seriously doubt the 480 will use 150w gaming.
The part about 30 watts from PCIe is not true. Actually, very few gaming cards do that, Fury X at 25w via PEG is an example as well as R9 290X in reference design at slightly above 30w another. But most Geforce cards (ref.) are around 50-ish watt. Some partner cards (AMD & Nvidia!) go well over 60, some select few even over 70 watts via PEG.
 
Last edited:
I've been monitoring that reddit thread about the AoTS benchmark, it turns out the FPS figure given by AMD is an average for all three runs; 62.5fps, 1.83x utilization. It appears that the slide shown at Computex(showing 51% utilization) was incorrect, according to AMD.

I'm glad Robert finally and definitively confirmed they showed average framerate for all batches. One has to wonder what were they thinking showing '51% utilization'. Very misleading to say the least.
 
I'm glad Robert finally and definitively confirmed they showed average framerate for all batches. One has to wonder what were they thinking showing '51% utilization'. Very misleading to say the least.

Not a good look for them at all. When you ran the bench yesterday were you using v1.12.19928, or is it not public yet? Because your 290 result would be very close to a single 480x given what we know now.
 
There was one more claim made by AMD in that AotS part of presentation which kinda slipped by everyone. And that is 2x RX480 also being more (power) efficient than single 1080.

So what AMD is claiming is that single RX480 is <90W?
JwlsbY6.jpg
 
Not a good look for them at all. When you ran the bench yesterday were you using v1.12.19928, or is it not public yet? Because your 290 result would be very close to a single 480x given what we know now.
That's the most recent version on steam, yes. gog.com hasn't been updated yet, there it's still 19917.

There was one more claim made by AMD in that AotS part of presentation which kinda slipped by everyone. And that is 2x RX480 also being more (power) efficient than single 1080.

So what AMD is claiming is that single RX480 is <90W?
FWIW, it's not directly said there that both are referring to the other card shown there, although it certainly is set up in a way to creat the impression. But it could also be in relation to earlier AMD cards.
 
FWIW, it's not directly said there that both are referring to the other card shown there, although it certainly is set up in a way to creat the impression. But it could also be in relation to earlier AMD cards.
If that was the case I'm certain they would have mentioned the name of the card. It's almost a no-brainer that the $700 card they were comparing against is the GTX 1080, which they alluded to throughout their presentation.
 
If that was the case I'm certain they would have mentioned the name of the card. It's almost a no-brainer that the $700 card they were comparing against is the GTX 1080, which they alluded to throughout their presentation.
I am pretty sure, they got this covered by legal department. Remember, it's marketing: Designed to conjure up an image in your mind while at the same time being as little binding as possible.
 
There was one more claim made by AMD in that AotS part of presentation which kinda slipped by everyone. And that is 2x RX480 also being more (power) efficient than single 1080.

So what AMD is claiming is that single RX480 is <90W?

One interesting implication would be that the silicon is able to compensate for the fact that two boards also injects twice the non-execution power consumption for things like power delivery, fans, PCIe, and an extra memory bus (although this also helps performance). A larger chip on such an architecture could see better efficiency figures.

However, just as the original 51% utilization was used in an overbroad manner, what if the efficiency figure comes partially based on the CPU-limited scenario where the GPU is more heavily gated?

Less seriously, let's take it literally:
Better performance + Efficiency = 62.5 fps

What if two cards run at 60 fps + 2.5x efficiency = 62.5?
Math has not had many friends in marketing for the last few days.
 
As far as I know most cards draw(excluding ones that don't have any power connectors) at most about 30w from pcie. Even cards like the 295x2. For cards that wants to stay in spec, they will draw only about 30w + max external power. Every card AMD has made with only a 6 pin has only drawn around 110w or lower at load.

I seriously doubt the 480 will use 150w gaming.


Yeah, I think its going to use around 130 watts.
 
IF that was the case I would had expected them to present say Rise of the Tomb Raider using PureHair (better visuals for AMD) :)
Or show the improved tessallation behaviour-performance of Polaris with say latest Witcher 3 game.

I think AoTS was carefully chosen for the dual cheap GPU beating expensive single card narrative, but if there substantial gains anywhere else I would had expected that game to be showcased as well.
It may be the GPU does have extra surprises in other games, and AMD missed a trick with presenting that at same time as AoTS.

Cheers

Yes, but both Rise of TR and Witcher3 games are nvidia gamework title. Despite that it use technology from AMD in the case of TR.

Honestly, i have the feeling they wanted to show as little possible of "performance" of thoses gpu's.... nearly no benchmark have been shown outside this small comparaison on AoTS. Like the 150W "Power" who is way enough vague. Even the >5Tflops instead of give a right number.

Several reason maybe behind it. 1060Ti who should launch soon. or they have yet not completely fix the clock rate.

The last time i have seen this from AMD, it was on the FuryX or Nano launch ( not remember ), and the gpu was finally coming with a bit more Tflops than previous initial presentation and higher core speed. ( should have been the Nano )
 
Last edited:
One interesting implication would be that the silicon is able to compensate for the fact that two boards also injects twice the non-execution power consumption for things like power delivery, fans, PCIe, and an extra memory bus (although this also helps performance). A larger chip on such an architecture could see better efficiency figures.

However, just as the original 51% utilization was used in an overbroad manner, what if the efficiency figure comes partially based on the CPU-limited scenario where the GPU is more heavily gated?

Less seriously, let's take it literally:
Better performance + Efficiency = 62.5 fps

What if two cards run at 60 fps + 2.5x efficiency = 62.5?
Math has not had many friends in marketing for the last few days.

Or maybe they meant that since the cards were actually at 91.5% utilization, that the materially false 51% utilization claim(performance) + 40.5% utilization(efficiency) = 62.5 fps.
 
Or maybe they meant that since the cards were actually at 91.5% utilization, that the materially false 51% utilization claim(performance) + 40.5% utilization(efficiency) = 62.5 fps.
I don't understand these strange theories, Robert Hallock posted it straight on Reddit, 62.5 FPS is the average FPS result, and utilization per batch was 51%, 71.9% and 92.3 %, it's dead simple.
62.5 FPS is more than 58.7 FPS > Better Performance
Better efficiency means simply that they achieve more FPS per watt in that test, than a GTX 1080 does
 
Status
Not open for further replies.
Back
Top