[WCCFTECH]Possible AMD Radeon RX 490 Performance Numbers Show Up in DX12 AOTS Benchmark – On Par Wit

[Pariah]

You should read it again. He said that objective gamers "who have not yet" upgraded might consider AMD cards. Not that objective gamers only buy AMD cards.

Uhh..no, considering the 2nd sentence of his post attacked nvidia users as "NV loyalists who will never buy AMD," the clear implication is that people who buy Nvidia are not objective and just buy based on brand.

 

[linkgoron]

Uhh..no, considering the 2nd sentence of his post attacked nvidia users as "NV loyalists who will never buy AMD," the clear implication is that people who buy Nvidia are not objective and just buy based on brand.

RS himself has two 1070 cards.

 

[hawtdawg]

Why do you believe this?

RX 480 GPU die only consumes under load 110W. HBM2 consumes with 2 stacks just 10W. Thats 120W for whole GPU package. And Vega 11 would most likely have at max 48 CU's - 3072 GCN cores. Reasonable Max TDP of 175W. Vega 10 - with 225W - that is competiton for Pascal.

And on top of this, we do not how in real world the patents we got in previous months affect the efficiency of the GPUs.

110w? lol. The 480 pulls north of 160w at stock speeds. HBM2 would save, at most, 15w off of that. A 1080 is around 1.8x faster than a 480 and a GP102 is 2.3x or so faster. You do the math.


If AMD pulls magic out of a hat, they'll have a GP102 competitor with about a 300w TDP. Vega was never going to be anything more than a big Polaris GPU with HBM2 (that it needs for no other reason than for power consumption). It was never going to compete with anything faster than the 1080. HBM2 is the only reason we didn't see this card months ago. You people are kidding yourselves if you think otherwise.

 

[Keysplayr]

You should read it again. He said that objective gamers "who have not yet" upgraded might consider AMD cards. Not that objective gamers only buy AMD cards.

Sorry, but we kind of read between the lines on that one. He just keeps becoming more evident as time goes on.

 

[Pariah]

RS himself has two 1070 cards.

So what? All that means is he isn't an objective gamer looking for the best price/performance based on his own definition.

 

[Shivansps]

Devastate Nvidia?, how they will pull that one off? AMD is now on the worse position EVER, i dont remember a case like this where AMD had nothing to compite with the high end sector for months. To make things worse, nvidia is now better positioned on the RX460 vs 1050 and 1060 vs 480 than it ever was with the R7 360 vs 750 and 960 vs R9 380.... not to mention Nvidia was BADLY positioned on the 970 vs R9 390, and this was only 6 months ago.
The only thing that AMD is now in a much better position is the RX470 vs 1050TI/1060 3GB, compared to what 950 vs R7 370 ever was.

And taking in consideration that nvidia is better in power consumption, less heat, superior media decoder and encoder, you cant really blame people for not buying AMD, i really hope they can get out the RX490 Vega ASAP.

 

[IllogicalGlory]

110w? lol. The 480 pulls north of 160w at stock speeds. HBM2 would save, at most, 15w off of that. A 1080 is around 1.8x faster than a 480 and a GP102 is 2.3x or so faster. You do the math.

In the broad strokes, you may be right, but the RX 480's GPU consumes 110W, which was what he was saying. The entire card is 150-160W.

 

[linkgoron]

A new 1070/1080 level GPU can't come fast enough IMO. Prices are too high IMO for that level of performance, so hopefully competition will drive prices down. Vega itself is already quite late (vs GP104/102), and I don't see AMD getting close to GP102, but maybe they'll be able to at least shake things up a bit. Hopefully we'll see some significant perf/watt improvements too, from process/arch and HBM2.

Also, when was the last time AMD released a non entry level GPU without a cut variant? Fury/Fury X, 480/470, 390/390x, 290/290x etc. I assume we'll see the same here.

And taking in consideration that nvidia is better in power consumption, less heat, superior media decoder and encoder, you cant really blame people for not buying AMD, i really hope they can get out the RX490 Vega ASAP.

For some reason, I recall people saying that Vega will have a new media encoder, I don't have a source though.

 

[HurleyBird]

Sorry, but we kind of read between the lines on that one. He just keeps becoming more evident as time goes on.

People in glass houses shouldn't throw stones, Keys.

 

[stahlhart]

All of you, stop it, and get this thread back on topic, now.
-- stahlhart

 

[PontiacGTX]

Project Quantum.


I don't think there will be cut down version of this GPU. Vega 11 will be the RX 490, and Vega 10 will most likely be RX Fury.

they could rehash cards from 400 series onto 500 series or make 485 and 495 models which would be the full gpu die or refresh then the complete Vega die could be RX Fury or wahtever they call it

290x did happen. AMDs problems are clockspeed and power consumption. Thats about it.

difference between Kepler and GCN were just 10-30w 290x power consumption was the only one which was out of range but then with Fury (X) / R9 Nano architecture tweaks and HBM the power consumption was evne lower than Hawaii

110w? lol. The 480 pulls north of 160w at stock speeds. HBM2 would save, at most, 15w off of that.

HBM saved more than 15w.HBM2 saves 8% more

 

[Pariah]

A new 1070/1080 level GPU can't come fast enough IMO. Prices are too high IMO for that level of performance, so hopefully competition will drive prices down. Vega itself is already quite late (vs GP104/102), and I don't see AMD getting close to GP102, but maybe they'll be able to at least shake things up a bit. Hopefully we'll see some significant perf/watt improvements too, from process/arch and HBM2.

Has AMD indicated that all vega products will come with HBM or are we just assuming that is the case?

1080's were selling comfortably below $600 for black Friday which is in the same price range that aftermarket 980's were selling for. If the 490 is in the 1080 performance range what price do you think AMD needs to hit to force a response from Nvidia.

 

[HurleyBird]

HBM saved more than 15w.HBM2 saves 8% more

Going down to 2 stacks, which seems fairly well corroborated (I'd love to be wrong and see GPUs with 1 TB/s bandwidth) should save a bit more on top of that.

 

[PontiacGTX]

Going down to 2 stacks, which seems fairly well corroborated (I'd love to be wrong and see GPUs with 1 TB/s bandwidth) should save a bit more on top of that.

well 2 Stack wouldnt be enough to get bandwidth for a card with 3072SP it barely has 256GB/s and that would be using max frequency speed , the RX 480 is bottlenecked with current Bamdwidth(and probably also half rops)

 

[hawtdawg]

they could rehash cards from 400 series onto 500 series or make 485 and 495 models which would be the full gpu die or refresh then the complete Vega die could be RX Fury or wahtever they call it

difference between Kepler and GCN were just 10-30w 290x power consumption was the only one which was out of range but then with Fury (X) / R9 Nano architecture tweaks and HBM the power consumption was evne lower than Hawaii

HBM saved more than 15w.HBM2 saves 8% more

All that does is compare HBM to GDDR5 at the same bandwidth. Of course 512bit GDDR 5 is going to gobble up power.

 

[hawtdawg]

http://www.anandtech.com/show/9266/amd-hbm-deep-dive/4

Here's anands own take on it. There is not even that much power usage to save. if HBM2 used 0 power, it would drop polars TDP by about 30w. 8% more than HBM1 = not a significant amount of power savings.

Like I said, roughly 15-17 watts. They might get a GP102 competitor low enough that they don't have to ship reference cards with a 240mm closed loop cooler.

 

[hawtdawg]

In the broad strokes, you may be right, but the RX 480's GPU consumes 110W, which was what he was saying. The entire card is 150-160W.

That's not true either. DRAM power consumption would take up, at most, 30w of the overall power usage in the Polaris application.

 

[IllogicalGlory]

That's not true either. DRAM power consumption would take up, at most, 30w of the overall power usage in the Polaris application.

You can argue with that, but GPU-Z and MSI Afterburner, which measure GPU power usage, agree with Glo. Maybe it's closer to 120W, but the order of magnitude is correct. Some may consume 110W. That famous GTR card reviewed by that youtube guy uses like 95W on the GPU.

Even this particularly power-hungry Gigabyte model only uses 126W on the GPU:

https://www.youtube.com/watch?v=9MWMyQ-slLI

Here's a watercooled one that draws 105W:

https://www.youtube.com/watch?v=wstG14DmZVk

 

[tential]

This card is a GTX 1070 competitor.
They are simply showing the card in its best light, which yes, allows it to compete with a card a tier above it. It's not like the 1070 and 1080 are far apart.
If the RX 490 was as fast as a 1080, that would leave a gap in the space that the 1070 occupies.... There is no way AMD does that.

This is really just common sense.

 

[Azix]

This card is a GTX 1070 competitor.
They are simply showing the card in its best light, which yes, allows it to compete with a card a tier above it. It's not like the 1070 and 1080 are far apart.
If the RX 490 was as fast as a 1080, that would leave a gap in the space that the 1070 occupies.... There is no way AMD does that.

This is really just common sense.

are you trying to say AMD would limit the power of the card just because of how nvidias products are setup?

 

[hawtdawg]

You can argue with that, but GPU-Z and MSI Afterburner, which measure GPU power usage, agree with Glo. Maybe it's closer to 120W, but the order of magnitude is correct. Some may consume 110W. That famous GTR card reviewed by that youtube guy uses like 95W on the GPU.

Even this particularly power-hungry Gigabyte model only uses 126W on the GPU:

https://www.youtube.com/watch?v=9MWMyQ-slLI

Here's a watercooled one that draws 105W:

https://www.youtube.com/watch?v=wstG14DmZVk

That's meaningless. Show me someone measuring that low of a power draw from the wall with one of those.

 

[thesmokingman]

That's meaningless. Show me someone measuring that low of a power draw from the wall with one of those.

Why don't you go buy/get one and do it, then show us?

 

[hawtdawg]

Why don't you go buy/get one and do it, then show us?

Why would I do that when all of the sites that have measured 480s from the wall have it drawing north of 160w? To the point that AMD had to release drivers to correct it? Suddenly an Afterburner TDP reading is proof to the contrary? What is wrong with you people?

 

[thesmokingman]

Why would I do that when all of the sites that have measured 480s from the wall have it drawing north of 160w? To the point that AMD had to release drivers to correct it? Suddenly an Afterburner TDP reading is proof to the contrary? What is wrong with you people?

lol, exactly you just bs'ing then.

 

[hawtdawg]

lol, exactly you just bs'ing then.

right, so I'm BSing because I wont buy a 480 to test myself, when there are tons of reviews on the internet showing Polaris 10's power draw to be north of 160w, to the point that AMD had to publicly address it. But apparently that's not proof, and some Youtube videos of Afterburner readings are?

Which do you think is more likely to be accurate, actual power draw numbers, or Afterburner?