[News/Rumor] AMD Radeon RX Vega loses to GTX 1080 at 1080p

[StyM]

http://www.tweaktown.com/news/57313/...80p/index.html

TPU's relative performance chart shows the RX Vega being 5% slower than NVIDIA's stock GeForce GTX 1080, and 27% slower than the GTX 1080 Ti, at 1080p. Radeon RX Vega beats the Radeon R9 Fury X by 31%, and is just 11% faster than the GTX 1070 at 1080p. These results should be taken with massive handfuls of salt, and real-world testing will need to happen to ensure we're not being fooled.

The listing shows that AMD has 8GB of HBM2 on a 2048-bit memory bus, with the Vega GPU clocked at 1GHz stock, while it can be boosted up to 1.2GHz in reference form. The 8GB of HBM2 is clocked at 1GHz, and shouldn't be confused with the 11GHz+ on the GDDR5X on the new GTX 1080 11Gbps and GTX 1080 Ti cards because they use GDDR5X compared to the HBM2 tech on AMD's upcoming Radeon RX Vega.

The card will rock 9.8 TFLOPs of compute performance, which is impressive considering the GTX 1080 Ti has 10.6 TFLOPs of performance. All of this will come in a dual-slot card with a 225W TDP, powered by a 8+6-pin PCIe power connector up. AMD will reportedly have the Radeon RX Vega measuring in at 267mm.

[Kai Robinson]

What i take away from this, is that Vega is doing in 1 to 1.2Ghz what takes the 1080 1.5 to 1.6Ghz to accomplish. Clock for clock it looks like Vega beats the 1080p. More than likely, the card that has been benched is an engineering sample, not a final release.

[kromosto]

looks promising but price is very important. Even if we believe it can match 1080 performance when it is released still it has to be $50 lower than 1080 for getting attention from nvidia customers.

[kemo]

Even if it match the 1080 or slightly beats it , needing 8+6 pin power to do so is a step backwards.

[Tim]

What i take away from this, is that Vega is doing in 1 to 1.2Ghz what takes the 1080 1.5 to 1.6Ghz to accomplish. Clock for clock it looks like Vega beats the 1080p. More than likely, the card that has been benched is an engineering sample, not a final release.

I'm probably missing something big here but I don't really see how that matters a whole lot. Let's say it runs hotter and is at higher power at 1Ghz then the competition at 500-600mhz more, then, does it really matter if it does more work per clock? Sure, if they can get the clocks up on a retail stepping, great, but in the past we've seen with AMD it does come at a cost of increased heat and power.

Anyway, what Vega needs to do is beat the Ti fair and square, only then can the terrible delay (year+ after 1080) be somewhat ignored. The last couple of generations/refreshes AMD has been trailing further and further behind, and I'm sure nVidia could have already put them back in such a way that they would never catch up. GP102 is ultimately a bit of a laugh really, compared to nVidia really has.

I'm not paying much attention to these results but it does paint a picture of what everyone has been saying already.

[BeepBeep2]

These aren't results, that graph clearly says "estimation"
We're looking at a placeholder. :|

https://www.techpowerup.com/forums/t...4#post-3645267

I don't. Cooperation with AMD has really been sluggish, chaotic and unorganized, didn't get any RX 500 samples, I had to organize my own. Going by that we might not even get a Vega review sample.

[AbortRetryFail?]

I feel certain that RX Vega will be flying off shelves to owners of 1080p monitors at 60- to 75Hz ...

[StAndrew]

So it pretty much matches Ryzen then ?

[vario]

So it pretty much matches Ryzen then ?

These are not the results , its just approximation and people read too much into it.
I bet its pretty close tho ;-) , in 1080p, in dx11 titles.Which means faster than 1080 in dx 12 , and almost catching Ti in vulkans doom.
Anyhow, this is the price range where most money is made, the 1070 and 1080, have high sales and a big fat margins.
I wouldnt call that dissapointment.

[zanzabar]

if that 225W is accurate it will fall through the floor face first. the 480 runs at like 180W at 1400mhz so this thing with almost 2x the shaders running at 225 with only an 8 and 6 pin wont have the head room for overclocking or even reasonable clocking.

first they go with samsung for fail lithograph and then the hmb2 delay, it is just destined to fail.

[drmrlordx]

So it pretty much matches Ryzen then ?

. . . what?

[Aleki]

That metric also shows the 1060 matching up with a Fury.. These figures are really suspect..

[dasa]

agreed but it can really depend on the game selection
in doom for example fury x sits right between 1070 and 1080 right where that shows vega to be
however then there is games like anno 2205 where a 1060 flogs fury x at 1080p and gets close to it a 4k