AMD VEGA confirmed for 2017 H1

hyperseven · 1 May 2017 at 13:19

N19h7m4r3 said:
Can you show me that the information is accurate?

Though the onus was to disprove it. What concrete irrefutable proof is there that this is fake? Please submit a counter benchmark

muon · 1 May 2017 at 13:22

I really doubt this is a fake benchmark. Yes it might not be as helpful as a full review, just like the cherry picked Ryzen benchmarks we got before release, but its infirmation nonetheless. Ryzen did end up being pretty damn good.

eddyr · 1 May 2017 at 13:22

muon said:
I don't think it is useless unless someone can show this informatiom is fake.

Which nobody can. Only that it doesn't line up with people's prior expectations of a massive IPC improvement. In fact I would bet against one given all the information from the last few months. I can be hopeful there will be one, but that is all it is.

If it is a 'legitimate leak' of a Vega chip, without context it ends up useless anyway. Cut down/engineering sample/features disabled etc, we don't know. Whether it is fake or not doesn't end up mattering.

akarypid · 1 May 2017 at 13:23

muon said:
I really doubt this is a fake benchmark. Yes it might not be as helpful as a full review, just like the cherry picked Ryzen benchmarks we got before release, but its infirmation nonetheless. Ryzen did end up being pretty damn good.

I doubt it's showing true performance.

muon · 1 May 2017 at 13:24

eddyr said:
If it is a 'legitimate leak' of a Vega chip, without context it ends up useless anyway. Cut down/engineering sample/features disabled etc, we don't know. Whether it is fake or not doesn't end up mattering.

I agree its not ideal, but when did we on this website start ignoring ES results just because they aren't promising.

We wouldn't be having this conversation if the results blew the gtx 1080 away.

N19h7m4r3 · 1 May 2017 at 13:29

muon said:
Hardware ID identifies the card as an RX Vega and it has been run on 3D Mark which does basic checks on fake results.

TPU have identified the same specs separately.

Over to you.

muon said:
We wouldn't be having this conversation if the results blew the gtx 1080 away.

Just like the Polaris ones where they were matching GTX 980Ti, and they were claiming that Polaris is Fury X performance for 1/3 the price according to WCCFT?
We know how true that turned out.

muon · 1 May 2017 at 13:39

N19h7m4r3 said:
Just like the Polaris ones where they were matching GTX 980Ti, and they were claiming that Polaris is Fury X performance for 1/3 the price according to WCCFT?
We know how true that turned out.

I wouldn't be able to comment on the veracity of those benchmarks. A submitted 3dmark time spy bench is very different to a website claiming a few things.

Wccft have a bad reputation anyway. They post too many rumours with contradicting information (only some whuch turn out to be true). This was evident with Ryzen.

humbug · 1 May 2017 at 13:39

muon said:
I agree its not ideal, but when did we on this website start ignoring ES results just because they aren't promising.

We wouldn't be having this conversation if the results blew the gtx 1080 away.

N19h7m4r3 said:
Just like the Polaris ones where they were matching GTX 980Ti, and they were claiming that Polaris is Fury X performance for 1/3 the price according to WCCFT?
We know how true that turned out.

This ^^^^

Its from the same Chinese guy who posted 980TI beating RX 480 3DMark results, all you do is edit the GPU identifier in the XML before sending it, you can make it look like any GPU you like, this is so blatantly an ancient Fiji card.

I can't believe we are still debating the validity of a 3DMark result from a known faker, Guru3D like WCCF and TPU are just click baiting, again... they know its a fake just like they knew anything in the past from this account was fake, they still post this crap because they know it gets people sucked in and excited, it results in lots of clicks for their websites.

Is there any real news?

LoadsaMoney · 1 May 2017 at 13:42

WCCFs take on it.

Prototype of RX Vega based graphics card scores 5721 points in TimeSpy, clocked at 1200 MHz

Note: The score shown in these benchmarks is not indicative of the final product by any means, as is clear from the massive clock speed difference as well as the fact that this could be a different die being incorrectly read as the 687F:C1. A cut down variant is also not out of the question.

Before I go any further, the clock speed of this sample is a mere 1200 MHz. This means you are looking at a number of 9.8 TFLOPs and not the performance Raja Koduri promised. With a single precision compute of 12.5 TeraFLOPs per second on a GPU with 4096 cores, you are looking at a RX Vega 10 graphics card that needs to be clocked at roughly 1526 MHz. Considering the fact that the already revealed MI25 is passively cooled, it should be able to achieve this mark in due course and even exceed it. The specifications mention the HBM2 stack with 512 GB/s of bandwidth and we already know from the RX Vega Doom demos that the product will have total vRAM of 8 GB. The card will consume 225 Watts of power.

AMD’s troubles with achieving clock speeds the likes of the Geforce GTX 10 series makes me suspect that we are probably going to be looking at an RX Vega that will go, at most, toe to toe with the GTX 1080. Anything else, would be wishful thinking given the evidence given. There is however, one caveat with all of this theorizing. AMD is releasing multiple variants of Vega, and there is a possibility that this particular benchmark is of a cut down chip that is meant to go head to head with the GTX 1070. The device ID used in this case is 687F:C1, and while the first 4 digits are the same, the last digit is actually different from the ID we usually catch: 687F:C3.

http://wccftech.com/timespy-radeon-rx-vega-benchmark-12-april-demo/

eddyr · 1 May 2017 at 13:43

muon said:
I agree its not ideal, but when did we on this website start ignoring ES results just because they aren't promising.

We wouldn't be having this conversation if the results blew the gtx 1080ti away.

In an ideal reaction to the leak, it would not have spawned any conversation! It just makes sense to not bother. To have any conversation based on it requires the assumption that this should somehow represent Vega's full performance and capability and what we can expect from the released product. If the leaked results blew the 1080ti away I'd be saying essentially the same, there are no worthwhile conclusions to be drawn.

muon · 1 May 2017 at 13:43

Original RX 480 leak.

https://videocardz.com/61005/new-amd-radeon-rx-480-3dmark-benchmarks

Seems accurate to me.

N19h7m4r3 · 1 May 2017 at 13:45

LoadsaMoney said:
WCCFs take on it.

http://wccftech.com/timespy-radeon-rx-vega-benchmark-12-april-demo/

You're missing the part with his own opinion on it from facebook.

eddyr · 1 May 2017 at 13:45

muon said:
Original RX 480 leak.

https://videocardz.com/61005/new-amd-radeon-rx-480-3dmark-benchmarks

Seems accurate to me.

Was that before or after AMD told the world their Hawaii performance target with Polaris?

akarypid · 1 May 2017 at 13:47

humbug said:
This ^^^^

Its from the same Chinese guy who posted 980TI beating RX 480 3DMark results, all you do is edit the GPU identifier in the XML before sending it, you can make it look like any GPU you like, this is so blatantly an ancient Fiji card.

I can't believe we are still debating the validity of a 3DMark result from a known faker, Guru3D like WCCF and TPU are just click baiting, again... they know its a fake just like they knew anything in the past from this account was fake, they still post this crap because they know it gets people sucked in and excited, it results in lots of clicks for their websites.

Is there any real news?

What would be awesome is if this faker is on their payroll! Makes for good clickbait!

muon · 1 May 2017 at 13:47

Wccft leak.

http://wccftech.com/radeon-rx-480-3d-mark-benchmarks/

Again accurate, RX 480 ~ R9 390.

eddyr · 1 May 2017 at 13:49

muon said:
Wccft leak.

http://wccftech.com/radeon-rx-480-3d-mark-benchmarks/

Again accurate, RX 480 ~ R9 390.

wccftech said:
Its been a few weeks since AMD unveiled the new RX 480 graphics card based on the Polaris 10 GPU.

muon · 1 May 2017 at 13:50

eddyr said:

My bad. But I want to see the RX 480 > 980ti article. I can't find it.

Burr · 1 May 2017 at 13:50

I just wish they'd hurry up and release the darned card already. I understand that they want to make sure their drivers are where they should be but this is getting a beard gorgeous. Upgrading my monitor in a couple of months, AMD better have Vega on shelves by then.

humbug · 1 May 2017 at 13:52

TPU claiming RX 480 is as fast as 980TI, was fake... https://www.techpowerup.com/222007/amd-polaris-10-ellesmere-as-fast-as-gtx-980-ti-rumor

Fudzela http://www.fudzilla.com/news/graphics/40537-amd-polaris-10-gpu-performance-close-to-gtx-980-ti

Guru3D also making the same claim for sub $300.... http://www.guru3d.com/news-story/am...ffer-near-980-ti-performance-for-299-usd.html

That was 5 minutes googling, As we now know... All fake ^^^^

This latest one is even more obvious as it matches an OC Fury-X exactly.

TNA · 1 May 2017 at 13:52

eddyr said:

When I read that comment instantly my reaction was :rolleyes:

Those guys had so much fake news about the RX 480 and he is using that link to say "again accurate". Got to be having a laugh...