Radeon RX 480 "Polaris" Launched at $199

[AllBodies]

https://forums.overclockers.co.uk/showthread.php?t=18720374

Here

Cheers.

Judging by that, it does suggest the performance will be a touch above the 980 and 390X. But not as high as 980Ti and FuryX.

Obviously, want to see full reviews.

But expectation of ~Fury pro performance for £160-180, stunning


And what is the $300 card? Everything crossed for a 480X which is mildly faster (like 5%) than a 1070 for £260-280!

 

[SiDeards73]

https://forums.overclockers.co.uk/showthread.php?t=18720374

Here



I am a dirty old Docker and forgive my ignorance but showing a test where your GPUs can only hit 50% utilization doesn't really do them any favours. They are all clued up with Oxide as well but like I said, smoke and mirrors.

I agree, but the demo was using DX12's mgpu and that is solely down to the Dev to implement and set the performance for the gpu's, i guess it was the only thing they could really use at the time. They did not want to show single GPU stats so that left them with either using Xfire or Mgpu, Xfire is a joke as mos of us know, so they chose the easy route of DX12 and mgpu, i think AOTS is the only game using that.

With that said, its down to Oxide to push the gpu's in an mgpu setup, not AMD, but i also wish they could have shown single gpu stats.

If NDA ends on the 29th, i doubt we will be seeing anything with regards to single GPU until except of course leaks.

 

[hyperseven]

With that said, its down to Oxide to push the gpu's in an mgpu setup, not AMD, but i also wish they could have shown single gpu stats.

Its an AMD partnered game, so why shouldn't their engineers help implement it?

 

[SiDeards73]

Its an AMD partnered game, so why shouldn't their engineers help implement it?

Dunno, from what i understand the dev's set the aggressiveness of mgpu, theres probably quite a bit of work involved in it, and it was new tech so i doubt any work was done to harness the new tech etc.

In some respects it actually paints the 480 in a better picture, as the cards were not being pushed hard by the game as its not coded for the cards i guess, so that level of performance on an unoptomised game is pretty promising.

 

[tavtavtav]

Cheers.

Judging by that, it does suggest the performance will be a touch above the 980 and 390X. But not as high as 980Ti and FuryX.

Obviously, want to see full reviews.

But expectation of ~Fury pro performance for £160-180, stunning


And what is the $300 card? Everything crossed for a 480X which is mildly faster (like 5%) than a 1070 for £260-280!

I'm not convinced there will be a cars faster than a 1070.

If there is I'd backlfip down to Singapore to pick one up personally...

 

[Greebo]

And the test was at 1440P, so that puts it quite a bit ahead of my AoTS run with an overclocked Titan X. That's pretty incredible really.

Average FR 49.5, GPU Duo(1) @1185/600, Normal FR 59.9, Medium FR 52.0, Heavy FR 40.6, CPU 5960X @4.9, AMDMatt
Average FR 46.8, GPU TitanX @1443/1778, Normal FR 54.9, Medium FR 47.2, Heavy FR 40.6, CPU 5930k @4.5, Klunt Bumskrint
Average FR 46.3, GPU Fury X @1150/545, Normal FR 55.0, Medium FR 47.9, Heavy FR 38.9, CPU 3770k @4.2, Dygaza
Average FR 45.5, GPU TitanX @1501/2002, Normal FR 51.3, Medium FR 45.1, Heavy FR 41.1, CPU 5960X @4.0, Kaapstad
Average FR 44.1, GPU 980 Ti @1470/2030, Normal FR 50.5, Medium FR 45.3, Heavy FR 38.1, CPU 4790T @2.7, Telecaster
Average FR 38.6, GPU TitanX @1386/2005, Normal FR 44.6, Medium FR 38.6, Heavy FR 33.9, CPU 3930k @4.2, Gregster

From the Reddit page, this is meant to be a single 480 result

http://www.ashesofthesingularity.co...-details/8b748568-fc96-4e48-9fed-22666a7149f5

Normal 45.5 Medium 41.8 Heavy 34.3

So on that s basis in AOS it beats the slowest Titanx on your list. Pretty impressive for a $200 card.

 

[CAT-THE-FIFTH]

Interesting thread on Reddit from AMD_Robert:

https://www.reddit.com/r/Amd/comments/4m692q/concerning_the_aots_image_quality_controversy/

Hi. Now that I'm off of my 10-hour airplane ride to Oz, and I have reliable internet, I can share some insight.

System specs:

CPU: i7 5930K
RAM: 32GB DDR4-2400Mhz
Motherboard: Asrock X99M Killer
GPU config 1: 2x Radeon RX 480 @ PCIE 3.0 x16 for each GPU
GPU config 2: Founders Edition GTX 1080
OS: Win 10 64bit
AMD Driver: 16.30-160525n-230356E
NV Driver: 368.19

In Game Settings for both configs: Crazy Settings | 1080P | 8x MSAA | VSYNC OFF

Ashes Game Version: v1.12.19928

Benchmark results:

2x Radeon RX 480 - 62.5 fps | Single Batch GPU Util: 51% | Med Batch GPU Util: 71.9 | Heavy Batch GPU Util: 92.3% GTX 1080 – 58.7 fps | Single Batch GPU Util: 98.7%| Med Batch GPU Util: 97.9% | Heavy Batch GPU Util: 98.7%

The elephant in the room:

Ashes uses procedural generation based on a randomized seed at launch. The benchmark does look slightly different every time it is run. But that, many have noted, does not fully explain the quality difference people noticed.

At present the GTX 1080 is incorrectly executing the terrain shaders responsible for populating the environment with the appropriate amount of snow. The GTX 1080 is doing less work to render AOTS than it otherwise would if the shader were being run properly. Snow is somewhat flat and boring in color compared to shiny rocks, which gives the illusion that less is being rendered, but this is an incorrect interpretation of how the terrain shaders are functioning in this title.

The content being rendered by the RX 480--the one with greater snow coverage in the side-by-side (the left in these images)--is the correct execution of the terrain shaders.

So, even with fudgy image quality on the GTX 1080 that could improve their performance a few percent, dual RX 480 still came out ahead.

As a parting note, I will mention we ran this test 10x prior to going on-stage to confirm the performance delta was accurate. Moving up to 1440p at the same settings maintains the same performance delta within +/-1%.

 

[Greebo]

It just comes across as all smoke and mirrors. I just wish they had shown a single card in that game (Yes, even that game) and that would have been a teaser enough. What they have done is left people scratching their heads over this. Why was only 50% GPU utilization being seen? Is CF scaling bad? etc.

I like simple things and they should have just shown a single card run. That would have sufficed for now and BS to upsetting the reviewers.

Its not crossfire though and that's why its only 51%. Crossfire has to be disabled for the dual cards to work in AOS. The AMD guy says this will improve but maybe not massively and its down to the Game developers, not AMD.

At least in medium and heavy it was almost 100% usage.

 

[TheRealDeal]

From the Reddit page, this is meant to be a single 480 result

http://www.ashesofthesingularity.co...-details/8b748568-fc96-4e48-9fed-22666a7149f5

Normal 45.5 Medium 41.8 Heavy 34.3

So on that s basis in AOS it beats the slowest Titanx on your list. Pretty impressive for a $200 card.

That result is using lower settings than in the bench thread. 4xaa v 8xaa and 8 million terrain shading samples v 16 million. Perhaps somebody could run these settings at stock clocks and see what they come up with.

 

[Mei]

i havent been keeping up has it been said what mhz these cards run at? the temps and power usage etc?

im guessing there will be a few cards with different price points?

 

[harney]

From the Reddit page, this is meant to be a single 480 result

http://www.ashesofthesingularity.co...-details/8b748568-fc96-4e48-9fed-22666a7149f5

Normal 45.5 Medium 41.8 Heavy 34.3

So on that s basis in AOS it beats the slowest Titanx on your list. Pretty impressive for a $200 card.

Very impressive if it is true ....Just so thank full we have AMD to keep prices in check .....Will be interesting what the exact UK price is after gouge, tax more gouge, and then more tax

 

[carrot1401]

This could be a silly question, but I remember crossfire & freesync not working together months ago. Did this get fixed in the end? Currently running a 290x and would be interested in grabbing a couple of 480's to go with my 1440p 144hz freesync screen (the Asus one!)

 

[ExoMale]

Because there are still no single cards capable of guaranteeing a minimum 60fps @ 4k resolution in all current games, never mind those coming over the next 12-18mths.

this is what you get for bringing out advance hardware ahead of time, companies push the boat out with technology in certain things like the televisions and monitors and nobody is able to catch up quick enough to properly utilize that new technology and its because of those last 2 words people have jumped on the band wagon which has resulted in some complaints because they expect everything to be at the same levels.

4k is nice, but no point if games cant run it properly and also no point if everything is recorded in lower quality for sky and the likes.

 

[Jono8]

So from that 1.51 scaling remark we know then that a 1080 is ~ 50% faster than a 480 meaning a 1070 will be about 25% faster and that is best case for AMD considering that is in AOTS.

Sounds like it is around 390x/980 performance then.

 

[HEADRAT]

I wouldn't touch CF with a barge pole, I will wait for Vega before upgrading my Fury X, life is too short to put up with the angst which is CF!

 

[fs123]

From the Reddit page, this is meant to be a single 480 result

http://www.ashesofthesingularity.co...-details/8b748568-fc96-4e48-9fed-22666a7149f5

Normal 45.5 Medium 41.8 Heavy 34.3

So on that s basis in AOS it beats the slowest Titanx on your list. Pretty impressive for a $200 card.

The card is looking better and better. If appears to actually match or beat a TX in AoTS. Part of that may be due to AMD being good with DX12 but I can't wait to see the DX11 results.
With GCN 4 updates, the new cards are expected to remove driver overhead and bottlenecks in DX11 so should run close to their theoretical capability.Potentially major boosts in DX11 games meaning the 480 will soundly beat the 390X in older games too.

 

[flopper]

The card is looking better and better. If appears to actually match or beat a TX in AoTS. Part of that may be due to AMD being good with DX12 but I can't wait to see the DX11 results.
With GCN 4 updates, the new cards are expected to remove driver overhead and bottlenecks in DX11 so should run close to their theoretical capability.Potentially major boosts in DX11 games meaning the 480 will soundly beat the 390X in older games too.

Total DominATIon with the 480 series.

 

[Toss3]

So from that 1.51 scaling remark we know then that a 1080 is ~ 50% faster than a 480 meaning a 1070 will be about 25% faster and that is best case for AMD considering that is in AOTS.

Sounds like it is around 390x/980 performance then.

Umm no one mentioned scaling, but utilization, which is something completely different. Both cards at 52% utilization could basically still be 1.5 times as fast as a single GPU, or it could be much lower (say 1.2 times). We simply don't know. So there's not really anything to derive from that number, other than that AMD and Oxide need to work on their software to utilize the hardware better.

 

[Toss3]

The card is looking better and better. If appears to actually match or beat a TX in AoTS. Part of that may be due to AMD being good with DX12 but I can't wait to see the DX11 results.
With GCN 4 updates, the new cards are expected to remove driver overhead and bottlenecks in DX11 so should run close to their theoretical capability.Potentially major boosts in DX11 games meaning the 480 will soundly beat the 390X in older games too.

I get that these cards will perform well under DX12, but saying they will be any better than the last generation in DX11 titles is a bit of a stretch. Call me skeptical, but I seriously doubt AMD has put any effort into improving DX11 performance, when every big title coming out this year is going to be utilizing DX12 or Vulkan.

 

[Jono8]

Umm no one mentioned scaling, but utilization, which is something completely different. Both cards at 52% utilization could basically still be 1.5 times as fast as a single GPU, or it could be much lower (say 1.2 times). We simply don't know. So there's not really anything to derive from that number, other than that AMD and Oxide need to work on their software to utilize the hardware better.

Read the thread. Amd bloke on reddit confirmed it meant 151% scaling.