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

[PontiacGTX]

When AMD released the Radeon RX 480, the first performance tests they officially showcased to the audience were from Ashes of the Singularity. AMD compared performance of two Radeon RX 480 cards against a single GeForce GTX 1080. Based on performance and price, the RX 480 Crossfire solution edged the GeForce card. The reason to choose AOTS for comparison was that the game fully utilized Asynchronous Compute. The new tech allows faster graphics performance for cards that support the new tech in DirectX 12 mode.

AMD 687F:C1 Graphics Product Is On Par With GeForce GTX 1080 – Possibly Vega GPU Based

Now there have been some rumors that this card could be a dual chip product but from the benchmarks, it seems like we are only looking at a single-chip solution. The card is listed with the device ID 687F:C1. AMD GPU database doesn’t list any such card and was the same case for the RX 480 prior to its release. So this could be an entirely new GPU which we haven’t seen or heard of before. Now this doesn’t mean that there won’t be any dual chip card. The possibility for that product also remain but I personally don’t think AMD will want to develop a dual chip solution based on Polaris 10. Their dual chip products are based only on high-end chips such as Fiji (Radeon Pro Duo) and Hawaii (Radeon R9 295X2).

[/URL]

AMD 687F:C1 DirectX 11 / DX12 Tests (1080P/1440P):
There are four benchmarks that were conducted with DirectX 11 API mode. These consist of Crazy, Standard and Low settings. The only score that has actually been compared to other GPUs was the Standard 1080P DX 12 test. This shows the card scoring 8400 points which is in the range of what GeForce GTX 1080 scores. The score was listed at 131th position before it was deleted. The fact that it was deleted could indicate that it was not supposed to be shown to public.

Spoiler: DX11

Spoiler: DX12

Source

 

[hawtdawg]

This is exactly what was expected was it not? AMD releases Vega 10 with a 250w TDP and HBM2, just to compete with Nvidia's small GPU.

The sad part is that the HBM2 is probably there just to keep TDP down as much as possible.

 

[PhonakV30]

so according to image , there is no mGPU then It's single GPU and Device ID is 687F:C1

RX 480 ==> 67DF:C7 ( Link )
RX 470 ==> 67DF:C4

so I think It's Little Vega.no way Full Vega when we have Titan Pascal.

 

[richaron]

so according to image , there is no mGPU then It's single GPU and Device ID is 687F:C1

I'm not familiar with AOTS or AMD's device ID's, how do we know this isn't a dual Polaris card?

 

[hawtdawg]

so I think It's Little Vega.no way Full Vega when we have Titan Pascal.

Unless Vega has a drastic efficiency improvement, they will not be able to build a GP102 competitor with it. They'd be looking at a 350+ watt TDP.

 

[Pariah]

I'm not familiar with AOTS or AMD's device ID's, how do we know this isn't a dual Polaris card?

The AOTS benchmark detects if the card is mGPU and lists it in the results.

Notice the card above the highlighted one. The highlighted card is supposedly a new AMD GPU.

 

[SlickR12345]

There are two Vega parts, the "big" Vega that is going to compete with Titan X2 and the smaller Vega that is going to compete with the 1070/1080. Sure they are late to the party, but with aggressive pricing, good reference design to SELL the card on first impressions it would shake up the market. Also with Polaris refresh somewhere around April they would have a great line of GPU's.

I mean RX 470 is UNMATCHED at $170, best price for performance. A strong refresh will devastate Nvidia!

 

[richaron]

*Snip*

Ah, thankyou.

I do hope the software dynamically tests for mGPU rather than relying on recognizing an ID string, since I am quite interested to see how AMD is doing with Vega.

 

[Pariah]

I mean RX 470 is UNMATCHED at $170, best price for performance. A strong refresh will devastate Nvidia!

You can't possibly believe that. Nvidia's market position is so dominant at this point that there is nothing AMD can realistically release that would "devastate" Nvidia. If AMD released this unknown card for $100, sure that would shake up the market, but that isn't happening. The best we can hope for is something from AMD that offers better price/performance than what NVidia currently has.

If the above card is small vega, then that is promising, because without comparable performance to Nvidia, nothing else would matter. One major problem is how late this product is coming out. By the time it hits the market Nvidia will likely be capable of releasing a Pascal refresh that will raise the performance bar well above the 1070/1080 resulting in no real shift in price performance unless AMD sharply undercuts Nvidia's current pricing which would have a very negative impact on AMD's bottom line especially if this card uses HBM2 which is undoubtedly much more expensive than the GDDR5X Nvidia is using in the 1080.

 

[SlickR12345]

You can't possibly believe that. Nvidia's market position is so dominant at this point that there is nothing AMD can realistically release that would "devastate" Nvidia. If AMD released this unknown card for $100, sure that would shake up the market, but that isn't happening. The best we can hope for is something from AMD that offers better price/performance than what NVidia currently has.

If the above card is small vega, then that is promising, because without comparable performance to Nvidia, nothing else would matter. One major problem is how late this product is coming out. By the time it hits the market Nvidia will likely be capable of releasing a Pascal refresh that will raise the performance bar well above the 1070/1080 resulting in no real shift in price performance unless AMD sharply undercuts Nvidia's current pricing which would have a very negative impact on AMD's bottom line especially if this card uses HBM2 which is undoubtedly much more expensive than the GDDR5X Nvidia is using in the 1080.

Nvidia played their hand, the refresh is not going to be a massive improvement in performance, for the 1000 series they went from a 5yo 28nm to a 16nm process, now they can just make smaller improvements to their core, maybe 5% gain from that and another 5-7% gain from overclocking and that is going to be it. There is nothing that Nvidia can release in 2017 that is going to be significantly faster than their Titan X2, which is already a super niche card at $1200.

AMD on the other hand still have all their cards and could play the four aces and go for the jugular and destroy Nvidia. 8GB of HBM2 memory, 4000+ shaders with significantly improved efficiency over the Fury X, lessons learned from the Polaris cards and improvements implemented and of course a more mature 14nm process node!

If you add an aggressive pricing on top of that and a great reference design to make a great first impression and you've got a winner. For the big Vega making it an exclusive card at say $1000 with Titan X2 performance levels, but with water cooling and HBM2, later on allowing custom made versions with air coolers for more market penetration, that would be a giant win for them!

 

[Pariah]

Nvidia played their hand, the refresh is not going to be a massive improvement in performance, for the 1000 series they went from a 5yo 28nm to a 16nm process, now they can just make smaller improvements to their core, maybe 5% gain from that and another 5-7% gain from overclocking and that is going to be it. There is nothing that Nvidia can release in 2017 that is going to be significantly faster than their Titan X2, which is already a super niche card at $1200.

Titan XP is a cut down gp102 471mm die with gddr5x. Full die has 30 SM's. Nvidia also sells a gp100 610mm die with 16GB HBM2 and a 250W TDP. Full die has 60 SM's. Previous generation Titan XP utilized a 601mm die. You really believe AMD is going to announce a card that is twice as fast as the current Titan XP in the next 6 months? Because that's what Nvidia is already producing. Certainly they don't want to be forced to sell them to gamers at much lower margins than Tesla's. But if they are forced to, they have much bigger guns in reserve than the Titan XP.

Even just using a full gp102 die plus HBM2 along with any tweaks Nvidia has made over the last year will result in a performance gain of well over 5-7%. Nvidia has played their hand? Please at least attempt to argue with facts.

 

[RussianSensation]

You can't possibly believe that. Nvidia's market position is so dominant at this point that there is nothing AMD can realistically release that would "devastate" Nvidia.

He phrased it too dramatically. We already know that NV loyalists as they will never buy AMD. How many times does this have to be repeated?

ATi wiped the floor with NV with 9700/9800Pro, X1900/1950XTX in every key metric under the sun that matters, from price/performance to overall IQ to absolute performance. NV had inferior 2D and 3D IQ for generations, including completely broken Full RGB over HDMI for 1 whole decade. I am not going to link 100 articles on how NV had horrendous MSAA for generations and it took at least 1 full level up in MSAA to match ATI. ATI also had a 6-9 months roll-out lead with HD5000 series (NV had 0 response to HD5750/5770/5850/5870 for close 6-9 months depending on the price segment and yet the GTX500 still outsold HD5000 no the desktop! *facepalm*). Despite all this, it barely reached 44.5% desktop dGPU market share during that era:

What matters is the remaining AMD users and objective gamers who have not upgraded to Pascal and are open minded to the best price/performance cards of 2017. Certainty, there are millions of pre-Pascal gamers who are going to be purchasing a new GPU in 2017.

If AMD releases substantially better cards than my GTX1070 SLI, for a great price, I'll be dumping my Pascal cards.

Right now the cheapest open-air cooled AIB 1080 is $599.99 on Newegg. If RX 490 comes in 10% faster than GTX1070 and costs $399-449, it'll be a great deal. If NV releases refreshes, there will be price wars = consumers win! Great outcome again.

There are two Vega parts, the "big" Vega that is going to compete with Titan X2 and the smaller Vega that is going to compete with the 1070/1080. Sure they are late to the party, but with aggressive pricing, good reference design to SELL the card on first impressions it would shake up the market. Also with Polaris refresh somewhere around April they would have a great line of GPU's.

I mean RX 470 is UNMATCHED at $170, best price for performance. A strong refresh will devastate Nvidia!

Exactly. MSI RX 470 for $150 and PowerColor RX 470 for $150. Despite these cards leveling GTX1050Ti by 25-35% depending on the review, 1050Ti turd will outsell the far superior RX 470 by 4-5:1. What else is new? NV's GeForce 5, 7 and Kepler say hello -- you can sell garbage with great marketing as long as it's attached to a trusted brand and you have flooded every retailer and online wholesaler

R9 285/380/380X class => RX 470/480 replacements
R9 390/390X class => RX 490 replacement with 512GB/sec HBM2

Rumour:

Fury/Fury X class => We don't know when this GPU will release in 2017.

 

[PontiacGTX]

so according to image , there is no mGPU then It's single GPU and Device ID is 687F:C1

RX 480 ==> 67DF:C7 ( Link )
RX 470 ==> 67DF:C4

so I think It's Little Vega.no way Full Vega when we have Titan Pascal.

if this has a cut down model at 370-350usd with HBM(2) it would be interesting

 

[casiofx]

He phrased it too dramatically. We already know that NV loyalists as they will never buy AMD. How many times does this have to be repeated?

ATi wiped the floor with NV with 9700/9800Pro, X1900/1950XTX in every key metric under the sun that matters, from price/performance to overall IQ to absolute performance. NV had inferior 2D and 3D IQ for generations, including completely broken Full RGB over HDMI for 1 whole decade. I am not going to link 100 articles on how NV had horrendous MSAA for generations and it took at least 1 full level up in MSAA to match ATI. ATI also had a 6-9 months roll-out lead with HD5000 series (NV had 0 response to HD5750/5770/5850/5870 for close 6-9 months depending on the price segment and yet the GTX500 still outsold HD5000 no the desktop! *facepalm*). Despite all this, it barely reached 44.5% desktop dGPU market share during that era:

What matters is the remaining AMD users and objective gamers who have not upgraded to Pascal and are open minded to the best price/performance cards of 2017. Certainty, there are millions of pre-Pascal gamers who are going to be purchasing a new GPU in 2017.

If AMD releases substantially better cards than my GTX1070 SLI, for a great price, I'll be dumping my Pascal cards.

Right now the cheapest open-air cooled AIB 1080 is $599.99 on Newegg. If RX 490 comes in 10% faster than GTX1070 and costs $399-449, it'll be a great deal. If NV releases refreshes, there will be price wars = consumers win! Great outcome again.



Exactly. MSI RX 470 for $150 and PowerColor RX 470 for $150. Despite these cards leveling GTX1050Ti by 25-35% depending on the review, 1050Ti turd will outsell the far superior RX 470 by 4-5:1. What else is new? NV's GeForce 5, 7 and Kepler say hello -- you can sell garbage with great marketing as long as it's attached to a trusted brand and you have flooded every retailer and online wholesaler

R9 285/380/380X class => RX 470/480 replacements
R9 390/390X class => RX 490 replacement with 512GB/sec HBM2

Rumour:

Fury/Fury X class => We don't know when this GPU will release in 2017.

Depends on the market. RX470 4GB is $221 while 1050ti can be had for $160 here. Both prices including vat. Laid down the choices, my relative chose 1050ti. If RX470 was $150 i would definitely got them.

The fast sync feature of 10 series is pretty awesome though

 

[Glo.]

if this has a cut down model at 370-350usd with HBM(2) it would be interesting

No chances. If this GPU has GTX 1080 level of performance it will cost at least 499$, to match 499$ price tag of Zen CPU.

What I believe AMD is trying to do with Zen+Vega is provide ultimate gaming platform for under 1000$.

Normally you would have to pay 1089$ for 5960X, and 699$ for GTX 1080. That is the target for AMD, IMO.

Also, this is only my speculation. You do not have to agree with me on this.

 

[PontiacGTX]

No chances. If this GPU has GTX 1080 level of performance it will cost at least 499$, to match 499$ price tag of Zen CPU.

What I believe AMD is trying to do with Zen+Vega is provide ultimate gaming platform for under 1000$.

Normally you would have to pay 1089$ for 5960X, and 699$ for GTX 1080. That is the target for AMD, IMO.

Also, this is only my speculation. You do not have to agree with me on this.

Cutdown not this GPU

 

[Yakk]

I have to say I'm in for a Zen+Vega setup.

Actually I wouldn't be surprised to see AMD refreshes their whole product line including SSD & DRAM from the sell-off I'm seeing on those parts lately to empty the sales pipeline. Maybe we could see a complete AMD Zen+Vega+SSD+DRAM system soon.

 

[Glo.]

I have to say I'm in for a Zen+Vega setup.

Actually I wouldn't be surprised to see AMD refreshes their whole product line including SSD & DRAM from the sell-off I'm seeing on those parts lately to empty the sales pipeline. Maybe we could see a complete AMD Zen+Vega+SSD+DRAM system soon.

Project Quantum.

Cutdown not this GPU

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.

 

[hawtdawg]

There are two Vega parts, the "big" Vega that is going to compete with Titan X2 and the smaller Vega that is going to compete with the 1070/1080. Sure they are late to the party, but with aggressive pricing, good reference design to SELL the card on first impressions it would shake up the market. Also with Polaris refresh somewhere around April they would have a great line of GPU's.

I mean RX 470 is UNMATCHED at $170, best price for performance. A strong refresh will devastate Nvidia!

No, I think the small vega is going to compete with the 1070, and the big vega with the 1080. AMD doesn't have the TDP headroom to make a GP102 competitor.

 

[Glo.]

No, I think the small vega is going to compete with the 1070, and the big vega with the 1080. AMD doesn't have the TDP headroom to make a GP102 competitor.

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.

 

[Yakk]

Oh, and this is also a perfect occasion for AMD to introduce another Titan killer GPU to go along with the Zen. With Vega's developement schedule it seems about right on timing.

 

[Azix]

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

 

[Vesku]

HBM2, node improvements and design improvements should allow for a nice jump from a similar die size Polaris.

 

[Pariah]

He phrased it too dramatically. We already know that NV loyalists as they will never buy AMD. How many times does this have to be repeated?

I don't know. Why are you complaining about something you decided to bring up on your own?

ATi wiped the floor with NV with 9700/9800Pro, X1900/1950XTX in every key metric under the sun that matters, from price/performance to overall IQ to absolute performance. NV had inferior 2D and 3D IQ for generations, including completely broken Full RGB over HDMI for 1 whole decade. I am not going to link 100 articles on how NV had horrendous MSAA for generations and it took at least 1 full level up in MSAA to match ATI. ATI also had a 6-9 months roll-out lead with HD5000 series (NV had 0 response to HD5750/5770/5850/5870 for close 6-9 months depending on the price segment and yet the GTX500 still outsold HD5000 no the desktop! *facepalm*). Despite all this, it barely reached 44.5% desktop dGPU market share during that era:

I don't understand what you are responding to here. Doesn't seem to have anything to do with the RX 490 this thread is about or what market effect it could have.

What matters is the remaining AMD users and objective gamers who have not upgraded to Pascal and are open minded to the best price/performance cards of 2017. Certainty, there are millions of pre-Pascal gamers who are going to be purchasing a new GPU in 2017.

This is where I stopped reading your post. Objective gamers only buy AMD? Don't waste my time with this absurd generalization.

 

[linkgoron]

This is where I stopped reading your post. Objective gamers only buy AMD? Don't waste my time with this absurd generalization.

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.