RTG3rd March – Will Discuss Polaris, Fury X2, VR, DirectX 12 and More

[therealnickdanger]

I'm kind of amazed that rebranding worked as well as it did.

"I would never buy a Toyota, I'm a Lexus kind of person."

 

[railven]

The funniest thing is that the 290(X) also seemed to have its image improve with the 390(X) launch. Having a big public bunch of reviews that start with saying that this 390 is basically the same as the 290 followed by a heaping of benchmarks showing the same sort of performance that had made the 290 the bargain of the half decade did a lot for its image.

I'm kind of amazed that rebranding worked as well as it did.

Considering the arguments had over "rebrand/rebadge" and "refresh" it wasn't simply like they changed the name.

The 390/390X are far better products day 1 than the 290/290X were, and a good portion of that is solely on the coolers. The added perf is of course excellent. Look at poor Nano/Fury X getting slapped with coil/radiator whine issues out the gate.

These little nuances can make or break a product.

 

[LTC8K6]

"I would never buy a Toyota, I'm a Lexus kind of person."

Nissan / Infiniti - didn't really work all that well...

 

[therealnickdanger]

Looks like FuryX2 talk is off limits.

 

[BaronVonMehl]

This whole AMA is really a bummer after all this hype ....

 

[swilli89]

This whole AMA is really a bummer after all this hype ....

Mainly with the exception of some great in-depth explanation of what's possibly in DX12 vs DX11 and other API related stuff.

But yeah.. disappointed we got nothing new in terms of hardware news.

 

[caswow]

told you so

 

[tential]

told you so

What were you expecting tons of new info?

Expect some PR answers, and maybe some small new tidbits. AMD has awhile to go before this all launches, they're playing the marketing game and just trying to stay in the news up to Polaris launch.

 

[therealnickdanger]

Sounds like GDC (March 14) is when they'll drop some more interesting news.

 

[Face2Face]

This whole AMA is really a bummer after all this hype ....

To be honest, they always are.

AMD : Okay guys, we're going to be answering some of those burning questions of yours...

<Insert Polaris or Zen question here>

AMD : Sorry guys, I can't answer questions about Polaris or Zen, you know the only two things anyone cares about, I like keeping my job Toodles

 

[poofyhairguy]

To be honest, they always are.

AMD : Okay guys, we're going to be answering some of those burning questions of yours...

<Insert Polaris or Zen question here>

AMD : Sorry guys, I can't answer questions about Polaris or Zen, you know the only two things anyone cares about, I like keeping my job Toodles

I thought the part about Trueaudio was interesting.

 

[Vesku]

Did confirm that the upcoming GPUs are 14nm GF and will be launching "mid-year" which should mean sometime this Summer.

Edit:

Polaris will have DP1.3 interface

"DirectFlip support shortly" for AMD drivers and the horror show that is Windows 10 store games limited capabilities does not apply to all DX12 games.

Polaris - "h.265 encode and decode up to 4K"

No planned official support for GCN1.0 in AMDGPU driver (linux).

Even their own PR focused people run Intel. (I7-6700K + Fury X) ;p

HDR displays expected to hit retail in 2H this year.

Regarding Arctic Islands and relation to Polaris: "I don't know how to answer this. Polaris 10 is Polaris 10, and Polaris 11 is Polaris 11. There are no other names."

"25 million people actively use the Gaming Evolved application every day. Not just installs, but active users."

Regarding Virtual Super Resolution: "We have hardware that performs real-time frame resizing with 0% performance impact above and beyond the change in resolution selected by the user. This hardware became more advanced as we iterated GCN.
It is possible to explore shader-based methods, but that can introduce specific performance penalties of their own.
We will continue to look at adding additional resolution options to VSR."

 

[Azix]

https://www.reddit.com/r/Amd/commen...ologies_group_qa_is_happening_here_on/d0mawke

sounds like its all 14nm global foundries.

 

[Piroko]

Regarding DX11 multithreading:

1) Because DCLs are useless. They've been inappropriately positioned as a panacea for DX11's modest multi-threading capabilities, but most journalists and users exploring the topic are not familiar with why DCLs are so broken or limited.
Let's say you have a bunch of command lists on each CPU core in DX11. You have no idea when each of these command lists will be submitted to the GPU (residency not yet known). But you need to patch each of these lists with GPU addresses before submitting them to the graphics card. So the one single CPU core in DX11 that's performing all of your immediate work with the GPU must stop what it's doing and spend time crawling through the DCLs on the other cores. It's a huge hit to performance after more than a few minutes of runtime, though DCLs are very lovely at arbitrarily boosting benchmark scores on tests that run for ~30 seconds.

That last part was new to me

Regarding drivers:

Sort of. We're clearly interested in doing a major feature-rich driver every year, and that is on-track for 2016 as well.

So the Radeon Settings / CCC dichtomy is there to stay for a while longer, ah well.

 

[Silverforce11]

In regards to TressFX (and PureHair in Rise of Tomb Raider, ~10% perf hit on all GPUs).

NVIDIA had success with TressFX because we designed the effect to run well on any GPU. It's really that simple. They were successful because we let them be. We believe that's how it should be done for gamers: improve performance for yourself, don't cripple performance for the other guy. The lesson we learned is that actual customers see value in that approach.

I wonder if they see enough value to buy an AMD GPU over NV?

 

[Bacon1]

Regarding DX11 multithreading:
That last part was new to me

Regarding drivers:
So the Radeon Settings / CCC dichtomy is there to stay for a while longer, ah well.

They do have multithreaded DLCs already as well:

Using the DX11 Multithread sample:

    DEVICECONTEXT_IMMEDIATE,                // Traditional rendering, one thread, immediate device context
    DEVICECONTEXT_ST_DEFERRED_PER_SCENE,    // One thread, multiple deferred device contexts, one per scene 
    DEVICECONTEXT_MT_DEFERRED_PER_SCENE,    // Multiple threads, one per scene, each with one deferred device context
    DEVICECONTEXT_ST_DEFERRED_PER_CHUNK,    // One thread, multiple deferred device contexts, one per physical processor 
    DEVICECONTEXT_MT_DEFERRED_PER_CHUNK,    // Multiple threads, one per physical processor, each with one deferred device context

Immediate -> 8-9fps
ST Def / Scene -> 8.5-9.5 fps
MT Def / Scene -> 23-24fps
ST Def / Chunk -> 8-9 fps
MT Def / Chunk -> 19-20fps

https://code.msdn.microsoft.com/Direct3D-Multithreaded-d02193c0

Using a 290

But yes its other engine limitations + driver overhead that cause it to be slow on DX11.

 

[crisium]

In regards to TressFX (and PureHair in Rise of Tomb Raider, ~10% perf hit on all GPUs).



I wonder if they see enough value to buy an AMD GPU over NV?

I've seen people on NeoGAF state they buy Nvidia specifically because Nvidia sabotages AMD performance and they want full compatibility. Never underestimate the consumer's willingness to side with unethical practices.

 

[Silverforce11]

I've seen people on NeoGAF state they buy Nvidia specifically because Nvidia sabotages AMD performance and they want full compatibility. Never underestimate the consumer's willingness to side with unethical practices.

Exactly. Lots of folks don't give a damn and just buy whatever that works best without issues. If AMD gets issues with GameWorks (Gears of Wars just now), it looks bad for them. They aren't going to feel pity and buy AMD just because! lol

We suspected Gemini (Fury X2) would be link to VR, as their push for 1 GPU per eye to minimize latency was often the selling point in LiquidVR.

The product schedule for Fiji Gemini had initially been aligned with consumer HMD availability, which had been scheduled for Q415 back in June. Due to some delays in overall VR ecosystem readiness, HMDs are now expected to be available to consumers by early Q216. To ensure the optimal VR experience, we’re adjusting the Fiji Gemini launch schedule to better align with the market.

Working samples of Fiji Gemini have shipped to a variety of B2B customers in Q415, and initial customer reaction has been very positive.

 

[Elixer]

Well... can't say I didn't try.
And I got a nice [REMOVED].

All I was asking about is Vulkan integration, don't know why the mods got their panties in a bind over that.
Perhaps it was a too technical of a question.

 

[MrTeal]

In regards to TressFX (and PureHair in Rise of Tomb Raider, ~10% perf hit on all GPUs).



I wonder if they see enough value to buy an AMD GPU over NV?

Barring other incentives, I'll probably be buying an nVidia GPU despite my pretty much exclusive track record of AMD GPUs since Geforce 2. I've spent close to 5 figures on AMD GPUs over the years (obviously not just for gaming), but if for gaming an nVidia GPU gives 10 % lower FPS on average at the same price point but doesn't suffer random -50% GW related drops, that's a pretty compelling argument.

 

[Vesku]

Barring other incentives, I'll probably be buying an nVidia GPU despite my pretty much exclusive track record of AMD GPUs since Geforce 2. I've spent close to 5 figures on AMD GPUs over the years (obviously not just for gaming), but if for gaming an nVidia GPU gives 10 % lower FPS on average at the same price point but doesn't suffer random -50% GW related drops, that's a pretty compelling argument.

Personally, I don't pay ransoms. Generally AMD manages to work around nVidia sabotage within a few weeks of a game launch and the launch FPS isn't unplayable just makes for great benchmark charts for nVidia marketing.

 

[jackstar7]

I'm really hoping they put out a compelling product in this next release. I'd like to swing back to AMD and then have Freesync options to explore (or that Dell OLED) with DP 1.3 hitting the scene.

 

[coercitiv]

if for gaming an nVidia GPU gives 10 % lower FPS on average at the same price point but doesn't suffer random -50% GW related drops, that's a pretty compelling argument.

What do you reckon your upgrade cycle will be, as you switch over? To put things into perspective, would you be using Kepler cards right now?

It seems to me that both brands have good value propositions for their customers, but that value implies certain consuming habits in order to shine.

 

[airfathaaaaa]

Barring other incentives, I'll probably be buying an nVidia GPU despite my pretty much exclusive track record of AMD GPUs since Geforce 2. I've spent close to 5 figures on AMD GPUs over the years (obviously not just for gaming), but if for gaming an nVidia GPU gives 10 % lower FPS on average at the same price point but doesn't suffer random -50% GW related drops, that's a pretty compelling argument.

we havent seen anything related to gw on dx12 and if a small example of what is coming on vulkan is the locked down demo of nvidia well things are going to be interesting:sneaky:

 

[prtskg]

No planned official support for GCN1.0 in AMDGPU driver (linux).

Open source work for it has already started.
http://www.phoronix.com/forums/foru...t-to-older-amd-gcn-gpus/page9?_=1457077650155