Vega/Navi Rumors (Updated)

[Crumpet]

An end of July launch means the UK won't see the cards for another few weeks, it could be August into September before I get a card.

At which point, we're only a few months off the next generation Nvidia stuff.

All the while i'm sitting here with a dying R9 290 that's degrading all the time, and a freesync monitor I bought in preparation for AMD's new flagship GPU's and i'm starting to lose faith.

 

[Qwertilot]

Their graphics demos have been underwhelming since the Polaris E3 2016 stuff. I say just do it like NVIDIA - anounce a presentation date, walk up to the stage and launch the card/announce launch dates. No need for demos and teasers.

I think the problem here is that they haven't the resources to deliver consistently. With NV recently you've basically been able to set your calendar by their GFX card releases. So they don't need to trail things much as people know something new is coming.

AMD are much more erratic recently of course, so......

 

[Malogeek]

I'm assuming they showed Prey due to the marketing with Bethesda but agree that using CF certainly wasn't the ideal choice, even if they wanted to use it to tout the PCIe lanes in TR.

 

[OatisCampbell]

I had no idea when I skipped fury x I'd have to wait this long.
At least I know next time around amd isn't above shady tactics like a frontier edition to pretend as if they hit a deadline.

I'm holding my disappointment for the performance until we see the gpu at launch. This is amd... They've still got so much driver develop left to do.

Since we'll see the frontier edition soon though we'll know how good Vega is.

Edit : if Vega wasn't competitive why even bother? They would have scrapped it in the planning stage. It has to scale past Polaris better than simply expanding Polaris....

You do know the designs take years to bring to market and they have multiple design paths in play correct?

If Vega is the best of their designs they can bring to market, they "have" to bring it to market. It's been two years since their last high end launch in a few weeks, they aren't selling that any longer.

NVIDIA is selling a lot of 1070s and 1080s, AMD needs a piece of that market. Not to mention that level of performance is enough for the vast majority of gamers. The resolution poll here shows only 13.8% of AT folks "need" 1080Ti or higher performance:

http://www.portvapes.co.uk/?id=Latest-exam-1Z0-876-Dumps&exid=thread...-mid-2017-thread.2507267/page-3#post-38915177

A 1070 does very well at 3440 X 1440 and anything resolution below. When you throw in Freesync monitors being a couple hundred cheaper than GSync and making drops below 60fps largely unnoticeable I can see several reasons why AMD would "bother".

 

[tential]

You do know the designs take years to bring to market and they have multiple design paths in play correct?

If Vega is the best of their designs they can bring to market, they "have" to bring it to market. It's been two years since their last high end launch in a few weeks, they aren't selling that any longer.

NVIDIA is selling a lot of 1070s and 1080s, AMD needs a piece of that market. Not to mention that level of performance is enough for the vast majority of gamers. The resolution poll here shows only 13.8% of AT folks "need" 1080Ti or higher performance:

http://www.portvapes.co.uk/?id=Latest-exam-1Z0-876-Dumps&exid=thread...-mid-2017-thread.2507267/page-3#post-38915177

A 1070 does very well at 3440 X 1440 and anything resolution below. When you throw in Freesync monitors being a couple hundred cheaper than GSync and making drops below 60fps largely unnoticeable I can see several reasons why AMD would "bother".

Your post doesn't remotely address what I'm talking about.
I said amd has no reason to make a design that scales worse than Polaris and you went on that off topic rant to educate no one.

 

[Gikaseixas]

Assumptions and more assumption... you guys need to be a bit more patient as AMD is clearly struggling to coordinate all these launches. Please remember that it is not easy to do it with a small budget.

In the end, if you can't wait just get a 1080 or 1080TI now, otherwise just wait and hope for the best

 

[Glo.]

https://twitter.com/GFXChipTweeter/status/869895814545514497

 

[IEC]

https://twitter.com/GFXChipTweeter/status/869895814545514497

Someone on reddit did an analysis. 2.7 tears in a frame = 162fps (60x2.7) or 388fps on 144hz or 81 on 30hz

First would be more in the realm of what's expected
Second would be nuts
Third would be mediocre

Nothing conclusive can be drawn without an FPS counter.

 

[Malogeek]

Yeah what was the stream output?

That's sampling 4 frames in a single refresh so at assumed 30Hz stream and looks like there's AA being applied, that's quite impressive.

 

[Det0x]

That's probably a 30Hz stream sampling and 120fps render without vsync. If it was a 60Hz stream then it's even more impressive.

*edited*

 

[Malogeek]

That's probably a 30Hz stream sampling and 120fps render without vsync.

Which is at minimum equal to a 1080ti for a single card.

 

[Despoiler]

I honestly don't understand what RTG is doing. Have they ever lead with their professional products? If they want to change up their product stack launch order then fine. However, launching Vega in late July, early August puts them really close to where Nvidia could launch Volta consumer variants. They would have 3-6 months of product until Nvidia launches. We saw Vega running faster than 1080 levels way back last year on debug level drivers. In hindsight they should have launched professional, consumer mid-high, and consumer midrange as soon as possible. Forget the highend. Get it later. That's what Nvidia does so you know that works. IMO the best strategy is to get whatever cards they can in the hands of consumers for the product cycles they are designed for. Time in the product cycle matters. Your name being in the game matters. It's not like ATI/AMD has ever ruled the highend roost for long anyways. Heck no one is really expecting them to take the crown outright. Holding the crown for what could be all of 3 months is going to be costly overall. They are chasing unicorns and missing the pot of gold they ran past.

 

[Malogeek]

One more thing to add from the all too brief demo. If it is indeed around 120fps then we're either looking at close to CF 100% scaling or Vega is even faster than previously thought.

 

[EXCellR8]

The amount of speculation surrounding the latest cards is staggering. I feel like until we see actual unboxing everything posted online is more or less forfeit. If the RX cards are late July, then it will probably be mid August by the time some of the better OC cards are available. I used to only buy reference models but after the R9 series and then the first gen RX cards, I'm paying much closer attention to the post-launch offerings.

 

[Kenmitch]

You guys is silly. AMD has been treading water for over a decade trying to keep alive and supply you with at least a alternative to the other guys. Currently they are in a position to achieve the goal and all you guys think about is yourself. If you were running AMD today most likely you'd be concentrating on the markets they look to be going after 1st. There is money in the dGPU arena but most of the BIG money these days is in the other PC markets. I'm too scratching my head about they way they're being silent, tight lipped, etc but in the end it doesn't really matter what I think or expect. Mobile chips do look somewhat interesting.

 

[zinfamous]

I honestly don't understand what RTG is doing. Have they ever lead with their professional products? If they want to change up their product stack launch order then fine. However, launching Vega in late July, early August puts them really close to where Nvidia could launch Volta consumer variants. They would have 3-6 months of product until Nvidia launches. We saw Vega running faster than 1080 levels way back last year on debug level drivers. In hindsight they should have launched professional, consumer mid-high, and consumer midrange as soon as possible. Forget the highend. Get it later. That's what Nvidia does so you know that works. IMO the best strategy is to get whatever cards they can in the hands of consumers for the product cycles they are designed for. Time in the product cycle matters. Your name being in the game matters. It's not like ATI/AMD has ever ruled the highend roost for long anyways. Heck no one is really expecting them to take the crown outright. Holding the crown for what could be all of 3 months is going to be costly overall. They are chasing unicorns and missing the pot of gold they ran past.

I think all of this is easier said than done when your group is working on something like 1/5th to 1/9th the operating budget of your competitor.

 

[T1beriu]

I'm too scratching my head about they way they're being silent, tight lipped

If they would have announced TR at Computex, not 2 weeks ago, or if they put a stop to the leaks, we wouldn't have seen the competition, in this case Intel, scramble around frantically to change their HEDT plans by bringing more than 12 cores to the platform with massive price decreases. Everything we see and old leaks point out that Intel responded to TR.

That's why sometimes is better to be tight lipped. You bring a bigger thunder and you take the market and competition by surprise, that translates into massive media coverage, good reception and profits.

 

[tamz_msc]

The PCGH.de coverage has some interesting tidbits about Vega 10 being the only GPU that is going to be released at SIGGRAPH, with Vega 11 coming late 2017 or even early 2018. Reference designs are expected in the beginning.

 

[Borealis7]

AMD pulled a "prepare to prepare to keep waiting". i was on the fence, now F them, i'm getting a Pascal.
Vega in stores around September (where i live...), and Volta a quarter or 2 later. that is too late for me.

 

[Valantar]

I think all of this is easier said than done when your group is working on something like 1/5th to 1/9th the operating budget of your competitor.

Considering that AMD's Market cap is roughly 1/9 of Nvidia, I think you're overestimating this. Remember, GPUs are the core of everything Nvidia does these days. Anything not GPU-driven (Denver ARM cores, for example) has been put aside/canceled. All their AI stuff, all their car stuff, it's all based on GPUs. All the while, AMD produces high end consumer, workstation and server CPUs too. I'd bet the RTG R&D budget is closer to 1/20 of that of Nvidia's GPU arch group.

 

[Despoiler]

I think all of this is easier said than done when your group is working on something like 1/5th to 1/9th the operating budget of your competitor.

I totally get that, but they can relatively easily build a pro card and then do consumer mid-high to low end with the same chip. The thing that is constantly just out of their reach is the very top enthusiast card. That's the one part that screws their execution. That being the case I'm making the case that they just don't allow chasing a top card to cause a delay to products you could be selling competitively. They literally cannot afford to do what they are doing.

 

[T1beriu]

The PCGH.de coverage has some interesting tidbits about Vega 10 being the only GPU that is going to be released at SIGGRAPH, with Vega 11 coming late 2017 or even early 2018.

It's just an inference, not a leak, the way I read it.

Source, btw.

 

[Mopetar]

At this point AMD really needs to beat the Ti out of the gate with RX Vega, just to make it worth the wait. The only other way it isn't a disappointment is if the prices are really aggressive.

 

[tamz_msc]

It's just an inference, not a leak, the way I read it.

Source, btw.

A German user on reddit said that the equivalent word for 'confirmed' was used, under behind the scenes talk.

 

[crisium]

if Vega wasn't competitive why even bother? They would have scrapped it in the planning stage. It has to scale past Polaris better than simply expanding Polaris....

If it's lagging this badly, then either there are hardware bottlenecks (most likely the fact that the front end is still as starved as Fiji, with only 4 shader engines we've seen on the Linux drivers) or else the drivers suck.

Vega was supposed to fix the 4 shader engines limit, but it doesn't look like that actually happened. If the product is really subject to the same kind of bottlenecks that held Fiji back, despite 2+ years to fix them, then Raja should be fired for incompetence and so should the design team.

I am biting my tongue since the Vega book has yet to be written. But you two are exactly right. AMD needs to move forward in a deliberate fashion. This card needs to have better scaling than Polaris. And that means it must fix the bottlenecks of Fiji.

Sometimes I think they should just port Polaris to the Hawaii config, clearly the best design that doesn't get bottlenecked anywhere. 44CU:64ROP and 512-bit with Polaris +~20% "IPC" and clock frequency for a GP104 competitor anyone?

While I'm backseat engineering, they should port the 40CU 384-bit Project Scorpio to a dGPU so that AMD can finally beat the GTX 1060 decisively. RX 590?

I still want to believe that Vega can compete with the GP102, which first launched in the 2016 Titan X in early August. Almost a full year before Vega. Hope...