It feels like the build-up to Vega has been just shy of forever. And in a way it has been: AMD first announced the architecture back in March of 2016, and we’ve known for almost as long that Polaris would be a volume play, not a play for the high performance market. It would be Vega that would be AMD’s next high-end GPU, and true successor to 2015’s unique (if mistimed) Fiji GPU.

Since then AMD has slowly been releasing details on the Vega architecture. Bits and pieces related to the Playstation 4 Pro, which incorporated some Vega features, and then a much more significant information dump at the start of this year with their Vega teaser, closely aligned with their associated announcement of the Radeon Instinct accelerator, which set some pretty specific guidelines for what kind of performance AMD was aiming for. Until today all we’ve known for sure is that Vega would be ambitious, but not unrealistic, and that AMD wanted to launch it by the end of June.

So for today’s AMD Financial Analyst Day, AMD has released a little bit more information as part of the next step of their campaign. The first Vega product to be released has a name, it has a design, and it has performance figures. Critically, it even has a release date. I hesitate to call this a full announcement in the typical sense – AMD is still holding some information back until closer to the launch – but we now finally have a clear picture of where the Vega generation kicks off for AMD.

Say hello to the Radeon Vega Frontier Edition.

AMD Workstation Card Specification Comparison
  Radeon Vega Frontier Edition Radeon Pro Duo (Polaris) Radeon Pro WX 7100 Radeon Fury X
Stream Processors 4096 2 x 2304 2304 4096
Texture Units ? 2 x 144 144 256
ROPs 64? 2 x 32 32 64
Boost Clock ~1.59GHz 1243MHz 1243MHz 1050MHz
Single Precision ~13 TFLOPS 11.5 TFLOPS 5.7 TFLOPS 8.6 TFLOPS
Half Precision ~26 TFLOPS 11.5 TFLOPS 5.7 TFLOPS 8.6 TFLOPS
Memory Clock ~1.88Gbps HBM2 7Gbps GDDR5 7Gbps GDDR5 1Gbps HBM
Memory Bus Width 2048-bit 2 x 256-bit 256-bit 4096-bit
Memory Bandwidth ~480GB/sec 2x 224GB/sec 224GB/sec 512GB/sec
VRAM 16GB 2 x 16GB 8GB 4GB
Typical Board Power ? 250W 130W 275W
GPU Vega (1) Polaris 10 Polaris 10 Fiji
Architecture Vega Polaris Polaris GCN 1.2
Manufacturing Process GloFo 14nm GloFo 14nm GloFo 14nm TSMC 28nm
Launch Date 06/2017 05/2017 10/2016 06/24/15
Launch Price (A profitable number) $999 $649 $649

How AMD is going to sell this card and its target market is a huge subject (and arguably the most important aspect of today’s announcement) but I think everyone would rather dive into the specs first. So let’s start there.

The Radeon Vega Frontier Edition is meant to be a powerhouse of a card for AMD. While we’re still not quite sure which Vega chip we’re looking at – AMD has confirmed two chips so far, and it now looks like the second chip is going to be farther down the line – the Vega that we are getting now is AMD’s high-end play. And this is reflected in the Vega Frontier Edition high performance specifications.

AMD previously indicated with the Radeon Instinct MI25 announcement that they would be targeting 25 TFLOPS or better of half precision (FP16) performance on their high-end Vega parts, and the Vega Frontier Edition will be delivering on the “or better” part of that, with performance in the ballpark of 26 TFLOPS. With the caveat that AMD isn’t providing precise numbers at this time – like the Polaris launch, they’re being a bit unspecific – this indicates that AMD is hitting their earlier performance goals.

Furthermore, as part of today’s announcement we finally know that the Vega Frontier Edition will ship with 64 CUs (4096 stream processors), which also allows us to narrow down the clockspeeds of the card. To hit 26 TFLOPS FP16, the card needs to be able to boost to around 1.59GHz. This is particularly notable because AMD’s reference Polaris designs only go to 1.34GHz, which means that AMD is expecting to clock Vega cards a good bit higher than their Polaris cards.

Hitting these high clockspeeds is important for AMD, because while Vega is slated to be a powerful chip, in terms of execution width it’s actually not an especially wide chip. With 64 CUs, the Vega FE is no wider than AMD’s last-generation Fiji-based cards like the Radeon Fury X, which had the same 64 CU configuration. This means that relative to its predecessor, Vega FE is going to rely on the combination of significantly higher clockspeeds and its architectural enhancements to deliver the big performance gains AMD is touting for the card.

Meanwhile, although we now have a good picture of the CU situation for Vega FE, we have a bit of a mystery on our hands with the ROP situation. AMD has not disclosed the number of ROPs, but it has disclosed that the card will have a pixel fillrate of about 90 Gigapixels/second, which doesn’t divide neatly into the earlier clockspeed. It’s quite likely that we’re looking at a 64 ROP design and that AMD’s figures are factoring in a more conservative clockspeed (e.g. the base clock rather than the boost clock). But for the moment, this dichotomy means that the ROP configuration of the Vega FE isn’t as clear as the CU configuration.

Moving on, AMD has confirmed that the Vega FE will have approximately 480GB/sec of memory bandwidth, and in a big surprise, that this will come from 16GB of HBM2 memory. AMD has been heavily promoting their use of HBM2 and what it means for their memory architecture – the High Bandwidth Cache Controller – particularly banking on the fact that the HBM2 specification allows for memory speeds of up to 2Gbps/pin. Coupled with the fact that we’ve already seen an early Vega chip with 2 HBM2 stacks (indicating a 2048-bit bus), and AMD’s figures confirm that they’re going to be coming very close to 2Gbps for the Vega FE. The card’s 480GB/sec of bandwidth means that the HBM2 stacks need to be clocked at 1.875GHz, just short of that 2Gbps figure.

However the far more interesting news here isn’t the bandwidth, but the amount of total HBM2 memory. AMD has revealed that the Vega FE will ship with 16GB of memory, a figure that until today didn’t look to be possible. And to be honest, we’re still not sure how AMD is doing it.

Along with the higher data rate of HBM2, the other major feature of the newer HBM specification is that it allows for much higher capacity stacks, up to 4GB for a 4 layer (4-Hi) stack, or 8GB for an 8 layer (8-Hi) stack, solving the capacity issues that dogged the HBM1-powered Fiji GPU. However to date, neither Samsung nor SK Hynix have offered 8-Hi HBM2 in their parts catalogs. No specific reason has been given, but the use of TSVs means that the greater the number of layers, the harder the memory is to assemble. For the moment, a pair of 8-Hi stacks is by far the most likely explanation for how AMD has put 16GB of VRAM on the Vega FE. However until we hear more from AMD, it’s not the only possibility, and given that 8-Hi HBM2 still isn’t on anyone’s parts list, it may be that AMD has adopted more creative measures to get this much VRAM on a Vega card.

Rounding out the specifications, as AMD’s focus in today’s announcement was on compute performance and memory, the company hasn’t disclosed too much else about the hardware itself. The TDP has not been announced, however as we’re looking at a high-end card, 250W or better is a very good assumption, especially as AMD’s renderings of the Vega FE cards all show the card sporting a pair of 8-pin PCIe power sockets.

What we do know is that there will be two different Vega FE cards, which AMD is calling the Vega Frontier Blue Edition and the Vega Frontier Gold Edition. Information on the differences between these cards is very limited, but from the information AMD has provided, the Blue Edition will be a traditional blower-cooled card, while the Gold Edition will be taking a page from the Radeon Fury X and will feature an external radiator attached via an integrated closed loop liquid cooling system (CLLC). The use of a CLLC on the Fury X in 2015 made it a very quiet card, but at the same time it was also a necessary move due to its 275W TDP. So this may be a sign that the Gold card will have a particularly high TDP.

Otherwise at an architectural level, AMD hasn’t revealed any further details on the Vega architecture, and indeed it’s not clear if there’s much they left uncovered in this winter’s Vega architecture teaser. HBM2, the improved geometry engine, the new FP16-capable CUs, and the tiling rasterizer were all touched upon in today’s announcement. And as mentioned previously, these new architectural optimizations, along with Vega FE’s high clockspeed, are going to be the major components of the card’s performance.

Finally, with today’s announcement AMD has further narrowed down the launch date of their first Vega cards. The Vega Frontier Edition will be the first cards to ship, and as per the company’s earlier promise of an H1’17 launch, the cards will be available in late June. AMD has not announced a price for the cards yet, and like the detailed specifications themselves, I expect we’ll see that closer to next month’s launch.

But what is clear right now is that the Vega Frontier Edition is not meant to be a consumer gaming card, and that takes us into what I think is the more interesting aspect of today’s announcement: the target market.

Vega Frontier Edition’s Target Market: AI, Machine Learning, and other Professionals
Comments Locked

134 Comments

View All Comments

  • roc1 - Wednesday, May 17, 2017 - link

    Going workstation first appears provoked by insufficient supply or inability to compete in terms of performance per dollar.

    I really wish AMD had more objective slides and comparisons in their slides.
  • xype - Wednesday, May 17, 2017 - link

    Newsflash: Not everything is a conspiracy.

    Workstation = big margins. That’s what AMD needs, more income. As a business decision it’s the only thing that makes sense.
  • JKay6969AT - Wednesday, May 17, 2017 - link

    I would tend to agree that AMD's choice to go workstation first is most likely being led by supply issues with HBM2 I would disagree that it is anything to do with performance per dollar, in fact I would presume the opposite as the pro market is already dominated by nVidia and AMD would be competing directly with them to gain market share, they can't really accomplish this with poor performing cards in comparison with nVidia unless they plan to add a value sector to the pro market which is unlikely due to this being a very profitable sector and if there is one thing AMD needs more than anything right now it's profits and lots of them.

    I imagine desktop VEGA to compete in performance with the 1070, 1080 and 1080Ti while costing a bit less but the issue I have there is that of RAM or Cache size on the VEGA parts. I imagine that VEGA will launch in 4GB and 8GB varieties with HBM2 and while 8GB is plenty of RAM, comparing it to the 11GB in the 1080Ti will affect people's decision on which card to get, I know far too many people who would buy a worse GPU because it had more RAM, 'cause more is better right? LMAO!

    I imagine this is part of the reason why AMD is showing how VEGA is able to run well with 2GB RAM with the High Speed Cache Controller enabled as it will help alleviate fears that 8GB isn't enough RAM or Cache. I also get why they call it cache, It reminds me of the old Pentium 2 Single Edge Contact Cartridge or SECC design which was inserted into Slot-1 Motherboards. The reason the CPU had the Slot form factor was because the cache chips were on the daughter board around the CPU rather than in later designs built right into the CPU itself. This is a very similar design strategy and should offer comparable speed benefits. I just look towards the next evolution of this process which should be the cache being baked into the GPU itself, there are issues with that as HBM memory is multi-layered and I imagine this would present significant design issues to incorporate this into the core of a GPU but hey, I'm a dreamer :-)
  • atragorn - Tuesday, May 23, 2017 - link

    Actually the reason they went with slot-1 was to prevent AMD from sharing a socket with them.
    Super socket 7 was very good for AMD , external cache was the way it was done up to that point.
    The K6-2 and K6-3 were clocking at 500mhz plus on socket 7.
    The k6-3 was a beast with that extra cache but most people only saw it wasn't clocking as high and didnt realize the performance it offered.
  • Kvaern1 - Wednesday, May 17, 2017 - link

    Volta wants a word with you.
  • lefty2 - Wednesday, May 17, 2017 - link

    You forgot to mention that Vega was 50% faster than P100 in Deep Bench.
  • nevcairiel - Wednesday, May 17, 2017 - link

    If any comparison would be meaningful, that would be to a Tesla V100, beating the competitors last years product a year later should be a given.
  • lefty2 - Wednesday, May 17, 2017 - link

    except that V100 isn't available til August
  • cap87 - Wednesday, May 17, 2017 - link

    It'll be over a year since the P100 is released and only a month after the VEGA FE that the V100 is released. It should be compared against the more recent one.
  • Meteor2 - Wednesday, May 17, 2017 - link

    Nobody outside the DOE will be buying V100s this year.

Log in

Don't have an account? Sign up now