Tuesday, May 16th 2017

AMD Announces Radeon Vega Frontier Edition - Not for Gamers

Where is Vega? When is it launching? On AMD's Financial Analyst Day 2017, Raja Koduri spoke about the speculation in the past few weeks, and brought us an answer: Radeon Vega Frontier Edition is the first iteration of Vega, aimed at data scientists, immersion engineers and product designers. It will be released in the second half of June for AMD's "pioneers". The wording, that Vega Frontier Edition will be released in the second half of June, makes it so that AMD still technically releases Vega in the 2H 2017... It's just not the consumer, gaming Vega version of the chip. This could unfortunately signify an after-June release time-frame for consumer GPUs based on the Vega micro-architecture.

This news comes as a disappointment to all gamers who have been hoping for Vega for gaming, because it reminds of what happened with dual Fiji. A promising design which ended up unsuitable for gaming and was thus marketed for content creators as Radeon Pro Duo, with little success. But there is still hope: it just looks like we really will have to wait for Computex 2017 to see some measure of details on Vega's gaming prowess.

Vega Frontier Edition is the Vega GPU we've been seeing in leaks in the last few weeks, packing 16 GB of HBM2 memory, which, as we posited, didn't really make much sense on typical gaming workloads. But we have to say that if AMD's Vega truly does deliver only a 1.5x improvement in FP32 performance (the one that's most critical for gaming at the moment), this probably paints AMD's Vega as fighting an uphill battle against NVIDIA's Pascal architecture (probably ending up somewhere between GTX 1070 and GTX 1080). If these are correct, this could mean a dual GPU Vega is indeed in the works, so as to allow AMD to reclaim the performance crown from NVIDIA, albeit with a dual-GPU configuration against NVIDIA's current single-chip performance king, Titan Xp. Also worth nothing is that the AMD Radeon Vega Frontier Edition uses two PCI-Express 8-pin power connectors, which suggests a power draw north of 300 Watts.
For now, it seems AMD actually did its best to go all out on the machine learning craze, looking for the higher profits that are available in the professional market segment than on the consumer side of graphics. Let's just hope they didn't do so at the expense of gaming performance leaps.

After an initial throwback to AMD's times since he became lead of Radeon Technologies Group, where Raja mentioned the growing amount of graphics engineers in AMD, including their commitment to the basics of graphics computing: power, performance, and software. Better basics in hardware, software, and marketing are things that Raja says are responsible for AMD's current market outlook, both from a gamer and content creator perspective, which led to an increase in AMD's graphics marketshare.
RTG's chapter two of Radeon Rising, going beyond the basics, will allow the company to go after premium market dollars, with an architecture that excels on both gaming and CAD applications. Raja Koduri said he agreed with NVIDIA CEO Jensen Huang in that at some point in the future, every single human being will be a gamer.
The final configuration of Vega was finalized some two years ago, and AMD's vision for it was to have a GPU that could plow through 4K resolutions at over 60 frames per second. And Vega has achieved it. Sniper Elite 4 at over 60 FPS on 4K. Afterwards, Raja talked about AMD's High Bandwidth Cache Controller, running Rise of the Tomb Raider, giving the system only 2 GB of system memory, with the HBCC-enabled system delivering more than 3x the minimum frame-rates than the non-HBCC enabled system, something we've seen in the past, though on Deus Ex: mankind Divided. So now we know that wasn't just a single-shot trick.
Raja Koduri then showed AMD's SSG implementation and how it works on a fully ray-traced environment, with the SSG system delivering much smoother transitions in the system. AMD worked with Adobe on integrating SSG capability into Adobe Premiere Pro.
Raja then jumped towards machine intelligence, which Raja believes will be dominated not by the GPU (NVIDIA's green) or CPU (Intel blue) paths, but in true heterogeneous computing.
Raja took to stage results on DeepBench, a machine learning benchmark where NVIDIA dominates at the moment, joking about AMD's absence from the benchmark - since they really didn't have a presence in this area. In a benchmark, AMD pitted Vega against NVIDIA's P100 architecture (interestingly, not against NVIDIA's recently announced V100 architecture, which brings many specific improvements to this kind of workloads), delivering an almost 30% performance lead.
Add your own comment

91 Comments on AMD Announces Radeon Vega Frontier Edition - Not for Gamers

#26
TheGuruStud
Why is there not a single video of this presentation online?

And where would they get a volta to test from? The video card fairy that appears out of ether? All 20 they made so far were sold LOL
Posted on Reply
#27
cotes42
Can someone please tell me what the 4 screw holes are for at the end of the card! I have a Radeon Pro Duo with the same 4 screw holes and I asked everywhere and no one can make a guess. Thanks!

Posted on Reply
#28
FordGT90Concept
"I go fast!1!11!1!"
Server cases can have mounts on both sides of the card to prevent it from sagging.
Posted on Reply
#29
ZoneDymo
xkm1948p100, not v100
yeah he knows, he is asking how AMD is suppose to pit their stuff against something that has only been announced...aka not here yet.
Posted on Reply
#30
Camm
People really need to get it into their heads that AMD is NOT an enthusiast company anymore. (but some of our interests may intersect).

Why? Well for better or for worse, we didn't buy their products (even when they did have superior designs), so they've moved onto other markets where they will.

So of course Vega is targeted at FP16, and is why we've had a tonne of talk from AMD about working with companies to target shaders for FP16 (games don't really need FP32 for most tasks). Most computational tasks use FP16, so guess where the bulk of Vega's design is!

Its like Ryzen with its CCX setup. Cross CCX calls hurt gaming tasks, but most batch computational tasks don't care about the increase in latency, and is beneficial overall as it allows AMD to avoid the increasingly large and complex amount of silicon in order to maintain cache coherency like Intel designs (allowing more cores, lower power usage and faster cores overall).
Posted on Reply
#31
W1zzard
Cammis targeted at FP16
If you have FP16 support then FP16 is twice as fast as FP32 because 16 vs 32 bits. Not a huge miracle I'd say
Posted on Reply
#32
R-T-B
eidairaman1Too bad this isn't the days of past where workstation boards could be bios modded for gaming.
Wouldn't make a difference as cost is sure to be the limiting factor, not the bios.
Posted on Reply
#33
xkm1948
So according to this


Vega will have less memory bandwidth than the FuryX?

If everything is optimized, shouldn't the card utilize more bandwidth? Also the 16GB is only 2048bit I would assume?

On a side note, that metallic blue color surely is beautiful. The gold one is not that bad as well.
Posted on Reply
#34
FordGT90Concept
"I go fast!1!11!1!"
HBM bandwidth is directly tied to how many stacks. It has less bandwidth than Fiji because it has two stacks versus four and presumably at a slightly lower clock speed too. Thing is, those two stacks are 8 x 1 GiB layers so it is much denser.

Vega should be cheaper to manufacturer than Fiji.
Posted on Reply
#35
xkm1948
FordGT90ConceptHBM bandwidth is directly tied to how many stacks. It has less bandwidth than Fiji because it has two stacks versus four and presumably at a slightly lower clock speed too. Thing is, those two stacks are 8 x 1 GiB layers so it is much denser.

Vega should be cheaper to manufacturer than Fiji.
I was hoping these to be 4 stacks of 4GB at 4096bit. Oh well.

The rumor of Vega in limited quantity could be true. If the gaming performance doesn't justify Vega in gaming GPU ranks, I can totally see AMD rebrand it as deep learning GPU and sell it for way better profit than gaming GPU.
Posted on Reply
#36
TheinsanegamerN
CammPeople really need to get it into their heads that AMD is NOT an enthusiast company anymore. (but some of our interests may intersect).
The existence of ryzen seems to disagree with you. It's performance is the definition of "enthusiast". It's a whole 2-5% slower in games (while still maintaining over 100FPS) and wallops intel in creative production, with better price/perf to boot and an 8 core chip in the standard socket that doesnt suck.
Why? Well for better or for worse, we didn't buy their products (even when they did have superior designs), so they've moved onto other markets where they will.
That simply isnt true. When AMD was competitive, IE the 5000 and 7000 series, They sold quite well.

AMD could never get their driver game together when they had good hardware, and that scared away quite a few consumers. By the time they got their drivers mostly straightened out (400 series), their hardware was a generation behind.

Given how complicated GPUs are, it will take years for RTG to straighten out the damage that AMD did to the ATI GPU. With ryzen selling well and vega finally out, R+D should finally start catching up to nvidia.
Posted on Reply
#37
Camm
TheinsanegamerNThe existence of ryzen seems to disagree with you. It's performance is the definition of "enthusiast". It's a whole 2-5% slower in games (while still maintaining over 100FPS) and wallops intel in creative production, with better price/perf to boot and an 8 core chip in the standard socket that doesnt suck..
Quite the contrary. Ryzen is an example where our interests align. However a truly aimed at enthusiasts chip wouldn't have used a LPP process for the chip, and would likely have been a single heterogeneous design. It just works out that because Intel has been bleeding the market from a monopoly perspective that Ryzen works 'well enough' in gaming that we are more than happy to lap up its other advantages.
Posted on Reply
#38
cdawall
where the hell are my stars
And what do you know consumers are delayed again. I am no longer worried that I made the wrong choice. Thank you AMD for confirming I shouldn't wait. This is frustrating I just want to see the succeed, but it is like they are determined to let us down at every single turn.
Posted on Reply
#39
Brusfantomet
TheinsanegamerNGiven how complicated GPUs are, it will take years for RTG to straighten out the damage that AMD did to the ATI GPU. With ryzen selling well and vega finally out, R+D should finally start catching up to nvidia.
The driver problems for the radeon line stems from the ATi time, not from the AMD time (first AMD designed radeon was the HD 7970 with GCN. Previews designs was based on VLIW design dating from the 9700 series)
Posted on Reply
#40
evernessince
CammQuite the contrary. Ryzen is an example where our interests align. However a truly aimed at enthusiasts chip wouldn't have used a LPP process for the chip, and would likely have been a single heterogeneous design. It just works out that because Intel has been bleeding the market from a monopoly perspective that Ryzen works 'well enough' in gaming that we are more than happy to lap up its other advantages.
Yes, because you know what Ryzen needed more than AMD's engineers. I must have missed that it's a rousing success.

FYI enthusiast does not equal forget efficiency, especially in today's world.
Posted on Reply
#41
R00kie
I seriously hope that HBCC isn't something that game developers have to code for, and it is just driver dependant...
Posted on Reply
#42
TheGuruStud
cdawallAnd what do you know consumers are delayed again. I am no longer worried that I made the wrong choice. Thank you AMD for confirming I shouldn't wait. This is frustrating I just want to see the succeed, but it is like they are determined to let us down at every single turn.
Give them a few billion and you wouldn't have to wait.

What's that? Intel stole it all and got away with it? No way....
Posted on Reply
#43
cdawall
where the hell are my stars
TheGuruStudGive them a few billion and you wouldn't have to wait.

What's that? Intel stole it all and got away with it? No way....
Keep clinging to that one. I have played that card and played that card the entire time that the FX series was their lineup. It isn't going to change how piss poor the FX design was and how horridly it ruined their name in the market. That was on them, not intel, not the consumers. They did a bad job and payed for it with market share.



Mind showing me on here where intel cheated AMD? oh wait that was the P4 era. 2006 was the release of the Core 2 architecture and AMD never came back, they never released a product that performed better in that entire time either.
Posted on Reply
#44
FordGT90Concept
"I go fast!1!11!1!"
2006 was the year Core 2 Duo debuted and AMD bought out ATI (so AMD had no capital). Bulldozer debuted in 2011.
Posted on Reply
#45
cdawall
where the hell are my stars
FordGT90Concept2006 was the year Core 2 Duo debuted and AMD bought out ATI (so AMD had no capital). Bulldozer debuted in 2011.
Yup core 2 marked the end for AMD having competing market share. Bulldozer only made it worse. Here is to the hope that Ryzen can start the deep dig out.
Posted on Reply
#46
OneCool
Not surprised really. I'm sure I'm not the only one that's noticed the trend in AMDs releases.
Massive work station/rendering CPUs, upgrades to their workstation pro line GPUs...the threadripper!!?? Coming too

AMD invested a lot into consule gaming with Sony, Nintendo and Xbox.
Why would they want to improve their PC gaming hardware?
Posted on Reply
#47
sweet
xkm1948So according to this


Vega will have less memory bandwidth than the FuryX?

If everything is optimized, shouldn't the card utilize more bandwidth? Also the 16GB is only 2048bit I would assume?

On a side note, that metallic blue color surely is beautiful. The gold one is not that bad as well.
They introduced new compression with Polaris already. Less bandwidth but more effective now.
Posted on Reply
#48
ShurikN
Oh my god, the watercooled golden version is sick. Probably the best looking piece of hardware I've seen since In Win 904.
Posted on Reply
#49
RejZoR
People complaining Vega isn't positioned against V100. Dudes, all of you, V100 DOESN'T EXIST YET.

Also, what's up with "It's not a gamer card" and then they place it on a graph next to R9 Fury X. Ugh?
Posted on Reply
#50
RejZoR
AbsolutionWelp, always one step behind AMD :(
Why people assume it's one steep behind just because their release schedules don't line up as they used to and AMD's isn't the first? Time wise, sure, it's "late", but that's how it is. Product releases don't match up with release timeframes anymore and I don't think they'll line up any time soon. I mean, that would mean releasing Vega now and releasing Navi sometime in Fall 2017. Which is just not going to happen.
Posted on Reply
Add your own comment
May 8th, 2024 07:01 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts