AMD’s Vega isn't finished yet, Vega 11 goes into production to replace Polaris

AMD Radeon RX Vega 64 specs

Things have finally settled down for AMD after Vega. While we all sit back and wonder what could’ve been, AMD have already set their sights on the next iteration in AMD’s mainstream graphics card family, Vega 11.

Here are the best graphics cards you can buy right now.

Polaris-based GPUs have served us well but their time is coming to an end. Vega 11 is a cut-down version of the Vega we’ve seen so far, which is intended to replace 400- and 500-series graphics cards.

Packaging orders for Vega 11 have been sent to Siliconware Precision Industries (SPIL) - the company responsible for packaging Vega chips - according to a source close to DigiTimes. Global Foundries, AMDs go-to chip manufacturer, are also expected to continue their partnership and will be manufacturing these future Vega 11 GPUs.

So far, Vega architecture has been paired with HBM2 memory. However, cut-down Vega 11 is expected to utilise GDDR5, which is found in the majority of graphics cards on the market today. Due to the high cost and low availability of HBM2, it will be tricky for AMD to create budget-minded graphics cards while utilising the higher-spec memory technology.

Nvidia Volta GPU specs

AMD have been struggling with Vega chips, and it seems they would only be in a worse position, attempting to meet the demands of the mainstream market with Vega tech, if they tried matching it with HBM2. It seems too optimistic for AMD to entirely ditch GDDR5 at this point and it’s doubtful whether AMD will be using exclusively HBM2 in their Navi-based graphics cards down the road.

Also hinted at in the report is Vega 20. It is reportedly going to be manufactured by TSMC using a 7nm FinFET process and is expected to be the most powerful Vega architecture part from AMD. But before you get too excited, this is intended entirely for AI and supercomputers.

AMD’s public roadmaps have so far touted the next-gen Navi architecture as their first 7nm design, with the exception of leaked internal slides, of dubious veracity, from 2016 that reference 7nm Vega 20 chips. Whether we will see 7nm Vega chips is still a mystery, as nothing from AMD has been clear on the matter. AMD will take the better part of 2018 to even get close to profitable yields with the 7nm process, so don’t hold your breath for 7nm, whatever the case may be.

AMD GPU roadmap

While Navi is touted by AMD as the killer architecture for the next generation, it’s not expected until 2019. We could see another high-end Vega-based card in the meantime, alongside Vega 11-based graphics cards. AMD’s roadmaps indicate we are still expecting a refreshed 14nm+ Vega-based chip before Navi is available. The final iteration of 14nm, which has been used since Polaris, is supposedly this 14nm+ refresh, before AMD move onto a 7nm process with Navi.

Vega 11 graphics cards are supposedly arriving sometime in 2018 and 14nm+ Vega could be the next mid- to high-end option for AMD customers around the same time.

Paladins
Sign in to Commentlogin to comment
Firerod avatarSirEpicPwner avatar[HFA]Dragonstongue avatardeksroning avatarJacob Ridley avatarDragonstongue avatar
Firerod Avatar
17
2 Months ago

The top Vega 11 card would have to be close to a 1070 at least, meaning RX 56/64 will have to be refreshed to be faster when Volta comes out.

2
SirEpicPwner Avatar
3
2 Months ago

I think that's the 14nm+ Vega in the roadmap slide from AMD. Probably will be higher clocks, and hopefully some power draw improvements as well.

1
[HFA]Dragonstongue Avatar
5
2 Months ago

Vega 56 is already "matched" with 1070 really, so, I would imagine the stepping stone for Vega 11 being "cut down" would be to be a bit faster then RX 570/580 similar price and raining on the GTX 1060 parade, if they can squeak the Vega 11 in there (Vega 46? 42?)) there is a happy middle ground between 1060/1070 that RX 500 series kind of hits but IMO a tad pricey most cases, and if they do it just right it will be more or less not worthless for mining purposes but excellent for those wanting/needing to upgrade at the ~$200-$300 price point (CAD$ especially..AMD really needs to hammer partners to be more attentive to price points)

I hope they really make sure is "tuned" for the maximum efficiency (i.e less power for the required clocks right off the bat) instead of running higher power than it needs AND a good cooler. RX 400 to 500 really only thing was higher clocks so faster but more power, Vega 11 done just right can give better to similar performance at lower power and temps (Vega 64 "tuned" daws 174w getting a ton of mining performance ~39-47MH/s so there is potential there IF AMD takes the bit of time to ensure is as good as it can)

Anyways, time will tell, I am not overly fond of the new not catalyst drivers TBH. but shelling ~$90 more then should if not more for an "upgrade" over my much old 7870 now not pleased with the seeming greed of makers or sellers for RX/Vega being far overpriced than what should be, we already have 1 Ngreedia do not need other joining that camp ^.^

1
deksroning Avatar
6
2 Months ago

Vega 56 already outperforms GTX 1070.

Also bear in mind that when Vga 56 is undervolted at the core, and overclocked at the memory, its performance gets close to or exceeds GTX 1080 (with lower power draw than 1080).

AMD overvolts their GPU's to increase yields... hence why undervolting fixes power efficiency and performance (as it also removes thermal throttling).

But AMD is also making their GPU's on a manuf. process that's not suitable for high clock speeds. Nvidia has access to Samsung's manuf. process that allowed them to make Pascal (which is nothing more than an overclocked Maxwell).

Finally, Vega's architecture apparently is not optimized for games per Raja Koduri's statement, and there are several features in Vega that need to be used by developers in order to see more performance.

1
Dragonstongue Avatar
3
3 Weeks ago

Nvidia uses TSMC NOT Samsung, GF (Global Foundries) uses/shares IBM/Samsung 14nm designs, whereas TSMC uses their own 16nm process, the main reason why Nv 1000 series is "so fast" they optimized the design to cut out all the "extras" to focus purely on gaming tasks not the advanced stuff that can be found in hashing or "true" DX12 product stack which polaris/vega are capable of far far more than GTX 1000 series.

if Nvidia were to use more or less the same transistor density and extras to get all that DX12 offers without relying on software tricks/hacks to make it "seem" like they are fast as they appear, their apparent efficiency would go into the toilet as would their raw clock speed advantage drop like a stone....they can force transistors to operate at a higher frequency if they are "lean" read optimized for clock speed alone, but, if they are "fat" to be able to do more they simply cannot clock as high.

Polaris/Vega may not be able to clock as high, but, they have much performance for the clocks they CAN hit, much like Ryzen may have a clock speed deficit vs various core i models they compete on near even footing at a lower clock speed might suffer some on "optimized" high IPC apps/games just like GTX1000 series does.

anyways, long story short, Vega was meant to be a workhorse just like pretty much every radeon ever released, whereas GTX cards for years now have been going more and more towards gaming grunt with less "fat" design.

Also to guy below me, supposedly it is NOT Volta that was rumored to use GDDR6 (they were supposed to be using either HBM2 or HBM3 once AMD "first access" is gone) but likely it will be GDDR5x for Nv at faster speed and AMD to use GDDR6 first.

Historically AMD used the newer memory standards and much of the newer features with various DX/OpenGL versions and Nv came AFTER others proved it useful just so Nv could come in and force multi millions of $ down throat of folks like MSFT to optimize/tweak for their specific needs at the cost of gamers/devs.

Tesselation is a prime example of that, had MSFT NOT basically forced AMD to build according to NV whims crying "unfair advantage" after AMD spent many years and many millions of $ implementing it generation after generation they would likely have stomped Nv to the curb, instead, NV was allowed to "trick" software to make them appear so much better at it even if the density of the final image was subpar but faster than AMD, same with PhysX before NV took ability away from Radeons to use it (because it made many tier higher NV cards look crap in comparison)

Radeon tend to be more horsepower/grunt whereas GTX tend to be (for many years now) all about tuner type in comparison.

anyways :)

1
SirEpicPwner Avatar
3
2 Months ago

If AMD does end up using GDDR5 in Vega 11, would that put them behind Volta midrange cards? I heard a rumor that Volta would be using GDDR6, and have a lot higher memory bandwidth as a result.

1
Jacob Ridley Avatar
11
2 Months ago

HBM2 just seems far too expensive for the midrange market without huge losses on every card so I'm not sure if there are many other options available to AMD except hunker down and wait for Navi. GDDR5 may help keep prices down and AMD should still be competitive at their price points.

Whatever the outcome within reason, bandwidth is not going to be such a dealbreaker for cards aimed at lower resolutions than 4K.

1