Friday, September 8th 2017

AMD To Change Suppliers for Vega 20 GPUs on 7nm, HBM2 Packaging for Vega 11

AMD's RX Vega supply has seen exceedingly limited quantities available since launch. This has been due to a number of reasons, though the two foremost that have been reported are: increased demand from cryptocurrency miners, who are looking towards maximizing their single node hashrate density through Vega's promising mining capabilities; and yield issues with AMD's Vega 10 HBM2 packaging partner, Advanced Semiconductor Engineering (ASE). It's expected that chip yield for Vega 10 is also lower per se, due to it having a 484 mm² die, which is more prone to defects than a smaller one, thus reducing the amount of fully-enabled GPUs.

AMD's production partner, GlobalFoundries, has historically been at the center of considerations on AMD's yield problems. That GlobalFoundries is seemingly doing a good job with Ryzen may not be much to say: those chips have incredibly small die sizes (192 mm²) for their number of cores. It seems that Global Foundries only hits problems with increased die sizes and complexity (which is, unfortunately for AMD, where it matters most).
Due to these factors, it seems that AMD is looking to change manufacturers for both their chip yield issues, and packaging yield problems. ASE, which has seen a 10% revenue increase for the month of August (not coincidentally, the month that has seen AMD's RX Vega release) is reportedly being put in charge of a much smaller number of packaging orders, with Siliconware Precision Industries (SPIL), who has already taken on some Vega 10 packaging orders of its own, being the one to receive the bulk of Vega 11 orders. Vega 11 is expected to be the mainstream version of the Vega architecture, replacing Polaris' RX 500 series. Reports peg Vega 11 as also including HBM2 memory in their design instead of GDDR5 memory. Considering AMD's HBM memory history with both the original Fury and and now RX Vega, as well as the much increased cost of HBM2's implementation versus a more conventional GDDR memory subsystem, this editor reserves itself the right to be extremely skeptical that this is true. If it's indeed true, and Vega 11 indeed does introduce HBM2 memory to the mainstream GPU market, then... We'll talk when (if) we get there.

As to its die yield issues, AMD is reported to be changing their main supplier for their 7 nm AI-geared Vega 20 from GlobalFoundries to Taiwan Semiconductor Manufacturing Company (TSMC), who has already secured orders for AI chips from NVIDIA and Google. TSMC's 7nm and CoWoS (chip-on-wafer-on-substrate) capabilities have apparently proven themselves enough for AMD to change manufacturers. How this will affect AMD and GlobalFoundries' Wafer Agreement remains to be seen, but we expect AMD will be letting go of some additional payments GlobalFoundries' way.
Sources: DigiTimes on Vega 11, DigiTimes ASE, DigiTimes on vega Shortages
Add your own comment

62 Comments on AMD To Change Suppliers for Vega 20 GPUs on 7nm, HBM2 Packaging for Vega 11

#51
Vya Domus
silentbogoThis is AMD's fault too, mostly for announcing a shitton of stuff simultaneously ahead of time with some over-optimistic time tables.
Except AMD never referred to any kind of "Vega 11" GPU officially as far as I know. Every one knew it was just one fully enabled GPU and one with some shaders disabled. How some made the conclusion that one of them is Vega 11 is beyond me. I simply fail to see how this is anymore confusing than any previous lineup from both AMD and Nvidia.



This leaked image dates all the way back to April 2016 and even then there was no mention of Vega 11.
silentbogoIf you've given up on keeping up with Kardashians Vega a few month ago, you expect Vega 10 to be already out and Vega 11 also being almost available or very near release date. Hence the confusion.
Well there you have it. It's about people not paying enough attention just as I said and the little attention that they had was aimed towards rumors. Can't blame that on AMD.
Posted on Reply
#52
efikkan
silentbogoVega 10 is what we have today as a magical and mysterious Vega64/56, a card that's meant to supersede Fury lineup.
Vega 11 is a future Polaris replacement w/ 6044 shader units at the top.
Vega 20 is an even "futurer" replacement-replacement to Vega 10.
I think you mean Vega 11 will have ~2560-3072 shader units.
Vega 20 is a compute/professional chip with more fp64 resources and possibly more memory controllers.
Vya DomusAMD never referred to any kind of "Vega 11" GPU officially as far as I know.
We know they have planned a Vega 11 GPU based on EEC filings.
But there is still very little information on it, which probably indicates that the release is not imminent.
Posted on Reply
#53
Th3pwn3r
OctopussWTF?
I give up.
I guess you have to be this geek or nerd or whatever thing to understand this shit. I have better things to do in my life than memorizing cryptic and confusing hardware names.
Someone didn't get his snack pack with his bagged lunch on that day. See ya.
Posted on Reply
#54
FordGT90Concept
"I go fast!1!11!1!"
Not much point in a Vega 11 when Vega56 holds its own against GTX 1070 and Polaris 20 holds its own to GTX 1060. If Vega 11 does exist, it may have a GDDR5X or GDDR6 controller instead of HBM. About the only reason I can think of for it to exist.

It's Vega 20 that's more interesting. Specifically, is it something they'll repackage to challenge the GTX 1080 Ti or is it going to be specifically aimed at Volta. I'm thinking the latter.
Posted on Reply
#55
StrayKAT
FordGT90ConceptNot much point in a Vega 11 when Vega56 holds its own against GTX 1070 and Polaris 20 holds its own to GTX 1060. If Vega 11 does exist, it may have a GDDR5X or GDDR6 controller instead of HBM. About the only reason I can think of for it to exist.

It's Vega 20 that's more interesting. Specifically, is it something they'll repackage to challenge the GTX 1080 Ti or is it going to be specifically aimed at Volta. I'm thinking the latter.
Those 580s are pretty inflated now... and the 1060 is only slightly more tolerable now too. There's just not any point in any of them (if you didn't buy them at the right time). It's almost time to forget about all of these chips soon enough (even though they're decent). Only 1080 and up.
Posted on Reply
#56
Th3pwn3r
StrayKATThose 580s are pretty inflated now... and the 1060 is only slightly more tolerable now too. There's just not any point in any of them (if you didn't buy them at the right time). It's almost time to forget about all of these chips soon enough (even though they're decent). Only 1080 and up.
I usually say a 1080 is a mid tier, almost high but the only 1080ti is a high end card. Especially when the price gap was around $300 between the two.
Posted on Reply
#57
jabbadap
FordGT90ConceptNot much point in a Vega 11 when Vega56 holds its own against GTX 1070 and Polaris 20 holds its own to GTX 1060. If Vega 11 does exist, it may have a GDDR5X or GDDR6 controller instead of HBM. About the only reason I can think of for it to exist.

It's Vega 20 that's more interesting. Specifically, is it something they'll repackage to challenge the GTX 1080 Ti or is it going to be specifically aimed at Volta. I'm thinking the latter.
Well direct replacement for polaris 20 with updated DirectX12.1 support and higher overclockability. AMD has done it before Tahiti->Tonga, two gcn chips with same core configs. Of course slap some faster gddr5x/gddr6 memory on the equation and it should be quite good performance uplift from RX 580.
Posted on Reply
#58
efikkan
FordGT90ConceptNot much point in a Vega 11 when Vega56 holds its own against GTX 1070 and Polaris 20 holds its own to GTX 1060. If Vega 11 does exist, it may have a GDDR5X or GDDR6 controller instead of HBM. About the only reason I can think of for it to exist.
There is a large gap between GTX 1060/RX 580 and GTX 1070/Vega 56 for anyone to fill. And keep in mind Volta will nearly be here by the time Vega 11 arrives, so it will be competing with "GTX 2060", so Vega 11 will replace Polaris in the lower-mid and low-end range.
FordGT90ConceptIt's Vega 20 that's more interesting. Specifically, is it something they'll repackage to challenge the GTX 1080 Ti or is it going to be specifically aimed at Volta. I'm thinking the latter.
It's a chip targeting Quadro and Tesla, it might not even show up in a consumer card.
Posted on Reply
#59
FordGT90Concept
"I go fast!1!11!1!"
Fiji ended up in Radeon Pro Duo and Radeon Instinct cards. AMD doesn't tend to produce a lot of silicon variations because they fundamentally don't nerf their chips.
StrayKATThose 580s are pretty inflated now... and the 1060 is only slightly more tolerable now too. There's just not any point in any of them (if you didn't buy them at the right time). It's almost time to forget about all of these chips soon enough (even though they're decent). Only 1080 and up.
Adding another Vega card won't fix that. Countries banning cryptocurrencies will.
Posted on Reply
#60
efikkan
Vega 20 will be a fp64 GPU for professional use:

It seems to be a node shrink of Vega 10 with extra fp64 resources and double memory bandwidth/capacity.
Posted on Reply
#61
FordGT90Concept
"I go fast!1!11!1!"
Global Foundries can't be that close to 7nm, can they? I sense delays. And we're talking a year away yet. If AMD really has nothing to offer between and now and then...damn.
Posted on Reply
#62
efikkan
FordGT90ConceptGlobal Foundries can't be that close to 7nm, can they? I sense delays. And we're talking a year away yet. If AMD really has nothing too offer between and now and then...damn.
These leaks (if genuine) were optimistic estimates made a long time ago. There is no guarantee these chips will be made by Global Foundries either.
Posted on Reply
Add your own comment
Dec 22nd, 2024 02:52 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts