Saturday, November 17th 2018

Samsung AMD's Second Foundry Partner for "Polaris 30"

AMD's "Polaris 30" silicon at the heart of Radeon RX 590 graphics card is the company's first 12 nm GPU. Unlike NVIDIA, which is exclusively sourcing its "Turing" family of GPUs from TSMC, the "Polaris 30" is coming from not one, but two sources. This, according to AMD in response to a question by TechPowerUp. The two foundries manufacturing "Polaris 30" are GlobalFoundries and Samsung. AMD did not provide us with visual cues on how to tell chips made from either foundries apart (such as serial numbering schemes). Packaging of dies sourced from both foundries is done in China, and the national-origin marking for the chip is on the package, rather than printed on the die.

GlobalFoundries' 12 nanometer FinFET node, called GloFo 12LP, shares a lot of similarities with Samsung's 11LPP, because both are "nodelets" that are derived from an original 14 nm FinFET process blueprint Samsung licensed to GloFo, deployed in its facility in upstate New York, where AMD's "Zen" processors are made. GloFo's 12 nanometer process is a refinement of its 14 nm node, in which 12 nm transistors are etched onto silicon using the same lithography meant for 14 nm. It doesn't improve transistor densities, but provides dividends in power, which explains why "Polaris 30" and "Pinnacle Ridge" have the same die sizes as "Polaris 20" and "Summit Ridge," respectively. This WikiChip article provides a good explanation of how GloFo 12LP is a nodelet.
AMD appears highly motivated to be first-to-market with both 7 nm CPU and GPU, which come in the form of its "Vega 20" GPU and "Rome" CPU, which could be formally launched before the end of 2018, with production being ramped up through 2019. AMD confirmed publicly earlier, that its first 7 nm chips could be sourced from TSMC, but even it realizes that TSMC has an exhaustive list of clientele for 7 nm, each with its own foundry allocation. AMD could try to find additional sources for 7 nm chips, and we predict Samsung to be one of them. GloFo canned its 7 nm plans, and AMD is keeping it busy with wafer contracts for "Zen," "Zen+," "Polaris," "Vega 10" GPU die, and the 14 nm I/O controller die at the heart of its EPYC "Rome" MCM.
Add your own comment

41 Comments on Samsung AMD's Second Foundry Partner for "Polaris 30"

#26
qcmadness
ET3DBoth Polaris 30 and Polaris 10 use the same production plants, right? Won't they come at the expense of one another anyway? I assume that AMD has quite a big stock of Polaris 10 to get rid of, but once that's gone, why would more Polaris 10 chips be made? (Of course, it's possible that AMD has so many Polaris 10 chips left that's it's an academic question.)
Judging from the previous lack of stock during the cypro-mining craze, I don't think AMD has that much Polaris 10 stock.
Posted on Reply
#27
$ReaPeR$
i dont think there will be a problem with this approach. if the manufacturing nodes are decent enough i dont see why there should be a problem with using 2 different manu companies.
Posted on Reply
#28
ixi
stimpy88OK, well we now have a silicon lottery regarding which chip we get, and which chip runs hotter, faster etc... Great AMD...

Samsung has the inferior process node and manufacturing quality. We found this with the iPhone...
And you want to tell me this is something new? Basically all the chips from intel, amd cpu's are the same, uses the same power voltage, can oc to the same GHz e.t.c, like really?
Posted on Reply
#29
medi01
theoneandonlymrkthey're 7nm is compatible with TSMC
How does that work?
Posted on Reply
#30
TheoneandonlyMrK
medi01How does that work?
I am not sure on the details but do know a few designs , like Polaris for Amd for example were designed to be compatible across foundries at 14nm , and I am fairly sure Asml ship similar machines to each foundrie, the difference between foundries stems from processing difference not an inate difference in starting posture or technology.
They will all get euv lines within the same few years but they still wont quite be used the same yet net result is the same, what they say is a 7nm chip.
Posted on Reply
#31
jeremyshaw
Totally"Fab 2, located in Woodlands, Singapore. This fab is capable of manufacturing wafers at 600 to 350 nm for use in selected automotive IC products, High Voltage power management IC and Mixed-signal products."
Ah, the old Chartered Semi plant. Glad to see it wasn't wasted.
Posted on Reply
#32
Vya Domus
hat600-350nm manufacturing? Who is still using this?
You're not going to make a micro controller in a washing machine on a leading node are you ?
Posted on Reply
#33
hat
Enthusiast
I guess? I thought smaller processes were cheaper? Why would you manufacture a gargantuan 600nm chip when you can make 14nm chips? Even if "leading nodes" like the ones currently making high end products like processors, ram, graphics chips etc are all loaded up, you could probably at least do 45nm or something.
Posted on Reply
#34
Vya Domus
hatI guess? I thought smaller processes were cheaper? Why would you manufacture a gargantuan 600nm chip when you can make 14nm chips?
The only chips that get to be gargantuan are GPUs and CPU and the likes of those. They get to that level because they have billions of transistors after all, simple chips don't come even close to that, so even on an antiquated node they are still small and use less power.
hatyou could probably at least do 45nm or something.
Some of them are, but simple 8-bit micro controllers or chips that are meant to perform some basic logic simply don't need to be on the latest manufacturing process.
Posted on Reply
#35
Totally
hatI guess? I thought smaller processes were cheaper? Why would you manufacture a gargantuan 600nm chip when you can make 14nm chips? Even if "leading nodes" like the ones currently making high end products like processors, ram, graphics chips etc are all loaded up, you could probably at least do 45nm or something.
More mature process means translates to less failures per wafer, demand for that process is probably lower therefore cheaper also outweighing the benefits of a smaller node.
Posted on Reply
#36
Prima.Vera
TotallyDoesn't anyone find it odd that Samsung doesn't have fab/foundry time for their own products but has no problem making time for other chip makers?
Noup. Because they get MONEYZZZZZ to fab chips for other makers, while doing for themselves it costs. ;)
Posted on Reply
#37
Vayra86
stimpy88OK, well we now have a silicon lottery regarding which chip we get, and which chip runs hotter, faster etc... Great AMD...

Samsung has the inferior process node and manufacturing quality. We found this with the iPhone...
We've also found that Samsung has fantastic results on DRAM and NAND flash. Its always a lottery. In all fairness I have a lot of faith that Samsung will be leading in the foundry business too, given time.

Even Intel's attempts to make every fab identical do not result in identical production runs, far from it even. Look at the 8700K.
Posted on Reply
#38
Casecutter
FordGT90ConceptThey might just sell the inferior chips as RX 580/RX 570. Perhaps as RX 585 and RX 575. It really depends on what kind of yields they're getting.
Yeah this only being a Polaris 30 (aka RX 590) I would believe they'd not need some second source. So likely AMD is going to go 12nm for the RX 580/570 also, and use up any lesser grade in either power or SU count.
Posted on Reply
#39
Totally
Prima.VeraNoup. Because they get MONEYZZZZZ to fab chips for other makers, while doing for themselves it costs. ;)
but at the same time they're holding their hands up saying "We need to charge more monies because don't have enough fab capacity for our wares"
Posted on Reply
#40
Vayra86
Totallybut at the same time they're holding their hands up saying "We need to charge more monies because don't have enough fab capacity for our wares"
Of course, they need the money to build more fabs they can shut down to control supply.

Sound logic no?
Posted on Reply
#41
Totally
Vayra86Of course, they need the money to build more fabs they can shut down to control supply.

Sound logic no?
Only if it's a monopoly. Coming full circle since they're clearly demonstrating behaviors of such an entity why isn't anyone taking action?
Posted on Reply
Add your own comment
Nov 21st, 2024 13:01 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts