Bookmark and Share


At the Common Platform Technology Forum last week GlobalFoundries and its partners from the alliance disclosed some of their long-term plans. As expected, both GlobalFoundries and Samsung Semiconductor have no plans to slowdown the evolution of process technology development. GlobalFoundries believes it will ramp up production using 14nm-XM process tech in 2014 and will follow up with 10nm-XM and even 7nm in the next four years.

In a bid to produce more power efficient and therefore more competitive chips for their clients, GlobalFoundries will initiate production using 20nm process technology in 2013 and will be able to ramp up manufacturing using hybrid 14nm-XM process (which combines 14nm FinFET transistors with 20nm back-end-of-line (BEOL) interconnect flow) in 2014. Apparently, there is also 10nm-XM hybrid process (which will utilize 10nm FinFET with 14nm BEOL) in plans for 2015 as well as 7nm technology due in 2015.

In a bid to speed up development and deployment of new process technology, members of the Common Platform alliance (also known as AMD fab club) will start deploying new process technologies before they get approved by IBM itself. In general, this should speed up time-to-market and maturity of the latest manufacturing technologies, even though at higher cost for foundries.

While it is crucially important to develop leading-edge fabrication processes, it is also essential to ensure high production volumes. While GlobalFoundries will be able to process around 190 thousand 300mm wafers per month (including 20nm, 28nm, 32nm, 40nm and other process technologies) by late 2013, its arch-rival Taiwan Semiconductor Manufacturing Company can already process 1.076 million 300mm wafers a month and will further boost output by the end of the year.

Tags: Globalfoundries, Semiconductor, IBM, Samsung, 20nm, 14nm, 10nm, 7nm, 14nm-XM, 10nm-XM, TSMC


Comments currently: 12
Discussion started: 02/11/13 03:05:58 PM
Latest comment: 02/15/13 03:15:52 AM
Expand all threads | Collapse all threads


Matching Intel's foundry tech. It's something AMD alone would have had no chance of doing, hence the spin-off. Now AMD is assured of keeping up with Intel's die shrinks. In fact when 22nm Bay-Trail is released in 6 months, don't be surprised if AMD hits them with the successor to Temash on 20nm just a quarter later.
2 3 [Posted by: linuxlowdown  | Date: 02/11/13 03:05:58 PM]

"AMD is assurred of keeping up"?

Hardly. First of all, Global Foundries has a history of making announcements like this and then missing deadlines and/or having very low yields.

Second, there is more to making good CPUs than just the lithography size. AMD has repeatedly failed on the design side the past few years, with the exception of their integrated GPUs (which are pretty good).
3 0 [Posted by: john42  | Date: 02/11/13 04:32:40 PM]
- collapse thread

So if AMD can keep tuning their designs and get 10% better performance with each generation, along with faster ramp of die shrinks, who will be laughing then? (Hopefully this ain't just a pipe dream/vaporware)
2 1 [Posted by: Pedro_mann  | Date: 02/11/13 09:35:52 PM]
@john42. I agree that there is more to CPU design than just die shrinks. If you were to compare AMD and Intel CPUs on the same node, AMD CPUs are more competitive (except for Bulldozer that was first gen of new arch). Intel are ahead in the game at the moment simply because they are one to two process nodes ahead. AMD CPUs are now better designed and tweaked after 2.5 generations. Kaveri will be gen 3. And as it happily turns out, this new CPU/APU architecture scales down better for efficiency than up. Once they catch up on die shrink, Intel will be outdone - in fact Temash has outdone them. The green tortoise and the blue hare. You know, with traditional PC sales down (and eventually big chip servers too) and Intel heavily weighted in engineering and management resources towards big powerful processors, I wonder how they're going to keep open those big expensive foundries that have given them the edge year after year. This is an issue that has been recently raised elsewhere on the net. Watch for negative signs after 2Q2013 when Haswell sales fail to impress.
3 2 [Posted by: linuxlowdown  | Date: 02/12/13 12:50:25 AM]

Optimism catches up to reality at some point. There's no way they're going to get 10 nm in 2 years, or at least not any sort of HVM/production version.

You have to develop what your customers want...and 80% of GloFo / TSMC's business is for older processes. Intel is their own customer, which is a nice advantage when you want to constantly be pushing the envelope.

@linuxlowdown. AMD isn't going to catch up at this point. The race isn't even close when Intel could buy AMD 10 times over (I'm not exagerrating...literally Intel is worth 10x as much as AMD). They're clinging to 28nm because they don't have the resources anymore to push lower: especially with a 15% engineering layoff that's in progress.
2 0 [Posted by: AnonymousGuy  | Date: 02/12/13 06:01:11 AM]
- collapse thread

But AMD is first to market - with Temash. Bay Trail is many months away. By the time Bay Trail comes out on 22nm, AMD will be moving on to its 4th gen core, most likely at 20nm. BTW, Intels 22nm is regarded by the rest of the industry as 26nm.

Quote: "In fact, GF's 14nm process may have smaller features than Intel's 14nm process because, says Mojy Chian senior vp at Globalfoundries, because "Intel's terminology doesn't typically correlate with the terminology used by the foundry industry. For instance Intel's 22nm in terms of the back-end metallisation is similar to the foundry industry's 28nm. The design rules and pitch for Intel's 22nm are very similar to those for foundries' 28nm processes."

Jean-Marc Chery, CTO of STMicroelectronics points out that the drawn gate length on Intel's ˜22nm" process is actually 26nm.

Furthermore Intel's triangular fins, which degrade the advantages of finfet processing could underperform GF's rectangular fins which optimise the finfet advantage."
3 1 [Posted by: linuxlowdown  | Date: 02/12/13 06:15:59 AM]
You're quoting a customer of GloFo who is probably cherry picking their "facts" to play down the fact that they're using an inferior process, because last I checked nothing can compete with Intel's performance or power metrics.

Not to mention Broadwell wafers are already taped out and being debugged, so 22nm is now old.
1 1 [Posted by: AnonymousGuy  | Date: 02/12/13 10:28:32 AM]
How do you know he's not wrong. Btw, no one other than Intel has done a x86 chip at 22nm. so there is nothing to compare with Intel's "superior" fab process at this stage. 14nm will be more apples with apples.
1 2 [Posted by: linuxlowdown  | Date: 02/12/13 03:16:55 PM]
How do you know he's right? Intel doesn't publish specs and there's probably dozens of different VLSI parameters. 22 nm has to be coming from somewhere within the process, following an industry standard. Otherwise Intel would be getting sued from both competition and investors for misrepresenting their technology.
1 1 [Posted by: AnonymousGuy  | Date: 02/12/13 03:38:12 PM]
You have an answer for everything! How about an answer to this: Intergraph vs Intel (1997). This is what happens when a company sues the Intel juggernaut... and wins. Intel then forces them into bankruptcy. Fricken evil, eh?
0 2 [Posted by: linuxlowdown  | Date: 02/13/13 04:41:52 AM]

AMD are terrible at executing on Road maps, and Globalfoundries are worse having been born from AMD themselves. 28nm has been on the shelves since over a year and we still have no AMD 28nm APU`s or CPU`s, apart from GPU`s...

Globalfoundries need to prove themselves before anything else, or else customers will never take them seriously. Road-maps and allot of hot air is only good for Investors.
2 1 [Posted by: keysplayer  | Date: 02/12/13 06:04:14 AM]

it is interesting how nobody understood a thing about article...

GF and TSMC are FOUNDRIES. they have many customers, only one of which is AMD, and not the largest one. Apple, NVIDIA, all the RISC bunch - all of them produce their stuff there.

when foundry says they will have Xnm-process at year 201X, it is not a form of Intel vs. AMD battle - surely, some of process capability will influence AMD also, but company has their products made at both TSMC and GF and isn't major customer at either. which also translates that others will be paying for transition to lower process-nodes, some of which moreso than AMD

so, when GF claims 10nm node, it doesn't translate in any way to AMD having 10nm CPU/GPU/APU/whatever at the exact same moment... and about that, what good did 22nm transition of Intel gave to consumers? not much at this moment, at least at desktop market... except the price remained the same, and chip is cheaper to produce - but not cheaper to buy
0 0 [Posted by: snakefist  | Date: 02/15/13 03:15:52 AM]


Add your Comment

Related news

Latest News

Wednesday, November 5, 2014

10:48 pm | LG’s Unique Ultra-Wide Curved 34” Display Finally Hits the Market. LG 34UC97 Available in the U.S. and the U.K.

Wednesday, October 8, 2014

12:52 pm | Lisa Su Appointed as New CEO of Advanced Micro Devices. Rory Read Steps Down, Lisa Su Becomes New CEO of AMD

Thursday, August 28, 2014

4:22 am | AMD Has No Plans to Reconsider Recommended Prices of Radeon R9 Graphics Cards. AMD Will Not Lower Recommended Prices of Radeon R9 Graphics Solutions

Wednesday, August 27, 2014

1:09 pm | Samsung Begins to Produce 2.13GHz 64GB DDR4 Memory Modules. Samsung Uses TSV DRAMs for 64GB DDR4 RDIMMs

Tuesday, August 26, 2014

10:41 am | AMD Quietly Reveals Third Iteration of GCN Architecture with Tonga GPU. AMD Unleashes Radeon R9 285 Graphics Cards, Tonga GPU, GCN 1.2 Architecture