28nm Continues To Perplex

The 28nm issue continues to perplex. Mike Bryant was the first to point out, back at January’s IEF2012, that customers for 28nm at TSMC were having issues.

But, although Altera, Qualcomm and Nvidia have publicly complained about lack if 28nm supply from TSMC, Xilinx has said it has no issues.

“We have all the capacity we want and yields are ahead of plan,”  senior Xilinx vp Victor Peng tells me.

He points out that Xilinx uses TSMC’s HPL process which is not used by any of the companies which have complained about lack of 28nm supply.

Curiously Peng says that Xilinx has shipped “1,000s of units.” Thousands seems rather small bearing in mind that Xilinx started sampling 28nm over a year ago.

Asked if the company was comfortable with the supply situation for the future, Peng replied: “We’re comfortable with what we’ve got. We have what we need right now.”

The most virulent complainant has been Qualcomm which said it was actively looking for multiple alternative suppliers. Qualcomm CEO Paul Jacobs did not blame yields but cited lack of capacity for constrained supplies which impacted Q1 revenues.

Altera  has made no criticism of yields but said it could not get enough supply to meet an up-tick in demand late in Q1.

However Altera, which has always exclusively used TSMC,  is not going to add to its fab sources nor go back to the jointly-owned fab model  in Washington State which  it used to have with TSMC.

Asked if going back to the shared fab model, or adopting a multiple foundry strategy, might be in Altera’s planning, Altera’s senior vp for military, industrial and computing operations, Jeff Waters, tells me: “One of the benefits of sticking to TSMC is the early access we get to new processes, another is the long-term support we get for our process technologies. If you’re hopping around from fab to fab, from foundry to foundry, it’s difficult to get a foundry to support you. And TSMC gives us  support.”

Jen-Hsun Huang of Nvidia is the only CEO to publicly accuse TSMC of having poor yields on 28nm and also announce that he’s off to Samsung to get 28nm fab. However some people have pointed out that Nvidia’s yield problem could be down to design flaws.

No one expects process transitions to go smoothly, but there had been a hope after the horrors of 40nm, that lessons learnt during that node would translate into a better experience at 28nm.

The jury is still out.



  1. EUV is not a requirement for 14nm at Intel. It would be nice but not essential. Despite not being linked, we expect the insertion point for both EUV and 450mm wafers to be at already proven 14nm processes for pilots but moving quickly to new 11nm EUV only processes for production.
    As for stopping Intel, the “rest of the industry” you mention is just dreaming. Intel’s lead will expand progressively over the next decade and a half with only Samsung’s memory processes keeping up. The others are simply not investing enough in advance process development.

  2. me 2 scunnerous

  3. I suppose the good thing for TSMC is that they’re not short of clients. If you look back Daniel had a real go at several people, including Mike, on the shutdown issue which “never happened”… “MANURE”. I note that Qualcomm has toed the line on the official TSMC “capacity constrained”, normal ramp story whereas nVidia has quite blatantly used the Y-word – tsk, tsk.
    In previous articles he does a good job of hinting that nVidia’s woe is self-inflicted, without actually saying it. The expression “design-limited yield issues” is a err, nice way of putting it in “TSMC 28nm Yield Explained!” on March 4. nVidia is not mentioned in that article but it’s cross-referenced in another of his articles.
    I dunno what to think but I’m guessing that the rest of the industry is hoping 14nm is going to stop Intel in its tracks, or at least delay them, so they can catch up: Intel is saying 2014 but EUV is said by others to be for 2015, with others even saying never. 14nm without EUV is a similar enigma so it’s all very murky to me.

  4. Ah yes Scunnerous on SemiWiki. Daniel seems to have set himself up as defender-in-chief of TSMC. But Jen Hsun Huang of Nvidia, who I gather is an old mate of Morris Chang, hasn’t held back on criticizing TSMC and Qualcomm has been a little acerbic and is looking elsewhere for fab and Altera says it has had short supplies and, as they say, the customer is always right!

  5. Hmm, I see that Daniel Nenni has had another poke at Mike Bryant, among others, on the TSMC (non-)shutdown. And a little gloat on the announcement that “TSMC 28nm Q1 revenue was 5%”.

  6. Since Xilinx’s top of the range chips make
    Intels seem relatively cheap, perhaps “thousands”
    at $1000+ each is correct.

Leave a Reply

Your email address will not be published. Required fields are marked *