- Intel's official stance is that its next generation 14-nanometer "Broadwell" processor is "on track" to go into production in 1Q 2014
- While this does not confirm/deny the Digitimes rumor that "Broadwell" will launch in 4Q 2014, this refutes the recent SemiWiki piece claiming further "yield problems"
- No word on the current plans for Cherry Trail, the first 14-nanometer Atom
In my recent piece, "Intel: Is This Rumor True?", I had commented on two related, but separate, rumors. They were the following:
- Intel's (NASDAQ:INTC) next generation "Broadwell" processor was delayed to a Q4 2014 launch
- Intel is experiencing further "yield problems" beyond what the company had commented on at its Investor Meeting
Shortly following my piece, I spoke with Intel representatives and the following was their official stance on the matter:
We continue to make progress with the industry's first 14nm manufacturing process and our second generation 3-D transistors. Broadwell, the first product on 14nm, is up-and-running as we demonstrated at the Intel Developer Forum in Q3'13. We're now planning to begin production this quarter with shipments to customers later this year.
Now, this addresses point (2), refuting the claim that Intel has run into any further "yield" problems, but the interesting thing is that this doesn't actually conflict with the Digitimes rumor. Indeed, the rumor claimed that Broadwel will be available in Q4 2014, which still lines up perfectly with what Intel has been claiming about Broadwell ever since the initial delay - that it would be a 2H 2014 launch.
In order to understand why a Q4 launch is likely, I'd like to take you through a sort of "guided tour" of how the launch of Intel's "Haswell" went during 2013.
Understanding Haswell's Timeline And Applying It To Broadwell
Look at the following chart to understand what the "timeline" for a product ramp from Intel looks like. I will use "Haswell" as the example:
Note that the pre-qualification builds began in the Q3 2012 timeframe, got much more aggressive in Q4 2012 and Q1 2013, and then the chips finally "qualified for sale" throughout Q2 2013. Note that Intel was in "volume production" of Haswell with those initial "pre-qualification" builds. Note also, that chips that were began being built in late Q3/early Q4 weren't available to the general public until June 2013 (and the Ultrabook ones, outside of the MacBook Air, weren't in systems until even later).
So, we can use this information to sort of put together a timeline for what the "Broadwell" ramp is likely to look like:
As you can see, the pre-qualification build of "Broadwell" likely begins in March 2014 (i.e. last third of Q1) and the qualification for sale/availability is likely to be late Q3 2014 (think high priority customers such as Apple), with broader availability (non-Apple) in Q4 2014.
I believe that Intel will have Broadwell "in production" in late 1Q 2014, which suggests product availability (in systems, on shelves) in Q4 2014 for non-Apple customers and likely late Q3 for Apple. So, while SemiWiki's claim that Intel is experiencing additional yield problems is unlikely to be true given the statements from Intel itself, the reality is that Digitimes is likely correct in claiming that Broadwell won't hit the shelves until 4Q 2014.
The real question - and one I couldn't get an answer to - is whether Cherry Trail (the 14nm Atom for tablets) will make it out this holiday selling season or not. My guess is as good as yours, but given that Intel will need to answer NVIDIA's Tegra K1 on the Windows side of things (Bay Trail is not good enough), I would hope that Intel makes that Q4 schedule with both Cherry Trail and Broadwell-ULX (for tablets).
Disclosure: I am long INTC, NVDA. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.