There have been a lot of rumors about why Intel’s (NASDAQ:INTC) Ivy Bridge is slipping, but they are all spin to cover the real issues. Several sources are telling SemiAccurate that there are problems, but they are nothing to do with what is being reported.
Anyone watching the PC industry knows that Ivy Bridge is, and knows it has been set to come out in March 2012 for months now. The latest spin is that Intel is delaying the transition to that chip in order to save money, and it is complete bull. Why? Ivy is built on a new process, 22nm, is technically complex, and requires lots of new fab upgrades to accomplish. Intel is the only company close to doing something this complex, others are years behind. To make the transition, you need to upgrade equipment, install it, calibrate everything, and then it takes about three months for the chips to start rolling off the line.
If the plan was to launch in late 2011, lets say Dec 1, the wafers to make those chips would have to be in the proverbial oven right now. Add a bit to the three months needed to process the wafers, two weeks to dice, test, package, and put them in boxes, plus another few weeks for the OEMs to make the boxes and get the channel stocked, and you are at easily four months. Initial ramps are slow, so add a month or so to get enough volume for launch, and you need at least 5 months from initial wafers in to boxes you can put a ribbon on and give to a spoiled child that really wanted an XBox anyway.
Chips out in late 2011 means chips in by late 2011 minus 5 months, or July 1st, give or take a bit. That didn’t happen. We wrote that it was 2012 last March, others such as Zol.com.cn have published roadmaps saying as much for months now, and they are also backed by numerous others. There is no way in hell that Ivy was a 2011 product, and hasn’t been since very early 2011, period.
So why is there a big song and dance about the new ‘slip’ until March 2012? The only excuse we can think of is that it is either a plausible excuse for the financial community, basically something to ‘explain’ a slip that actually didn’t happen in a way that seems positive instead of negative. Since there actually hasn’t been a slip, this one doesn’t seem very sane.
Slightly more logical but less likely is that it is fabricated by someone with a lot of shares that wants to move the stock to a more profitable position. In this case, Digitimes is the patsy, and basically every tech site repeating the bull is a sounding board, chipping away at their credibility too. No one in the chain’s echo chambers seem capable of using Google to do the merest of fact checking before breathlessly regurgitating nothing. In either case, the ‘delay’ is, was, and alway will be nothing, the date has been March 2012 since at least very early 2011.
That isn’t to say there aren’t delays though, lately the rumor mill has been saying that there are fresh new delays to a post March timeframe. Some of these rumors say that Ivy Bridge will be delayed by about a month, other say it is closer to a quarter, probably Computex for the launch. Both sets of rumors point to the same problem, 22nm process problems, and both have one thing in common, they are not confirmed.
SemiAccurate asked Intel about the possibility of a post-March delay last week, the question asked was, slightly paraphrased, “Any changes to the Ivy schedule recently?”. Intel responded, “Nothing major that I know of, and even if there were not sure what we would or wouldn’t say about it.” Since the official launch date has not been announced yet, their response was completely understandable. Fair enough, and no source will give anything more specific than what we mentioned above.
That leads us to two separate conclusions. First, the March ‘delay’, isn’t, it is simply hot air backed by either ulterior motives, laziness, or abject stupidity. Ivy Bridge has been March, and semi-officially still is March for a long time now. The other conclusion is that problems still remain, and a longer delay is still quite possible, but still seems a bit fluid. Chill people, and do a search or two before you panic next time, you may not look as silly.S|A
Latest posts by Charlie Demerjian (see all)
- You can now buy bare Snapdragon SoCs with 410E and 600E - Sep 28, 2016
- Spin Transfer Technologies talks about their ST-MRAMs - Sep 27, 2016
- ARM adds CMN-600 interconnect and DMC-620 memory controllers - Sep 27, 2016
- Globalfoundries 7nm process isn’t even close to the name - Sep 26, 2016
- ARM upgrades realtime offerings to v8-R and adds Cortex-R52 - Sep 21, 2016