You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
"Our first quarter results, combined with a variety of third-party insights from across the industry, and form our thinking for the balance of 2015. We expect the PC market to remain challenging, leading to a mid single-digit decline in the overall full year PC TAM."
The newly created Client Computing Group had revenue of $7.4 billion an 8% decline year-over-year, driven by a 16% decline in desktop unit volumes, partially offset by a 3% increase in notebook volumes.
Turning to the full year 2015. We expect revenue to be approximately flat to 2014, down from the prior guidance mid-single-digit percentage growth. We’re now projecting a mid-single-digit decline in the overall PC market.
Yes there will be contra-revenue for SOFIA. But it is Intel manufactured. You have to look closely at what Intel says in this area since they are deliberately vague. They called Sofia an integrated baseband and application processor. What they want you to think is that includes the modem but it does not. The modem is the part manufactured by TSMC and it can be kitted with Sofia but is a separate chip (this is not necessarily bad, Apple has a separate modem and application processor, for example).
Best error in the transcript is Brian talking about "Morris Law" where he presumably actually said "Moore's Law." Moore's Law is dead, long live Morris Law!
On the Altera deal, management only has two modes of operation. Either there is a deal, and it is official. Anything else is a rumor that they cannot comment on. If you asked Intel whether Facebook was buying them, instead of laughing at the very idea they would still have to say that they never comment on acquisition rumors. You can't not comment on ones that are false, even if ridiculous, because otherwise not commenting means there is substance to the rumor. Several analysts tried to get Intel to at least talk about their basic acquisition strategy but they wouldn't bite.
And another thing: it looks like their flash (with Micron) is really good and much denser than anyone else's for a few quarters. If they can get it to yield well, of course. They talk about SSDs all the time but most flash goes into mobile not SSDs, although SSDs are growing. So another way for Intel to hide weakness in mobile would be to take lower margins on flash and not call it negative revenue any more, but kit cheap flash with Sofia or their modem or whatever.
Don't ask questions about M&A
Don't ask questions about 10nm (still waiting for development "go" signal as far as I can tell)
No questions about foundry (I guess it's MIA)
Overall impression: They have to ride the ups and downs of a gigantic mature core market, which grows at the rate of the economy. The economy in the US is doing OK, but sputtering in Asia and elsewhere. So Intel bears the brunt of this. Other chip companies have a product mix that is biased toward younger, growth products--tablets and smartphones--and are in all likelihood going to grow much faster than Intel now and in the future.
IMHO, among Intel's four largest business units, their short term urgent task is to push Data Center Group (DCG) to grow fast enough to offset revenue decreased in the Client Computing Group (CCG). Because DCG is about half the size of CCG, this is not an easy task to achieve. The revenue for two other business units, Internet of Things Group and Software and Services Group, are too small to make enough impact on the declining CCG revenue.
1. Intel put their Non-volatile Memory Group and New Devices Group under the "All Other" category. It has been losing big money for the past 5 quarters. Intel is so proud about their leading edge technology in memory. But financially it's a drag of Intel's profit.
2. For the past 5 quarters, the operating loss from the "All Other" category is larger than IoT group's or Software & Services group's revenue! And other than 1Q2015, for each quarter in 2014, "All Other" category's operating loss was larger than its own revenue. It doesn't make sense and it's really bad.
3. Operating income from IoT group and Software & Services group are shrinking quickly despite their relatively flat revenue growth.
1. Intel put their Non-volatile Memory Group and New Devices Group under the "All Other" category. It has been losing big money for the past 5 quarters. Intel is so proud about their leading edge technology in memory. But financially it's a drag of Intel's profit.
2. For the past 5 quarters, the operating loss from the "All Other" category is larger than IoT group's or Software & Services group's revenue! And other than 1Q2015, for each quarter in 2014, "All Other" category's operating loss was larger than its own revenue. It doesn't make sense and it's really bad.
3. Operating income from IoT group and Software & Services group are shrinking quickly despite their relatively flat revenue growth.
That's a good point. In the conference call they say that IoT is doing well, but the financial statements tell a different story.
IoT is really what used to be embedded. I wonder what happened there that caused the embedded profits to nose dive.
That's a good point. In the conference call they say that IoT is doing well, but the financial statements tell a different story.
IoT is really what used to be embedded. I wonder what happened there that caused the embedded profits to nose dive.
It's quite strange when many news outlets or analysts were talking about Intel's earning conference call, it seems they don't really look into the actual financial statements Intel posted on their website. Let me post the link again here:
How can operating loss be larger than it's own revenue? Unless Intel is doing "Contra Revenue" sort of things on the "All Other" segment? Or is it because Intel has too many money losing projects or write-off that they decided to bundle them all under this segment?
Intel Custom Foundry was also not mentioned on the call, which I find interesting.
I also do not understand why the Altera CEO would not accept a $50+ offer for a $30+ stock. Just how does he think he is going to get his stock up that high when they fumbled at 28nm, 20nm, and it is not looking good at 14nm.
I have talked to dozens of people about this and not one could confirm or deny a deal was in the works. There are no secrets in Silicon Valley, absolutely.
I wonder about the future of that business. The key attraction was that Intel were way ahead of the other foundries, but now that is no longer true and it turns out, even if Intel is year ahead, their foundry business is a year behind. (Altera will deliver their first 14nm product a year after the first 14nm Core M part.)
Why pay for more for Intel's 14nm FinFET when you can get much cheaper 14/16nm Samsung/TSMC?
Intel Custom Foundry was also not mentioned on the call, which I find interesting.
I also do not understand why the Altera CEO would not accept a $50+ offer for a $30+ stock. Just how does he think he is going to get his stock up that high when they fumbled at 28nm, 20nm, and it is not looking good at 14nm.
I have talked to dozens of people about this and not one could confirm or deny a deal was in the works. There are no secrets in Silicon Valley, absolutely.
Intel put foundries business under the "All other" segment. The same segment includes Non-volatile Memory Group and New Devices Group. It has been losing money each quarter in 2014. This "All other" segment lost about $2.46 billion in 2013 and $3 billion in 2014. For 1Q2015, it lost $586 million.
From the Intel's statement and revised statement structure, this "All other" is the only operating segment losing money since 2013.
Is it because the foundry business losing big money or their memory business losing big money? Or both? No analyst asked this question during the earning conference.
I wonder about the future of that business. The key attraction was that Intel were way ahead of the other foundries, but now that is no longer true and it turns out, even if Intel is year ahead, their foundry business is a year behind. (Altera will deliver their first 14nm product a year after the first 14nm Core M part.)
Why pay for more for Intel's 14nm FinFET when you can get much cheaper 14/16nm Samsung/TSMC?
The speed at which the new processes are coming should also worry Intel. I'm told that Samsung will tape-out full products in Q4 (risk production) and TSMC is not far behind. That is one year after 14nm. Intel has not talked about 10nm but I can assure you that Intel Custom Foundry and specifically Altera will not be taping out 10nm anytime soon. Altera is still struggling with 14nm. The next ALTR and XLNX conferences calls will be interesting for sure.
I can't imagine ICF is losing big money. What is in the cost center other than wages? Certainly not CAPEX and such. I honestly do not know, anybody else?
Money is very easy to waste, trust me, I have seen it way to many times. An organization gets sloppy, especially a large mature one, and a billion is nothing. Tight use of resources and financial controls are usually the difference between competing and not competing. A percent here and a percent there and the profit quickly becomes a loss. Money is funny, valuation is in the mind of the beholder.
