TI’s lowered guidance, announced at the company’s midquarter update Wednesday, comes after Nokia last week sharply lowered its sales forecast. The Finnish handset maker, a big customer of TI chips, has been struggling to keep up in a smartphone market increasingly dominated by rivals like Google Inc. and Apple Inc.

Read the rest of this post on the original site »

The Dallas-based company had already issued warnings about the impact on its two factories in the quake zone. In reporting first-quarter results Monday, TI put costs associated with the disaster at $30 million, which reduced earnings per share for the period ended in March by about two cents. In addition, the company suffered about $20 million in lost sales, said Ron Slaymaker, TI’s vice president of investor relations.

Read the rest of this post on the original site »

Here’s a sampling: Fairchild Semiconductor is up nearly 5 percent in after-market trading. Analog Devices is up a little more than 4 percent. Maxim Integrated Products, 5 percent. Linear Technology about 3 percent. On Semiconductor more than 5 percent. Intersil, 8 percent. STMicroelectronics, 3 percent. You get the idea: Investors are clearly hoping that another acquisition is soon to happen in the world of analog semiconductors.

TI was already the big gorilla in the analog business. It’s a $42 billion business and TI accounted for about $6 billion worth of analog sales in 2010, and the analog business amounted to 43 percent of TI’s business last year. With this acquisition it goes from being a chip company that supplies the microprocessors for most of the world’s wireless phones to a mostly analog company. A Gartner market ranking of analog chip makers ranked TI number one and National fourth. The combination makes TI an even bigger gorilla yet.

Who’s number 2? Analog Devices, ahead of Maxim and Linear Technology, who are numbers three and five respectively in the Gartner rankings. Further down the list you see names like STMicro, On Semi, Intersil and Fairchild.

So what’s so great about analog chips? They go into everything: From wireless phones to games to medical devices to military and industrial gear. They manage power; help convert light and sound into digital information; and measure temperature or pressure or any other quantity that occurs in the real world. And with all the digital gadgets we rely upon these days, we’re relying increasingly on analog chips of every stripe. Analog sales grew 37 percent last year, which was faster than the chip business as a whole, Gartner analyst Steve Ohr told me.

As is the case with TI and National, there’s bound to be a lot of overlap in products and customers among all these other companies. That suggests that more combinations are likely, though with valuations inflating on the speculation, any company thinking of making an acquisition would probably want to sit tight until the prices come down a little bit before making a move.

In a company statement announcing the deal, TI CEO Rich Templeton described the deal as being about “strength and growth.” He said National had done a good job boosting its profitability and reining in expenses, and that the combination will, upon the close, boost TI’s earnings per share.

TI specializes in wireless chips sold into mobile phones while National specializes in analog chips. TI is in the analog chip business too and has 30,000 analog products in its portfolio, which will now be supplemented by the 12,000 analog products that National sells. In one go, analog products will become more than 50 percent of TI’s business.

Another key factor: National’s manufacturing prowess. It owns chip factories–usually referred to as “fabs”–in Maine, Scotland and Malayasia, and TI says it will continue to operate them.

Terms of the agreement call for National shareholders to receive $25 in cash for each share of National common stock they hold at the time of closing. TI expects to fund the transaction with a combination of existing cash balances and debt. As of December 31, TI had about $3 billion in combined cash and short-term investments, and no long-term debt. TI also had access to a variable-rate revolving credit facility that gives it access to a combined $1.9 billion until August 2012; I’m presuming that the remaining $1.5 billion or so will be borrowed. National also has $879 million and change in cash on its books, but a little more than a $1 billion in long-term debt.

Investors seem cautiously optimistic about the deal. TI shares are up slightly in after-hours trading. National shares are naturally soaring by 70 percent to catch up with the valuation of the deal. One other company whose price I just checked is that of Fairchild Semiconductor. It’s a rival to National in the analog and power-management business. Its shares are up 94 cents–or more than 5 percent–after hours, and there’s no mistaking the speculative hope that it could be the next takeover target.

As usual, the idea behind the teardown is not only to figure out who Nintendo’s component suppliers are and what parts are being used, but to estimate how much all the components cost to help guess how much of a profit margin Nintendo is making on each unit. And it looks like a decent margin. ISuppli says the cost of all the parts in the device itself plus what’s in the box amount to $103.25 for a device that’s selling at retail for $249. The cost works out to an increase of about $25 over the Nintendo DSi, the most recent Nintendo handheld, released in 2009, which cost about $78, when iSuppli tore it apart that year.

While most of the components come from Japan, it’s not entirely clear if the supply of any of the parts used come from areas affected by the earthquake and tsunami, says Andrew Rassweiler, an iSuppli analyst who supervised the teardown. “Many of these component should have a greater risk exposure to supply chain problems, though we don’t know about any specific disruptions at this point,” he said.

The most expensive component, as is often the case with consumer electronics, is the displays. The 3DS uses two Sharp displays that cost a combined $33.80. The headliner is the top screen 3D. It’s a 3.5-inch 800-by-240 pixel display that uses an LCD-based parallax barrier panel sandwiched to the back of the color LCD which alternates between the left and right images at a high rate of speed to produce the 3D effect. “It looks like a conventional LCD from the outside, but when you open the display you see that on one side of the glass is essentially the conventional color element, and on the other side of the glass is a monochrome element,” Rassweiler told me. “It’s a clever bit of display engineering.”

The handheld’s main chip is an applications processor. It’s a custom ARM-based chip manufactured by Sharp, that at a cost of $10.02 is only slightly more expensive than the chip in the previous Nintendo DSi. However, Nintendo has quadrupled the amount of flash memory in the 3DS versus the DSi to 16 gigabytes, and Samsung, the world’s largest manufacturer of flash, supplied it. Fujitsu supplied another type of memory known as fast-cycle RAM. Rassweiler says for this particular type of memory, Nintendo has used a type of chip that’s only made by Fujitsu, which is odd because FCRAM is widely available, and its unusual for consumer electronics manufacturers to “single source”–that is, rely upon a single supplier for an important component. The combined cost of memory on the 3DS worked out to $8.36, more than twice the cost of the memory found on the DSi.

Three chips related to the user interface cost a combined $6.81: an accelerometer from STMicroelectroncis, a gyroscope from Invensense, and an audio chip from Texas Instruments. Atheros, the Wi-Fi chipmaker that’s being acquired by Qualcomm, supplied a $5 Wi-Fi chip. TI and NEC supplied power management chips that cost $3.63. The 3DS contains three cameras, and though it’s not clear who supplied them–camera suppliers have gone to great lengths to hide their identities in recent years–iSuppli reckons their combined cost at $4.70.

Since I often get asked this question, let me say that iSuppli’s analysis focuses strictly on the materials used and doesn’t account for the cost to develop software or to license any patents. Nor does it account for the cost of any shipping or distribution or marketing. It’s just the raw cost of the hardware.

Japan in 2010 accounted for 35 percent–or $11 billion worth–of the $31.5 billion global market for automotive infotainment electronics, iSuppli says. On top of that, Japan is responsible for about one-third–$7.3 billion–of the $23 billion market for chips used in cars overall. Aside from chips, Japanese companies produce LCD panels and optical sensors used to make in-car systems.

ISuppli says Renesas Electronics, Texas Instruments, Freescale Semiconductor and Fujitsu, all of which supply components to the auto industry, have all been affected by shipping problems and difficulty in obtaining raw materials. The problems could last weeks or months.

Problems like this aren’t just hitting infotainment systems. As The Wall Street Journal reported today, shortages of a single electronic part made by Hitachi Automotive that measures airflow in car engines have forced companies like General Motors, Toyota and PSA Peugeot-Citroën to cut their output at auto plants in the U.S. and Europe. The plant that makes the component is located to the north of Tokyo and has been shut down. Hitachi makes about 60 percent of the world’s supply for this type of part. The Journal said Toyota on Wednesday warned employees to expect a production halt at some plants in the U.S. and Canada.

Meanwhile, Japanese automakers have stopped production at several plants in order to conserve electricity following the loss of generating power from the troubled Fukushima Daiichi nuclear plant. There has been a 10 percent reduction in electrical capacity, iSuppli says. Additionally, companies like BMW, Volkswagen, Continental and Bosch have removed their expatriate employees from Japan.

Separately, Dow Jones reported yesterday that ZTE Corp., a Chinese maker of networking gear, is suffering quake-related supply problems. A company exec said it expects the problems to last as long as six months.

Sony Ericsson

Wireless handset maker Sony-Ericsson just issued a statement on the extent of disruptions it expects.

“Although the full impact of the current situation on our business will take additional time to assess, Sony Ericsson anticipates disruption to its supply chain operations,” the company said in a statement. “As part of our business continuity plan, we are in contact with all our key suppliers in the region and we are identifying the possible relocation of certain component manufacturing, and looking at secondary sources of supply.”

Intel and Qualcomm

Some analysts had speculated that plant shutdowns by Mitsubishi Gas would slow chip production by cutting supplies of certain chemicals used in the chip making process. Both Intel and Qualcomm told Bloomberg News that they’re careful to avoid situations where important supplies come from only one source.

Texas Instruments

Wireless chipmaker Texas Instruments said on Monday that a plant in Miho, about 40 miles northeast of Tokyo, had suffered “substantial damage” and that it may be July before the plant is back up to full production. The plant is responsible for about 10 percent of TI’s overall production, and about a third of its capacity is devoted to its Digital Light Processor, and the rest to analog components. Shares of Diodes, Inc., a TI rival, jumped when analyst John Vinh of Collins Stewart said the company stood to benefit from TI’s troubles.

Oracle

Analyst Derrick Wood of Susquehanna Financial said in a note to clients that fears about Oracle’s exposure to the Japanese market, which has driven the shares down in recent days, are probably overblown. Even though Japan accounts for about five percent of Oracle’s revenue, he said, fears are “likely overdone.” Most of Oracle’s revenue from Japan comes from recurring maintenance fees, so the risk of a serious hit to sales is minimal.

Raw Materials

The Wall Street Journal notes that Japan supplies about 90 percent of the world’s bismaleimide triazine, an important material used in making printed circuit boards for wireless phones. Japan also supplies much of the world’s silicon wafers that are used to make chips.

(Map via Wikimedia Commons)

But the teardown that Wall Street and the investment community is waiting on is the one from the market research firm IHS iSuppli, whose team spent all day Saturday in a furious effort to dissemble a 32-gigabyte iPad 2 and estimate the cost that Apple paid for every component. They gave me an exclusive early look at their findings.

The point is to form a partial picture of the gross profit margin on every unit, a figure that Apple generally keeps to itself. This information is useful to investors and analysts who then factor the findings in with other assumptions they use to predict how much of a profit Apple is going to report over the next few quarters.

The headline of iSuppli’s teardown researcher is always the estimated bill-of-materials cost, which is the sum cost that it thinks Apple has paid for all the hardware inside the iPad 2. It doesn’t take into account the cost to develop software, or other things like packaging, shipping and distribution, or manufacturing.

In this case the estimates are for the 32-gigabyte, 3G version of the iPad which sells for $729, and there are two estimates, one for the AT&T version–$326.60, and one for the Verizon Wireless version–$323.35. Some of the wireless chips used in the AT&T version are a little more expensive or require an extra part. For example, on the Verizon version, GPS is integrated with the Qualcomm-made wireless baseband chip. On the AT&T version, an extra GPS chip had to be added along with the Broadcom-made Bluteooth and Wi-Fi chips, adding an extra cost of $1.50 per unit.

The baseband wireless chips were naturally different because AT&T and Verizon use different wireless technologies. Intel, the new owner of the former wireless chip division of Infineon, supplied the main wireless chip in the AT&T version, with supporting chips coming from TriQuint Semiconductor and Skyworks for a combined cost of $18.70.

Qualcomm supplied the main wireless chip Verizon version, with supporting chips coming from Skyworks, Avago Technologies, and Murata for a combined cost of $16.35. While there had been some speculation that Apple had used a Qualcomm chip in both versions, but it turned out not to be the case.

Aside from the wireless chips, the components are otherwise identical across both versions. Both sport Apple’s A5 chip, and iSuppli says that Samsung is still manufacturing it for Apple at a cost of $14. While there had been some talk in recent weeks that Apple was moving its chip manufacturing contract to Tawain Semiconductor Manufacturing Corp, there’s no evidence that it has made such a move, at least not yet.

The most expensive component by far is the touch-sensitive display, coming at $127. ISuppli says that the LCD portion the unit they tore apart was built by LG Display, but Apple is known to use other sources for displays, including Samsung, and possibly ChiMei Innolux. The glass assembly covering the display is thought to come from TPK or WinTek. ISuppli says costs on the display are going up because manufacturing yields on LCDs have been lower. Apple is also thought to be using a more expensive glue to improve the efficiency of the process of bonding a new thinner type of Gorilla glass to the display.

Samsung supplied Apple with the NAND flash memory used in the iSuppli sample, holding on to a relationship that goes back several years to the days of the first iPod nano, though Toshiba is also known to supply Apple with flash. It is the world biggest consumer of flash memory, after all. Elpida supplied the DRAM memory. ISuppli estimates the combined cost of memory, both flash and DRAM plus a Micron-made MCP memory chip at $65.70.

Then there’s a set of components seen in the iPad 1 that remained the same in the iPad 2. STMicroelectronics supplied the gyroscope and the accelerometer, and AKM Semiconductor supplied the electronic compass. Broadcom supplied touch interface chips, while Texas Instruments supplied a touch screen driver chip. Analog Devices supplied a capacitive touch controller.

Finally there are the two cameras. ISuppli hasn’t yet named the suppliers there, though the usual candidate is Aptina, the former camera unit of Micron, though it’s possible that Apple sources them from more than one place.

ISuppli’s estimates are a lot higher than the findings of another teardown shop, UBM Techinsights. The Wall Street Journal reported that UBM’s cost estimate is about $270, but that estimate was made before it conducted its actual teardown, and didn’t change once it had.

Despite the headwinds, Intel delivered on its quarterly results. Profits were up 48 percent year-over-year, and Intel set all-time records in sales and profits both for the quarter and the full year.

As he pointed out in his prepared remarks in today’s quarterly earnings conference call, sales at Intel’s Data Center Group grew by 15 percent in the fourth quarter and finished 2010 up 35 percent. What’s driving such healthy growth? The Internet, of course. Its growth is causing a mushrooming demand for servers of all kinds, which means demand for Intel server chips is going up right along with it.

Traffic on the Internet was 245 exabytes in 2010, he said–greater than all previous years combined, and within five years another billion people getting online will push that to over 1,000 exabytes a year.

(What’s an exabyte? 1,024 petabytes. What’s beyond the exabyte? The zettabyte. Incidentally, Royal Pingdom released some interesting numbers about the size and scope of the Internet today, and they’re worth reading.)

Hear the snippet from Otellini’s remarks below.

Later, in response to a question from Patrick Wang of Wedbush Securities, Otellini talked about where enterprises are in their PC refresh cycle, one of the few bright spots for PC sales that Gartner and IDC mentioned in their sales estimates yesterday. Reiterating an earlier comment that he expects PC sales to grow at a rate in the “low to mid-teens” in 2011, Otellini said he thinks many corporations haven’t yet updated to using Microsoft’s Windows 7, and so the upgrade cycle hasn’t yet, in his view, reached its midpoint. Then Intel CFO Stacy Smith jumped in with some comments about where cloud computing figures into sales of server chips. Another two-minute audio highlight is below.

Also asked during the call by Shawn Webster of Macquarie Research about Microsoft’s announcement that it will create a version of Windows for ARM-based chips from Qualcomm, Texas Instruments and Nvidia, Otellini said it didn’t seem to him like “new news,” and that versions of Windows for chip architectures other than Intel’s x86 family have been around before. In this 90-second audio clip he talks about the potential pluses and minuses of Microsoft’s move for Intel.

Intel said in its statement that PC Client Group revenue was flat. Other groups were stronger: Data Center Group revenue was up 35 percent, and Intel’s architecture group saw sales surge by 27 percent. Intel Atom microprocessor group, its low-power chip aimed at tablets and smartphones, saw revenue grow 8 percent. This despite word from Microsoft last week at the Consumer Electronics Show that it will develop a version of Windows for ARM-based chips from Texas Instruments, Qualcomm and Nvidia aimed at tablets and smartphones. Microsoft’s move calls into question Intel’s hopes to land design wins for the Atom low-power chip that it hopes to sell to manufacturers of smartphones and tablets, but which has yet to show any significant results.

This report of flat revenues for PCs comes a day after Gartner and IDC both said they saw weaker-than-expected sales of PCs in the fourth quarter of 2010.

For its 2011 first-quarter outlook, Intel said it expects revenue of $11.5 billion, plus or minus $400 million, and a gross margin of 64 percent, plus or minus a couple of points. For the full year, it expects gross margins to be 65 percent, plus or minus a few points. More after I go through the numbers and attend the conference call, which starts in about an hour.

First, the departure has jarred the confidence of investors who have pushed the value AMD stock up by more than 57 percent since September. Shares are down by more than 8 percent today.

Second there’s the problem of hiring a replacement for Meyer, who had been on the job only a little more than two years. I’ve been talking to people both inside AMD and longtime AMD watchers outside of the company and practically all of them have been having trouble coming up with a short list of potential candidates.

For one thing, I’m hearing from people familiar with the thinking of those involved in the hiring process that there’s a strong preference for an external candidate.

Among the criteria are someone with a proven record of running large technology companies, and one with some charisma who can get the marketplace excited about AMD again. While Meyer deserves credit for getting AMD back on relatively stable footing following the divestiture of its manufacturing operations–now GlobalFoundries — and his predecessor, Hector Ruiz, gets the credit for doing the heavy lifting of getting the complicated transaction related to that split, neither could be described as charismatic.

Historically, AMD knows what it’s like to have a charismatic CEO. Jerry Sanders who founded the company and ran it from 1969 until 2002, possessed plenty of it, and some of the more colorful anecdotes about Silicon Valley history concern him. The board wants someone who’s both capable and cool at the same time. Someone who can represent the company well to the outside world, bring an air of stability and competence and elaborate a vision that will move the company forward. That’s a tall order for a company like AMD, whose fundamental strategic problem can be summed up in a single phrase: Competing with Intel is brutal, no matter what you do.

The list of potential candidates isn’t obvious by any stretch. Still in my conversations today, a few names came up, some more idealistic than realistic. One internal candidate who will probably get courtesy consideration I’m told is Rick Bergman, senior vice president and general manager for AMD’s products group. He oversees both the graphics and microprocessor operations and came to AMD in 2006 as a senior executive at ATI, the graphics chip company that AMD acquired for $5.4 billion in 2006. His résumé includes time at Texas Instruments and IBM. He’s described by those who know him as hard-driven and competitive and a capable well-respected manager, though at the end of the day not likely to get the nod.

Another name that has come up is that of Pat Gelsinger, not necessarily because he’d be a candidate for the job, but more as an example of the kind of person AMD would like to hire. Gelsinger was Intel’s CTO from 2001 to 2005 and was senior corporate vice president for the Digital Enterprise Group until 2009, when he suddenly jumped to EMC as president, COO and apparent successor-in-waiting to CEO Joe Tucci.

Finally there’s Michael Capellas, whose name invariably comes up whenever a significant CEO slot comes open. He’s currently running Acadia, a private cloud computing joint venture between Cisco Systems and EMC with investments from Intel and VMWare. Capellas was the CEO of Compaq Computer when Hewlett-Packard acquired it in 2002, then went on to helm MCI and engineered its turnaround and sale to Verizon in 2006. His next stop was the payment giant First Data after it was taken private in a leveraged buyout by the private equity fund Kohlberg Kravis Roberts. He was considered for the CEO job at Hewlett-Packard, before Léo Apotheker was named, but was said not to be interested. He’s got the tech and management chops and has a proven record for getting troubled companies on solid footing. It’s unclear if he would be interested.

Whoever they pick, they may want to do it quickly. AMD has a tough road ahead of it, and uncertainty at the top certainly isn’t going to help.

Shares in ARM have nearly tripled in value from this time a year ago, and the most recent surge occurred in December, when the first reports emerged that Microsoft would do something that previously seemed almost unthinkable: Create a version of Windows designed to run on chips other than the x86 chips from Intel and Advanced Micro Devices. Microsoft confirmed the news two days ago. If 2011 is going to be the year of the tablet, then chances are it’s going to be the year of ARM chips.

The funny thing is, practically every year has been a good year for ARM chips. They’re so widely used already that there’s a good chance you use them, probably several of them, every day. During its most recent quarter, more than 900 million ARM-based chips were sold in mobile devices like smartphones and tablets, while another 600 million were used in devices as varied as TV, toys, cars, alarm clocks and remote controls.

ARM doesn’t build the chips itself; it designs the cores–or central brains–used on those chips. I like to compare it to selling a basic cake recipe. If you’re a baker whose expertise is making really great frosting, why bother dreaming up a brand-new cake recipe when you can use an existing one, and instead use your time and effort to make great frosting? A lot of semiconductor and electronics companies have reached the same conclusion, and paid to license ARM’s recipes for chips, and then built their own custom enhancements around the ARM core.

It’s a pretty popular recipe. The company issued more than 700 licenses as of last year to some 250 chip companies, which then turned around and sold the chips to more than 1,000 manufacturers. ARM estimates that in 2009 four billion chips based on its designs were sold, and that more than 20 billion have been sold in the two decades since the company launched.

Aside from the three ARM-based chips from Texas Instruments, Nvidia and Qualcomm that Microsoft demonstrated running Windows as part of CEO Steve Ballmer’s Jan. 5 keynote presentation at CES, the list of companies using ARM includes Samsung, Broadcom, Toshiba and scores of smaller chip companies.

ARM also has an interesting history. It was founded as a joint venture between Apple and a British outfit called Acorn Computers in 1990. Apple’s interest was to create and develop a chip that would run the Newton, and spur the development of a new-age handheld computer that the Newton was supposed to bring about. (As a few commenters note below, the first ARM chips were used in desktop computers sold primarily in the U.K.) The Newton went nowhere, but the vision for ARM as the chip of choice for mobile computing was right on the money. ARM chips from Motorola (now Freescale) landed in devices from Palm and early handhelds running Windows Mobile. ARM flourished and went public on the London Stock Exchange in 1998. Between 1998 and 2004, Apple sold off its ARM shares for combined proceeds of almost $800 million.

Now having built a considerable lead in the wireless world, ARM-based chips look awfully strong as the battle over tablets shapes up. And beyond that lies higher-end computing opportunities like servers. Some think Intel should be worried. Despite this week’s launch of its Sandy Bridge generation of PC processors, Intel’s shares are trading lower today than they did at the start of the week.

I caught up briefly with ARM Executive Vice President Antonio Viana by phone from CES to talk about the year ahead for ARM.

There’s been a lot of attention around ARM coming into the Windows fold, and everyone knows it from its strength in the wireless devices. How is 2011 shaping up for ARM?

We got our start more than 20 years introducing a chip architecture aimed primarily at the mobile industry. We offered a chip design that’s efficient in the way it consumes power. What happened was the technology moved beyond the cellphone: Into the home, cars, printers. And that trend is continuing. Consumers want features that require a lot more computing power. Some of these devices are handhelds, some aren’t. What makes the ARM architecture central to all that is that industry brings their own secret sauce, their own pieces to the table. The development with Microsoft is just a small microcosm of that.

Are there new licensees coming on?

Our roadmap is constantly evolving, and we’ve developed the architecture for a pretty broad set of use cases. We license to companies like NXP that are relatively simple 8- and 16-bit microcontroller chips that go into industrial equipment, or meters or toys. But because of the network connectivity requirements that are starting to come to those devices, you’re starting to see some of these move to more versatile 32-bit chips and the costs are manageable because developers are so used to working with ARM. Then if you swing way out to the other extreme we just launched our A15 architecture. That’s a multicore design, and it’s finding its way into next-generation servers.

Who’s using that?

Unfortunately, we can’t say yet. A15 was announced last year. We think we’ll start seeing production silicon in the latter half of next year, and there will be samples before that. When you start seeing samples then the partners working with it will start announcing products.

Obviously Intel has its Atom processor, which it has aimed at tablets and handhelds and many other market segments you’re involved in. What kind of competitive threat do you see from Intel?

The competitive threat is certainly there–x86 is incredibly robust and it has the incredible capital resources of Intel behind it. ‘Nuff said. Intel will be successful in various markets they go after. We’d be fools not to acknowledge that. But the question is who’s going to grow more? Who is going to leverage off the market trends right now? Tablets are a wonderful example of that. Right now about 90 percent of all tablets in the marketplace are ARM-powered. At a show like CES you see a lot of things that indicate the market trends. You always have to take a step back and wonder which of the things you see may never happen. But the trends are usually accurate. One of those trends is for always-on, always-connected power-efficient devices. When you look at it that way I’m pretty comfortable with ARM’s position.