GLG News by James "Jim" Cantore

Implications: 1. New Markets will be created, enabled by the 450mm revolution. 2. Opportunity is there for new equipment strategic alliances with major chip players. 3. Smarter, simpler electronics projects need very intelligent, very large 450mm wafer-based chips.

Analysis: If we believe a survey done in 2007, the expected arrival date of 450-mm wafers fabs may be 2013. Then nearly 40 percent said 450-mm wafers would never happen. Now we read that Intel, Samsung, TSMC want 450-mm wafer pilot lines in 2012. These companies intend this target date as the start for 450-mm pilot line, not high-volume production lines.   Just as in the pre-300 mm fab era, many now doubt whether 450 mm fabs will appear. They feel that the risks and costs of 450 mm overshadow any benefits. Semiconductor manufacturing equipment vendors have called 450mm fabs flawed and misguided. The risks are that 450mm fab tool development costs could run as high as $100 billion for semiconductor manufacturing equipment vendors. Leading companies that make semiconductor manufacturing equipment, have said time and time again that investments in 300-mm manufacturing equipment have yet to be recouped. They believe that moving to 450 mm wafer size would yield little or no cost benefit.  However, Intel, Samsung, TSMC and Toshiba agree that a 30% percent reduction in cost per wafer over time will be achieved by 450-mm fabs. This cost reduction may expand many electronics markets, notably consumer, and many emerging geographical regions. Some experts challenge assumptions that 450-mm equipment will have the same throughput as an existing 300-mm machines. Their models for such fab equipment as etch, ion implantation, inspection and lithography challenge throughput assumptions. Assuming there are 400 dice per wafer, a 300-mm tool with a throughput of 150 wafers per hour could process 60,000 die per hour. Larger 450-mm wafers would more than double the die count to 920. Thus, a 450-mm tool with the same wafer throughput could process 138,000 die per hour. That is where models fall apart, claim equipment manufacturers. They fear that larger 450 mm wafers, having much finer than 300 mm geometry, processed on 450-mm tools will process but 74,400 dice per hour. That is a far cry from 138,000 die per hour. Thus chip makers would need to buy twice as many costly 450-mm tools. Further clouding sales prospects is that equipment vendors may have a customer base of just 15 companies who will build leading-edge 450 mm fabs compared to the 27 who now own 300 mm fabs. Who will drop out at 450 mm?  

Implications: 1. AMD got itself in a precarious position by failing to execute. 2. Nvidia acquiring AMD is not the solution. Nvidia does not need an AMD headache. Neither aspirin nor anything stronger will help Nvidia. 3. AMD needs financial and engineering resources.

Analysis: Nvidia buying AMD makes no sense." Financing the transaction might be problematic in today's tight lending market, and the assumption of more than $5 billion in AMD debt - Nvidia is not dumb. What sense does it make to take on AMD's manufacturing problems and also become a direct competitor of Intel Corp. AMD has its share of difficulties and remaining indigestion since its acquisition of ATI. AMD-ATI lost significant graphics market share to Nvidia. Had ATI remained independent, it would be mentioned today as a potential buyer of AMD. ATI was purchased for $5.6 billion in cash and stock in October 2006, NVDA won't do any better Nvidia lacks enough experience in integrating any major rival. It would have to resolve major architectural platform issues not just with AMD but also with the former ATI. Trying to deal with AMD's Fusion could be another Barcelona product execution bust. Totally dependent upon foundries for its wafer supplies, Nvidia has never had to manage its own wafer fab with its own people, let alone another company culture. An Nvidia-AMD-ATI's chances are like those of someone who urinates at the feet of the Roman army and expects that they will avoid being crushed - in this case by Intel. Intel, with its vast resources will maintain a long term processor market share of 80% or greater against any other processor player. NVDA does not need an AMD/ATI headache.

Implications: 1. The NOR market is a money loser, no matter who the players are. 2. If Numonyx, (Intel + STMicroelectronics), was such a great money making proposition it would have been funded months ago. 3. The financial world is well aware that the last big NOR Flash venture between AMD and Fujitsu, Spansion, resulted in a public company that to date has never ever made a profit.

Analysis: At the end of December 2007 the news was "Amid turmoil in the debt capital markets, Intel, Francisco Partners and STMicroelectronics have delayed the closing of Numonyx, its joint flash memory venture, to March 28, 2008."
Wait a minute. Is there an elephant in the room? Could it be that gross margins in the NOR Flash market are so poor that all players are losing money? Could it be that there is too much NOR Flash capacity? Could it be that about 67% of all NOR Flash sold is consumed by cell phone manufacturers? 
Could it be that NAND Flash is pushing NOR Flash out of cell phones? Could it be the relentless NOR Flash price pressure from the big 5 cell phone manufacturers? Could it be a price war that was instigated by Samsung because it wants to be number 1 in NOR Flash before 2010?
Could it be the big mobile network operators' price pressure on the cell phone manufacturers? Or is it the pressure coming from worldwide demand for lower price cell phones? Is it the credit crunch?
No, these are all of the elephants in the room. Smart money did not get that way by investing in money losing markets.

Implications: Does AMD keep close enough tabs on the details? Risk underestimation cripples bold initiatives. Crisp execution and coordination of all the risks is something AMD needs to do better. AMD overpaid for ATI, ran short of cash and had to go overseas for capital.AMD will give up precious market share to Intel.AMD profitability remains just out of reach.At current market valuation AMD may be a takeover target.  

Analysis: AMD traditionally over promises; such were its claims to have the first quad-core processor. Sometimes such tactics work. Now AMD works feverously to beat market expectations. AMD consistently has been very bullish; now it finds itself running from the bulls of Barcelona. Barcelona came up short on critical technical details. Its more conservative competitor Intel is now locking up key designs with its own quad-core processors and rapidly regaining market share from AMD.It seems like just yesterday the Barcelona was going to be the killer processor. Now its too late, too slow and not yet running full blast on a 45nm process. AMD now races to recover from its bug fixes in time to hit a rapidly closing market window in a darkening economy.It seems like just yesterday AMD hailed its ATI acquisition as a key piece of a grand strategy to integrate graphics onto a future processor. Recently the market cap of AMD was less than it paid for ATI; clearly AMD overpaid. The financial resources AMD surrendered have slowed its R&D and capex strategies and plans. Profitability remains problematic. Moreover AMD's lower market cap may make it a takeover target for semiconductor players looking to add processors and graphics to their product portfolio. AMD's recent Middle Eastern investor may have had that in mind when it invested in AMD.

Implications: New product hype must match product reality Early adopters can become harsh critics A customer burned will cause lost sales

Analysis: Apple spent a lot of money to work early adopters into a "gotta have it" frenzy. In late June they grabbed iPhones hot off the production line and paid sticker prices. They had the latest hi-tech toy.

But then they had trouble activating their iPhones. They had to wait a long, too long a time for internet access. But they had bragging rights to their iPhone. Performance did not match up to the hype. A 2.5G cell phone is not so hot. A 3G phone is what the hype implied. Soon users realized what they bought was an overpriced iPod-like device with a non-state of-the-art cell phone thrown in.

Adding insult to injury, less than two months later Apple dropped prices by $200 and also dropped the 4GB iPhone. Some users became ophaned product owners. Clearly Apple sales did not match sales forecasts. Why do you think Apple dropped the price so quickly?

Apple probably punished its component suppliers with demands for lower prices. They Apple provided a new forecast for increased sales as bait to their suppliers.

Meanwhile, Apple must deal with burned customers and suppliers. What does Apple think about how future customers and valued suppliers will behave? Will Apple ever recover from angry customers and suppliers who"drank the Kool Aid" along with the hype? Apple, eat your humble pie. Get real; think about what you promise.

Implications: An Intel move had long been rumored in the last few years. In one big move the new JV is now the world's largest NOR Flash memory player. Whether or not this new JV creates a more rational NOR market is probably in the hands the major cell phone handset players. Based on past market shares, the new JV is the largest NOR Flash supplier. The infusion of financial resources from Francisco Partners will stablilize the transition of Intel and STMicro into the new JV. Layoffs are likely due to redundancies in Intel and STMicro. The new JV will put huge competitive pressure upon Spansion.

Analysis: An Intel move had long been rumored in the last few years. Intel had been signaling its disappointment. Its NOR Flash business was recently costing Intel about a negative three to four percent off its gross margins and negatively impacting Intel’s bottom line and stock price.
 
In one big move the new joint venture, JV, is now the world's largest NOR Flash memory player. The new JV is now the king of a red-ink-bleeding NOR market. All NOR Flash players have been losing money for the past few years. Market share wars had debilitated all NOR Flash players. A consolidation was long overdue.

Whether or not this new JV creates a more rational NOR market is probably in the hands the major cell phone handset players. Cell phone handset consume about 67% of all NOR Flash. The biggest negative forces confronting all players in the NOR Flash market came from the major cell phone handset players such as Nokia, Motorola, Samsung, LG and Sony-Ericsson.

Based on past market shares, the new JV is the largest NOR Flash supplier. The new JV will soon abandon the NAND Flash business since Intel already has a stake in the NAND market with its Intel-Micron Flash Technology JV, IMFT.

The infusion of financial resources from Francisco Partners will stablilize the transition of Intel and STMicro into the new JV. These resources are very likely critically needed by the new JV to maintain the CAPEX and business investment required to lead the NOR Flash memory market.

Layoffs are likely due to redundancies within Intel and STMicro. Marketing and sales layoffs will likely occur in the first year. Redundancies in other business areas will be eliminated in the next 24 to 30 months.

The new JV will put huge competitive pressure upon Spansion. Spansion has been losing money both before its IPO and ever since its IPO. The Intel-STMicro JV may likely push Spansion into even greater losses due to pricing pressures from the cell phone handset players since they consume so much NOR Flash. Also pressuring the NOR Flash market is the increased usage of NAND Flash in cell phone handset.

Implications: The PC Market has slowed down. Intel's weak Q2 sales revenue guidance confirms this.

Intel did a good job of cost cutting and making yield improvements, but not as good a job in booking business so far this year.

PC Bookings were pretty weak in Q1; this negatively impacted Intel.

Better bookings will occur in June if stronger than expected back-to- school PC builds materialize.

 

Analysis:

Intel managed to show improvement due to a very heavy duty focus on cost reduction efforts. Intel has been very good at managing its own costs. For example Intel had from 10,000 to 15,000 employee layoffs in the past 2 years.

However, the fact that sales slipped 1% indicates seasonal market softness in PCs. Also Intel’s projected forecast for the second quarter is also below expectations. Intel is not riding a wave caused by a great surge of PC demand. Intel must now wait for better bookings to occur in June, This will happen if stronger than expected back-to- school PC builds materialize.

However, at the cost of lower pricing and gross margins, clearly Intel made AMD pay dearly for its market share gains of the past 3 years. Intel has regained the momentum they lost to AMD during the market share war.

The viciousness of the processor market share and price war shows in the drop to 50% from 55% in Intel gross margins year over year. While this has been difficult and painful for Intel, clearly it’s been a lot worse for AMD. The market share losses to Spansion and continued losses in Intel’s Flash Memory business also contributed to Intel’s lower gross margins.

But Intel shows no signs of letting up with its new array of processors planned for the second half of 2007. Intel’s forthcoming rollout of a wave of new processors will more than counter whatever AMD might bring out in the meantime, such as Barcelona.

AMD has been pummeled and battered by Intel. So one has to ask the question “Has AMD reached rock bottom?” I think so. I think they’ve gone down about as far down as they can in this cycle. Whereas Intel on the other hand will perhaps in the next year reach a point where it is as good as it gets. Hopefully that is not the case if Intel has a very strong second half in 2007.

Intel has been able to hold the price of its processors due to its rapid migration from 65 nm to 45 nm process technology and yield improvement. This has allowed Intel to keep it prices fairly flat while widening its gross margins. Intel is pressing forward with a vengeance.

Implications: Intel, Samsung and other placing their bets on Phase Change Memory.

Customers may now try PCM in NOR Flash applications.

PCM is a potential replacement for NOR Flash memory.

Analysis:

Intel’s new 128 Mbit phase change memory is their first phase change memory, PCM, product offering. After a few years development Intel is testing the waters. Their hope is get customers interested enough to test the product in actual applications. Intel is not alone since Samsung has also rolled out a higher density 512 Mbit PCM. Both companies are putting their bets on the PCM memory table.

However, these may be just proof of concept products. PCM is a potential replacement for NOR Flash memory. PCM represents a hedge against NOR’s possible inability to scale below 65 nm processes. PCM may be coming soon to a NOR Flash socket near you but do not cash in your NOR Flash chips just yet.

Customers may now try PCM in NOR Flash applications, although such experiments are just the start of a significant shakedown cruise. A key target application for PCM is in cell phone handsets where it might replace NOR Flash. Two thirds of all NOR Flash goes into the cell phone market. By this time leading handset manufacturers may have PCM test beds. Beyond cell phone applications PCM might be integrated as embedded memory in microcontrollers. Low density PCM may be quite compatible with microcontroller circuitry.

Although some may claim PCM is a DRAM replacement, that may really be just a multi-year long shot at best. DRAM is not yet approaching near term process limitations, so such a claim may fuel some PCM laboratory experiments but little else for now.

But Intel, as long as it remains in the NOR Flash market must protect its franchise and hedge against other Flash memory players such as Samsung, Infineon, Elpida, Macronix, Renesas and Hitachi. Of these Infineon and Macronix must attempt to leapfrog over the NOR Flash market with PCM. Ovonyx, PCM’s inventor, is expecting that its licensees, many of the companies just mentioned, may eventually ship royalty bearing chips. Otherwise companies will be relegated to also ran status as Flash memory players.

PCM has many manufacturing hurdles to overcome. The memory business is an evolutionary, not revolutionary business. Evolution takes many years, not a few calendar quarters.

Implications: This is the time of year for worrisome sentiment.

Seasonality is often forgotten.

Potential for overcapacity does exist but the industry has ways to adapt.

Analysis: This is the time of year for worrisome sentiment.

Post-holiday industry sentiment mirrors life in general. Once a year the holidays and New Year optimism and good cheer are over and gloom sets in. After year's end the reality of the moment strikes the semiconductor industry. Fear grips technology pundits and industry. The semiconductor industry does not see strong incoming orders for Flash memory and it panics.

Seasonality is often forgotten.

Each year for the past few years, around the end of December and in January, Flash memory pricing and orders fall off dramatically.  The business climate chills for the next few months. The semiconductor industry reacts and shifts memory capacity away from Flash memory until pricing ceases its rapid fall and starts to tighten.

The potential for overcapacity does exist but the industry has ways to adapt.

What is left out of the picture is that other factors are at work; end market demand factors for one. The end of the seasonal holiday consumer electronics buying season reduces Flash memory demand. Flash memory suppliers habitually overshoot product demand for fear of missing new orders.

The introduction of Microsoft Vista has caused DRAM producers such as Samsung and Hynix to shift capacity away from NAND Flash to DRAM. The end market memory surge in demand for Vista-equipped PCs DRAM will cause DRAM and NAND memory production corrections. DRAM supply will increase. Another correction will be a cutback in NAND production for the next four months. Similarly, suppliers will push out plans for new Flash memory capacity expansion.

Thus, in time the memory industry makes its seasonal adjustments and the tough times for the Flash memory industry are mitigated.