AMD analyst day presentation notes December 14, 2006

Notes on AMD December 14, 2006 Analyst Day Presentation

These are the raw notes I took which I am sharing with you. See also my main AMD page.

December 14, 2006

AMD Analyst Day notes

Began with an Oracle presentation on Opteron for reducing power and cooling requirements while enhancing performance. Roadmap looks like it meets needs today and in future. Serving on-demand customers with Opterons.

Dirk Meyer (President & COO) says PCs are increasingly used for media; acquisition of ATI enables that. Industry moving from single-threaded performance to multi-core led by AMD. More and more general purpose cpus on a die will be a pointless war. Need specialized cpus built into chips. Allows ability for partners to differentiate their products. Converged devices (media - computer) will be a high growth area. ATI uses fabless production model which AMD plans to adopt heavily.

Chipset business is aligned with GPU business, but is driven by CPU roadmap.

Leadership opportunities include energy efficiency, visualization, and global affordable platforms (50x15).

AMD Live outsold Intel Viiv last month. Deep relationships with top China suppliers. Great progress with other local (non-US) OEMs. Partnered with Microsoft in Brazil with pay-as-you-go product. Energy-efficient Athlon 64 X2's flying off the shelves.

Aims to transform point product customers into portfolio customers. ATI sales force global and a great addition. ATI brand to continue to be used for graphics and chip sets. Has penetrated high-growth markets. Consumer electronics to be a big part of sales.

Server space a $6.5 billion opportunity in 2007.

Perfect alignment between new AMD and next generation of consumer products. $15 billion opportunity in desktop IT and game consoles. Mobile is a $9 billion opportunity. Fashion is important in notebook.

Vista will change the paradigm and criteria for device selection in 2007.

Now a leader in digital TV space. Handheld PDA & phone is a $4 billion opportunity. Designs with 5 of 7 top OEMs.

Continue to support OEMs: "we are there to support them, not supplant them."

Stability of platform is very important in server space. Independent assessment shows AMD continues to have performance per watt advantage over Intel.

DirectConnect architecture's superiority allows for superior blade designs; AMD has over 50% wins. Also high % design wins in rack space, desktop, and mobile PC.

2007 goals include continuing to set server center agenda; virtualization; be safe choice for servers; penetrate commercial client PCs.

Trinity - open approach on virtualization, management, support standards, interoperability.

Raiden - client reinvention, new form factors, usage models.

Quad core will be native, DirectConnect, with cache and core enhancements. Drop-in upgradability from 2 cores. It will show how to do quad core right.

Torrenza is designed for discrete acceleration. A particular piece of silicon for a particular task. One example is Stream computing.

In consumer desktop in November 2006 had 40% of shelf space. 23% of mobile shelf space. Has ability to increase SKUs in mobile. In past customers had to chose between battery life and graphics performance. In 2007 will have notebooks that switch automatically so you can have both (dynamic graphics power switching).

Surprise: will continue to use Nvidia and Broadcom for server chipsets.

Coming soon: hard drives with flash auxilary.

UVD in late 2007 will allow offloading of compression/decompression from CPU.

Emphasized commitment to open approach to ecosystem.

GPU business: DX10 and Unified Shader architecture. Avivo and UVD (Universal Video Decode). Crossfire. GPGPU. Claims lead in graphics with unified shaders, other R600 technology. Vista will drive with DX10 engine.

Fusion, apparently some time in 2009, will put the GPU on the CPU chip and have a customer-centric overall architecture.

Chipset business is major growth opportunity in 2007. Also transition of TV from analog to digital, both in signals and in screens. Xilleon is AMDs tech for this. Moving to network-to-glass integrated model. "Every notebook in the world is really a TV waiting to be turned on."

Mobile phone growth opportunity because of desire for easy integration of rapidly evolving graphics technology. Motorola RAZR use AMD Imageon technology, as do many other brands. Good gross margins in this business.

APUs - accelerated processing units - will combine CPU and GPU and possibly other technologies. Torrenza will be socket level accelerators. Later will integrate on silicon (Fusion).

Now at full volume production of 65 nm process at very high yields. Working on 45 nm validation and will have in volume by mid-2008. Manufacturing flexibility from having own fabs and a fabless system. New York fab could begin construction in mid-2007 to mid-2009.

Is already sampling Barcelona quad core processors.

Diversity in workloads really needs to be addressed.

Plan is to address fundamental limitations in notebook computers, starting with performance per watt.

Gained market share in 2006; grew at 3 times the industry rate.

Did lose ATI revenues at first because of some of Intel business going away, but there are cost synergies in 2007 of $250 million, increased revenues of 80 to $160 million. So will be dilutive in first half of 2007 and accretive in 2nd half. Gross margins of the combined company will decrease from traditional AMD. ATI would have lost money Q3 2006 because of inventory adjustments.

Q4 2006 350-400 million in ATI sales. $550 million in merger costs, much of it non-cash. Starting with Q1 2007 will report 3 segments: computing solutions, graphics, and consumer electronics. Computing solutions had $5.8 billion revenues in the last 4 quarters. Gross margins were 53%, operating margin 15%. Margins will increase going forward because of shift to 300 mm silicon, etc. Graphics segment was $1.3 billion, gross margin 23%, operating (3)%. Believes can increase graphics operating margins about 18%. Consumer segment revenue was $550 million, 48% gross margin, 17% operating margin, can increase margins somewhat in future.

Overall annual revenues were $7.7 billion, 47% gross 9% operating margin. Believes can reach 51 to 55% gross margins in future, with 18% to 24% for operating margins (but not in 2007).

Guidance: Growth in CPUs will be 2x industry rate of 10%, other segments at industry rate (graphics 5%, DTV 44%, handheld 19%). 2007 gross margin 48 to 52%; R&D and SG&A expense will go up about 20%, capital expenditures around $2.5 billion. Free cash flow will be negative $500 million. $380 million acquisition and amortization costs. 2007 stock option compensation of $135 million. No specific revenue or net income numbers.

Hector Ruiz recapped how far the company has come in the past 5 years. Accomplishments have emboldened the company.

Question and Answer Session

ATI third quarter loss to continue? Does not expect to lose money in any given quarter in 2007, but expects gains to rise as 2007 progresses. The 3rd quarter 2006 ATI loss was mainly due to Intel chipset business that went away.

Market share gains in servers? Started 2006 targeting 20%. Last reported quarter had 24%. Bullish will have higher percentage when reported at end of this quarter.

ASP assumptions on microprocessors? Client ASPs (prices) are likely to be neutral, mobile flat to slightly up, servers are apt to be down on average because of moving into 1P space, hard to say on comparable (2P or 4P) basis because competitor (Intel) continues to price aggressively.

Quad core timing? Sticking with mid-year 2007. Virtualization shows dramatic improvement for 4-core. Mobile design will be very power efficient. Server quad-core has improved performance and improved power efficiency.

Fusion processors? Have one design under development. Will be 3 to 4 variations. First design will be targeted at mobile space. Will be a SOI technology for power efficiency.

Retail space % drop? Desktop shelf space was sacrificed for mobile shelf space. Focusing much retail development energy outside the North American market.

Sell fewer servers because they are getting so efficient? Companies that want virtualization will choose AMDs superior technology. Will help maximize server utilization. Discrete acceleration capabilities will build AMDs market share. Cost effective for companies to switch from RISC or older x86 to newer chips, so size of market will continue to grow.

Vista power consumption for notebooks? True in general Vista needs more energy.

Supply side, both AMD and Intel are building capacity? AMD has been slightly on the cautious side, even so increased 32% this year. Likely to need a 35 to 38% increase in capacity next year to meet needs of clients acquired in 2006. AMD will continue to work with OEMs who are differentiating their brands. Big question is when to start New York plant. Ask Intel why their die size never goes down even though they shrink technology? AMD believes because they need to chew up a lot of silicon to stay competitive in the space because their architecture is poorly designed. Their quad core implementation is pretty darned large.

Is ATI integration into AMD to help profitability? Yes, consumer electronics is more profitable than AMDs traditional business. Chip sets should be profitable. Graphics should show profits once discipline is applied.

Upside surprise on ATI? Gross margins will be better than they originally thought. Equipment acquisition costs will be lower than expected. R&D will not have to grow as fast as in the two separate companies.

Pricing first half v. second half of 2007? Believes market growth overall will be 10%, will gain market share with OEMs, power of choice is being felt, likely to pick up 1 or 2 significant customers. Challenging scenario is in desktop, partly because of move toward notebook space. Opportunity to increase notebook share because they are behind there.

ATI GPU price increase? Need to reverse on recent market share losses of ATI. AMD knows how to fix the problems ATI had.

OpenIcon Analyst Conference Summaries Main Page

Disclaimer: Our analyst summaries may include both our condensations of statements made by company representatives and our own analysis. They are not covered by any warranty. We cannot guarantee anything said by company representatives is true. We try not to make errors, but it is possible. Before making or terminating an investment you should always verify any factual basis of your decision.