AMD just posted its best quarter result in years and slowly becomes more than just a headache for Intel. It was time to find out more about AMD’s success and climbing market shares. Randy Allen, corp. vice president of AMD’s server products division, talked to us about product strategy, upcoming technologies, 65 nm transition and the impact of the antitrust complaint against Intel.
The quarterly reporting of financial results typically is not quite what makes corporations and their market segment interesting, at least if your stock portfolio isn’t concerned or if you aren’t in the business of predicting the future for a particular company or industry. But once in a while, these reports indicate trends that affect us all – which makes them well worth watching. And such a trend could have revealed itself earlier this week, when AMD and Intel presented their most recent financial performances to the public.
Other than expected, Intel uncovered a rather sobering result, even if the company crossed the $10 billion revenue mark for the very first time in a quarter and profits exceeded $2 billion by far. The balance sheet however showed that basically all of Intel’s gains originate in the firm’s mobile processor business; many other traditional processor segments – including servers – revealed declines. AMD on the other side presented sharply increasing revenues and profits and claimed that it gained market shares in desktop, server and mobile market segments.
While actual numbers of potentially shifted market shares will not be available until Gartner and IDC release their quarter reports, the stock market already has reacted to the situation. In the beginning of Nov 2005, AMD was valued at about $8 billion vs. Intel’s $141 billion – or about 6% of Intel’s value. Until today, AMD’s market cap has grown to $15 billion, while Intel decreased to $135 billion. Suddenly, AMD is worth 11% of Intel.
We have seen AMD teasing Intel once in while over the past decade several times, but essentially, AMD has been locked down into second place. The stock market appears to believe that it could be different this time around. We should not forget that Intel still outsells AMD in every segment with a strong lead, but that lead appeared to be shrinking in the third and especially the fourth quarter of 2005.
Responsible for that shrinking lead is especially AMD’s server products group. Intel’s CEO Paul Otellini recently acknowledged that Intel had to give up market shares to AMD and will likely be forced to hand over more shares until the next generation of server chips arrives. The origin of this weakness is often seen in a particularly poor product lineup in the 2003 and 2004 time frame. The circumstance that Intel could not have paid not enough attention to AMD or market trends, may cause the company not just a headache now.
AMD’s Randy Allen explains in this conversation with TG Daily why he believes that Intel will need much more than a new processor to be able to slow AMD’s growth.
Gaining market shares
TG Daily: It is no secret that your server business has contributed a significant portion to AMD’s climbing revenue and profit in Q4. What portion of that success can be attributed to the server division?
Allen: We do not break those figures out. What I can say is that servers did exceptionally well. We do not have the final numbers until we work with Gartner an IDC, but we are certainly very pleased with the market share gains we have seen in the fourth quarter.
TG Daily: The financial world apparently does not view your Q4 result as an accident, but possibly as a trend. How would you explain the improvements of AMD’s server results over the course of the past year?
Allen: We believe that we have had a superior product for several years now. But it took a while for the OEM’s to broaden their product offering. 2005 was a phenomenal year when you look at all the OEM’s, including HP and IBM, and how they have widened their portfolios. The reason why they committed is because we have several advantages over our competition. We offer the highest performance, the best price/performance processor, the best value and – what has come on really strong in the second half of 2005 and into 2006 – a performance-per-watt advantage. Plus, our dual-core products have continued to fuel our market gains.
TG Daily: Intel conceded that it may continue to lose market share to AMD until the launch of their 65 nm Dempsey processor, which may arrive in customer hands in the second quarter of this year. Which disciplines will be critical for AMD to keep its current lead?
Allen: Our key advantage has been performance-per-watt all along. There are those folks who focus on the most performance and others that focus on performance-per-watt. The fact that we lead in both are the two factors that drive continued gains in market share. Looking at our roadmap, we will see significant improvements in both dimensions, performance-per-watt as well as performance. Our OEM partners seem to be very pleased not just with our current, but also with our future products. In fact, compared to 2005, the number of our platform solutions will significantly increase in 2006. Said in another way: I believe we actually will be gaining momentum as we are going into 2006.
TG Daily: AMD recently has been very tight-lipped on its plans for 2006. What exactly will ensure that you can keep your current pace, given the fact that your competition appears to be getting up to speed?
Allen: First, we have the transition to DDR2 in 2006, which presents new opportunities for OEMs and large Opteron platforms. In addition, the blade market continues to grow and this segment has selected Opteron because of its performance-per-watt advantage. Those are two factors that are driving the deepening commitment to Opteron in 2006.
TG Daily: Admitted, we haven’t seen Intel’s new processors and platforms yet, but will that be enough? What improvements do you have in store for the Opteron processor?
Allen: You will see an increase in frequencies, which will mean an increase in performance. You are going to see us add features that will become an important part of the platform. For example, we have worked with Microsoft, VMware and XenSource to launch a superior virtualization product in 2006. Of course, we are always making improvements to the core to deliver increasing performance. Summing up, you are going to see virtualization, higher performance and higher performance-per-watt in 2006. And if you look at the transition to DDR2, this technology proves a higher level of performance in itself.
TG Daily: 2006 appears to become one of the most interesting years in the history of the microprocessor and the competition between AMD and Intel. Still, AMD expects continued market share gains. Why do you think the market wouldn’t wait for Intel’s Merom processor core?
Allen: Good question. I believe Intel is simply too late. Sure, they are not standing still and do everything they can against the market share gains we are experiencing. But we are not standing still either. OEMs are looking at the roadmap Intel is presenting to them and they are looking at our roadmap. I am sure that they see that we can continue to extend our lead through 2006.
TG Daily: How far is Intel behind then in AMD’s view?
Allen: Look at IDF, where they probably make their most significant statements about upcoming technology. There they made no mention of coming up with a system architecture that is really competitive with direct connect architecture at least until 2008. If we look at their roadmap and everything we see they announced, they are not addressing the fundamental problem they got – which is their legacy front side bus architecture. I believe they are fundamentally disadvantaged until they address this bottleneck.
TG Daily: It has been a few months since AMD has filed an antitrust complaint against Intel. Given the media exposure of this move, do you already see an impact on your sales?
Allen: Absolutely. This really put a spotlight on the fact that AMD is very important for the market and provides people an opportunity to choose the best technology that is available. Yes, people looked into the lawsuit and our claims of their bad behavior. What is more significant, however, is the more people look at us the more they love us. People began to bring in Opteron system and benchmark them against a competitive system from Intel. We definitely have seen growing acceptance [for our systems]. Until today, 90 out of the top 100 companies have deployed Opteron systems.
65 nm production by the end of 2006
TG Daily: How long, do you think, will dual-processors be able to deliver enough performance to satisfy customers?
Allen: We target the introduction of multicore in the 2007 time frame. 2006 will be a year of significant transition from single-core to dual-core and we expect the majority of customers purchasing dual-core systems this year. In 2007, the transition to multi-core will begin.
TG Daily: How do you view AMD’s progress in improving the power consumption of processors? Do you already offer a reasonable level or do you see power consumption decreasing over time?
Allen: We currently offer Opteron processors that at 68W, a 95W and 120W. This basically allows people to trade off performance versus power. The folks who don’t have the power delivery and heat removal that is required for a 120W part can go with a 95W processor. Those guys who try to get very dense are going to choose the 68W chip over the 95W and 120W CPU every time. The question is not so much ‘What is the right power point?’ but rather that you should provide a broad product portfolio so that those individual users and OEM’s can make those choices for the particular market segment they are targeting.
TG Daily: … which means power consumption level will not decrease?
Allen: Our approach is to largely deliver ever increasing levels of performance within the same power budgets and thermal requirements. Of course we always listen to our customers, but in general we believe that we offer nice power points and we will continue to innovate in those.
TG Daily: What priority has your 65 nm transition at this time?
Allen: 65 nm is critical as we look into the long term. It’s going to allow us to reduce cost for our products. But more importantly, from the view of the market share gains we are seeing, it allows us to increase the capacity in our fabs. Right now, our 65 nm program is on track. We will be delivering 65 nm production by the end of this year. Interestingly, customers do not care if their processors are 90 nm or 65 nm; they care about product features. That said, we are delighted about the progress we are seeing from our new Fab 36. It will start initial production this quarter and it will transition to 65 production by the end of the year.
TG Daily: Compared to Intel, AMD will be just about one year behind in the 65 nm space. What disadvantages does that imply for your company?
Allen: 65nm will deliver a cost advantage over time, when the technology matures. AMD has been very successful at being very aggressive in terms of our ramp from one technology to the other. The key point is not so much when you introduce the technology, but at what point you are hitting a crossover – when the majority of your production runs at new technology. What technology you are running is a clear secondary issue, relative to the product attributes.
TG Daily: What will be your most important challenge over the next 12 months?
Allen: This clearly will be to flawlessly execute the roadmap that we have shared with our partners. If we can just do what we said we are going to do, then OEMs are ready and rolling to increase their commitment to Opteron, Turion and Athlon 64. Execution means execution on our DDR2 transition across our product lines, it’s about Fab 36 getting to the capacity we need, and it’s about the transition from 90 to 65 nm. We have a great product family, a great processor. We just need to stay the course.