Here are my notes from the ASM meeting. I finally got some of the stock and wanted to hear what they had to say about their potential and problems. I’ve used CRAY equipment in the past but only as a user of programs that ran on the machines. I won’t try to get into the details of architectures and such. If anyone else attended and can clarify items, please do so. It is inevitable that I got something not quite right. As always, when in doubt, call up the Investor Relations folks.
Cray Annual Stockholder’s Meeting 2000 – May 16, 2001
This is the first year that I have owned CRAY stock so this was the first year that I attended the stockholder’s meeting. They held it in their headquarters and probably had over 100 people show up. I suspect a lot of the late shows standing in the back were employees. They had some of components laid out as examples of their work. I’m not experienced at evaluating hardware so I found it interesting though not overly enlightening.
FORMAL MEETING
This was competently managed. There was 98% of the votes represented. There were also a lot of suits to introduce. Evidently some of them don’t normally wear suit and tie. Fine by me. This piece was over quick with no surprises. Every one running got elected.
INTRODUCTION to the PRESENTATIONS
The CEO started off by describing the 3 new products and their schedules. 1) The super cluster machine (what is it’s name again?) is due out in the second half of 2001. 2) The MTA-2 was booted for the first time in the first half of 2001, should ship by the end of 2001, and should add to revenue in 2002. 3) The SV-2 sounds like it will generate revenues in the second half of 2002.
VIDEO
There was a 15 minute video from a bunch of talking heads describing some of CRAY’s potential. The encouraging thing was that a lot of them were not CRAY employees. It sounds like there is enthusiasm for seeing what the new CRAY can do. A comment was made that there hadn’t been much innovation in the supercomputer industry lately. Now that CRAY was teamed with TERA, there was a lot of potential. In the interim many companies have been using clusters of low end machines to handle simple but big jobs. Basically the industry’s commercial customers aren’t interested in doing computer science. They just want results. The trick is to reawaken the interest in supercomputers and in vector operations. It doesn’t sound like it is a hard sell. There are always bigger problems out there that folks want to run on the next generation computer. Aside from the normal customers of autos, aerospace, petroleum, and government, there is also the potential for the bioinformatics industry to generate a lot of customers. There was a nice description of the various CRAY products and where they were in their life cycle. The folks at the door told me that the presentation was available on video tape or as a CD. Maybe they still have a few. I think I’ll be getting mine in the mail.
The CEO followed up with some comments. Evidently the CRAY technology that works well for analyzing codes for the government (decoding 1’s and 0’s) is also good at analyzing other codes (the A’s, G’s, T’s, and C’s of the genome). The folks being interviewed described the desirability of dealing with a company that has a wide array of offerings. He feels that all of the pieces are in place and is glad to see that CRAY is making the competition uncomfortable.
SALES & MARKETING
The VP of Sales got up and did a nice job of breaking down the market and the revenue streams. I get the impression that he only dealt with the hardware side of things because he was followed by someone describing the service side. CRAY expects to generate about $95M in 2001 from hardware sales. About 45% of the it from the T3E and about 40% of the total is lumped in with secret government stuff. It can be hard to judge though because Q3 and Q4 are the busiest quarters. The majority of the transactions (out of a total of about 75 transactions) are under $1M. They are trying to go for larger and more complex deals. So far it has all been with existing customers. This should improve though since there is an increase in the number of customer visits. CRAY is starting to act as the leader.
CUSTOMER SERVICE
The Customer Service segment was much shorter though it is still generating a lot of revenue (about $80M). For the most part it represents the work done by 284 folks world wide. They help with sys admin duties, optimization, and have just finished converting about 90% of the SGI contracts to CRAY contracts. I had to wonder if the ones converting the contracts were also doing the sys admin work. I didn’t ask. He didn’t say.
QUESTIONS & ANSWERS
These will be my paraphrased version of the question and the answer.
 The supercomputer market and the market share are defined somewhat arbitrarily but assuming it involves only those machines that cast more than $1M, then the market is about $1.5B and is growing at about 8% per year. CRAY currently has about $100M of that $1.5B. A third of the market is vector products. CRAY hopes to increase market share by adding products and feels that they have the resources to do the work.
 Their burn rate and cash needs might be okay. The dollars involved are so large and spaced out that even cash flow doesn’t happen. The CEO pointed out that they had to raise some cash a few months ago when the money markets weren’t favorable. Luckily, the trailing twelve months have been positive. The company does not have a strong working capital position but doesn’t need to go out and get funds either.
 Classified equipment sales are already included in the financial statements. They just won’t go into details about them.
 CRAY is probably not a takeover candidate unless they make a lot of money.
 There are no acquisitions planned, though partnerships like the one with NEC are more likely.
 The NEC agreement is being treated as an opportunity to present a larger product line within North America but NEC is still treated as a competitor elsewhere.
 The agreement with Sun is meant to help provide similar interfaces across all of CRAY’s platforms.
 Bioinformatics are being pursued via partnerships. The one they described most frequently was with the National Cancer Institute.
 Burton Smith was not there.
 The vector machine competition is basically NEC with $200M, Fujitsu with $200M, and CRAY with $50M. Fujitsu is not going to pursue vector machine any further.
 Managing a company across many locations is hard and takes a lot of flying.
 The bioinformatics market would add to the 8% growth mentioned above. Estimates of the size of the market vary wildly. Within 4 to 5 years the market is expected to be somewhere between a few hundred million dollars and ten billion dollars. The CRAY folks are expecting something closer to the lower number.
 The NEC agreement is set up for ten years though NEC can back out if CRAY does not meet certain targets. The agreement also allows NEC to access the US government market.
 The last year was spent getting the word out to the potential customers. The focus will now start swinging to the investment community.
 There was a New York Times article on supercomputing that didn’t mention CRAY. CRAY has been educating that reporter. They are happier with the relationship they have with the Wall Street Journal.
 The bioinformatics firms tend to buy machines for less than $0.5M. They suspect that is because the purchase is being made by a division, not at a corporate level. It doesn’t sound like CRAY will be setting up a machine to lease time out as a teaser to small businesses.
 There is no active prevention against a hostile takeover though the staggered board terms and the legal structure of the SGI agreement make a takeover much more difficult.
 Evidently Cray Research didn’t fit into SGI’s long term game plan and innovation suffered. The CRAY/TERA match up seems much better.
 The practical consequences of export restrictions may have loosened since the early 80’s. I can’t remember what brought that up unless it was the sales of weather equipment to India and Pakistan.
MY CONCLUSION
I buy and hold for the long term. I like CRAY and don’t think supercomputing will ever go away. I don’t know that they will become the market leader but even if they don’t they are projecting revenues in excess of the current market cap. The company seems to be at a critical stage. There is a lot of potential for CRAY’s new machines to grab lots of orders but a lot of that potential is a few months out. I can’t imagine that the economy is the best for them right now. If the regular product line hangs in there, orders are paid for to support cash flow, if the new machines live up to their promise and if the bioinformatics market kicks in, then CRAY may do some astounding things. Things could still be tough though if they hit a number of hurdles at once. That is true of any company but these folks don’t have the resources of Intel or Microsoft to carry them through. The next twelve months should be fascinating.
DISCLAIMER
Like I said above, I am not an expert and admit that I might not have the correct understanding of some of these items. Feel free to correct me on items and like I said before, when in doubt, go talk to Investor Relations.
In the meantime I will be staying tuned. |
