GOVERNMENT

By Steve Fisher, Editor in Chief -- Amid much fanfare last week the new HP began formal operations and hit the street. This week Supercomputing Online interviewed HP executives Ty Rabe and Steven Joachims to get their thoughts on the new HP’s place in, and potential impact on, the high performance computing industry.
SCO: How will the combined capabilities of the new HP affect the high performance computing marketplace? RABE: First of all we think having had, Steve was actually on the integration team doing planning prior to the acquisition being completed and we’re now in the process of looking at the information from both sides of the equation just for the past couple of weeks. I think we both conclude that when you put together the HP heritage organization, talents, product set etc., and the Compaq heritage organizations, product set, expertise, that we have a set up resources that is unmatched in the industry for providing solutions for high performance technical computing. We have an extremely complimentary product set, where the Compaq focus has been Alpha servers, RISC architecture, very high memory bandwidth and an approach to supercomputing that was distributed, what’s called these days ‘scale-out’ meaning lots of systems interconnected. HP is coming at this with a roadmap around their Superdome architecture which is a scale-up symmetric memory system architecture, way ahead of anyone else in the industry as far as implementing Itanium processor family systems, extremely capable of performance at very moderate prices. Steve has a team of very talented people focused on solutions in Richardson, Texas and we have the equivalent here in New England. They’ve (HP heritage organization) really been the leading solution provider in computer aided engineering, we’ve been the leading solutions provider in life sciences, they do some complimentary things in life sciences and we do some complimentary things in computer aided engineering. When you put all these pieces together we just think we’re going to have a much better capability to serve customers going forward than we ever had in the past working independently. I think the impact on the market will be, if you put the pieces together now, we’re the leading supplier. Our intention is to maintain and increase that lead and to really be the provider of choice for the majority of the folks in the market. JOACHIMS: We made a decision, a strategic decision on the integration team to form one very strong HPTC organization that is the aggregate set of resources from both Compaq and Hewlett-Packard. Everyone associated with this business right now is really excited about meeting their peers and finding a way to be able to build a particularly strong set of partnerships, develop some new innovative strategies with our partners that are helping us deliver solutions to the marketplace. That’s what the meeting we had today was about, introductions, and a good understanding…people understand a little about Compaq and HP because we’ve been competing with one another for sometime, til just about a week ago. So we understand the strengths of the other business and we’re now beginning to understand how it is that we might be able to meld some of the technologies and partnerships and so forth together, and aim that outward in the marketplace and provide more value to customers because we have stronger reach across the globe. SCO: Let’s talk a little bit about what’s going on with HP-UX and Tru64. JOACHIMS: HP-UX for our Unix business is our strategic operating system and particularly for high performance computing what we’ll see is a firm commitment to our customers and new buyers, for Tru64 we expect to be selling Tru64 and associated Alpha-based systems through the 2005 time frame with support until at least 2011. So we’re not trying to whittle away at that business at all. It’s a marketplace that that really gains benefit from that operating system and there are some applications out there that are not available on any of the other operating systems we have available in our portfolio. So it’s certainly strategic for that reason. We’re going to be merging some of the technologies from Tru64 into our HP-UX offering so the HP customer, whether it’s a customer from the HP side or the Compaq side will benefit by merging many of the best technologies for high performance, scalability and clustering. SCO: Specifically which aspects or capabilities of Tru64 will be integrated into HP-UX? RABE: I’m certainly not an expert in this area so there probably are a number of other things that might be included from our perspective for our customer base are the file system and the clustering technology. We have a single system image clustering file system and the associated clustering software that ties together clusters in a single root in such a way that it looks like one system to the system manager or systems administrator. That is the key technology for us to provide the scale-out capabilities for high performance technical computing customers and that’s moving into the HP-UX product. SCO: How about OpenVMS? RABE: OpenVMS is going to continue on pretty much under HP just as it has under Compaq and to some extent was the same under Digital before that. It’s a very, very capable operating system that has an extremely large base of customers who are very loyal and don’t want to move to anything else and we’re going to support them until they don’t want to use it anymore. And we’re going to do that on Alpha now, and it’s going to be moved to the Itanium product family platforms and customers will be able to use OpenVMS for just as long as they want to. SCO: High-end servers. Where do we stand now? Where are they going? I’m particularly interested in the “Q System.” RABE: Let me start with the Compaq side of the equation. We had announced quite some time ago, I’m not sure if in all the uproar about the merger and the proxy fight if people got the message but a long time ago HP committed to carry forward the Alpha product roadmap that Compaq was already on, so we’re on a path to deliver the next generation of Alpha servers based on the EV7 processor architecture and we’re going to follow that up with a speed increase sometime in calendar year 2004. That will be reflected in the full line of Alpha-based servers from dual processor systems at the low end to 64-way symmetric multiprocessing systems at the high-end and we’ll be implementing the AlphaServer SC supercomputing architecture based on those EV7 systems as well so that will be the replacement for the current generation of Alpha servers. That combination of products will form the basis for the solutions we deliver at places like the ASCI program at Los Alamos that you refer to. So we’ll be delivering Alpha-based solutions in fulfillment of our contractual agreements and obligations to Los Alamos later this calendar year and we have other deliverables based on subsequent generations of Alpha servers that we’ll be delivering next year and the following year. There is absolutely no negative impact of the HP acquisition of Compaq on any of those contracts or on our product roadmap. It’s continuing on just as we announced it would while we were in the merger evaluation stage. Now I’m going to turn this over to Steve to talk about the HP side of the equation. JOACHIMS: We’ve created some server roadmap slides that show a real balance here for both our PA-RISC business, which has been very strong for us in high performance technical computing and for the Alpha server business we have two more generations of products and associate higher performance technologies that will be introduced into those product lines over the coming couple of years. In the midst of all this, we are orienting much of our investment around the success of IA-64 or the Itanium processor family as we introduce yet another, you know second generation product here in very short order in conjunction with Intel and partners. Both companies have been talking about moving to that technology over the long haul so as we project out let’s say three years, we want to be in a position where we are moving the vast majority of our business to products that are based on Itanium, HP-UX with a number of Tru64 OS features, Linux, and even Windows64 where appropriate. RABE: I think an important point to make here is we’ll be doing that in time frames where the IPF (Itanium product family) based servers will clearly be the best price/performance solutions for customers. So it will be a logical movement or migration, not something we’re forcing on them. SCO: Alright. Consider this your opportunity to send a message to the HPC community. Feel free to go with it where you will. JOACHIMS: I would say first of all that Hewlett-Packard is absolutely committed to the business which has been formed and identified within Compaq and Hewlett-Packard. We believe that there’s plenty of evidence out there that the two businesses were very close to leadership position within respective pieces of the market, HP in the engineering enterprise and Compaq in science and research particularly at the high-end. We have formed one strong business entity within Hewlett-Packard that has significant visibility across the overall business and we’re enthusiastic about taking basically a leadership position in high performance technical computing and bringing some innovative new technologies and capabilities to the market. We believe that the breadth of the solution portfolio and partnerships gives us an unfair advantage against particularly the likes of Sun and SGI. IBM being a larger entity has its fingers in a little bit of everything, but we certainly believe that we can go toe-to-toe with them in terms of products, solutions, and services delivered to the marketplace. RABE: All I’d like to add is that I’ve been involved now in high performance computing from Compaq for the last five years and I look at what we can do right now and I’m more excited than I’ve ever been about the kinds of solutions we can offer, the growth opportunities for HP in this market, and our ability to serve our customers.