What is HP's Unix strategy in light of the company's decision to drop PA-RISC?
Unix is a very important part of our operating system and operating environment strategy. In the foreseeable future, I don't see a big change. We've got an operating system strategy, which we've been pretty consistent on, which is Unix with HP-UX, Linux, and NT and all the Microsoft environments. In addition to that, we've got some specialty operating systems like NSK (NonStop Kernel) and OpenVMS and others, which we will continue to support for a number of years as we port those to Itanium.
But our Unix business is really robust. And the way we see this Linux/Unix/NT battle playing out is everybody wins.
We are consolidating to microprocessor platforms around IA-32 and IPF (the Itanium processor family). And we are porting all our operating systems to IPF. In fact, HP-UX is up and running now on IPF.
How do you expect that business to pan out? ?
Is HP-UX on Itanium going to be a strategic system for HP for the long term? Absolutely. I believe it will be the operating system we're using to set a lot of the new benchmarks. That's all running on our Superdome platform with the Itanium processor in it.
So it's absolutely not the case that the move away from PA-RISC means moving away from HP-UX.
It is absolutely not the case - you're right. If you think about the customers that are on HP-UX and the applications they're running, these are probably the most sophisticated enterprise environments in the world. You don't switch platforms - it's really hard.
There is certainly a perception among users we've talked to that RISC and Unix go together and that distancing yourself from RISC naturally means a distancing from Unix. Are you fighting any such perception or confidence issue among your HP-UX users?
We were worried about it in the beginning. We went out and worked very carefully with our big enterprise customers.
The nice thing about the Unix space is the customers are really big: You know who they are; you can go sit down and have a conversation with them. We've literally lost, I think, one or two to a different platform, and they were people who were looking to move for other reasons. Even those have gone to Linux or NT.
The trick has been not so much us, but the ISV [independent software vendor] community porting to Itanium. If you get applications ported over, then you're in great shape.
What sort of a curve do you see for the HP-UX market? Upward, downward or remaining flat?
That's getting into what's happening in the Unix market overall. We see it as sort of flattish. But it's not a function of Unix. If you look at what's happening in the IT environment, it has been a little slow. If you look at our Unix business, we've done pretty well, relative to the market.
But will the HP-UX market curve be upward, downward or flat?
What I would expect it to do is a little better than the market.
I'm trying to get you to answer the question specifically.
I know what you're trying to get me to do, and you can see I'm not going to do it.
What do you think the Unix market overall is going to do?
I think it's a little hard to tell, but I think the Unix market is stronger than people realise because it's such an important part of all the major enterprise environments. People have got all their mission-critical business applications running on Unix platforms. That's not something you change quickly.
So you are unwilling to project where the Unix market as a whole is heading?
I don't have a crystal ball on the Unix market as a whole, so I don't know. And I don't think anybody else does, either. I think it's a mistake to underestimate the potential of a good Unix offering.
Would it be fair to characterise Unix as a casualty of the RISC-Intel war that Intel won?
No. One of the things to keep in mind is the economics of what's happening here. This is as much about economics as it is about speeds and feeds. What you're seeing is a consolidation in this industry, where you can no longer afford the investment required to support the entire ecosystem around a proprietary microprocessor architecture. So what we're moving to are more industry-standard platforms, where you get the economies of scale of having a partner like Intel provide a platform to the industry.
What's your response to someone who says Unix is becoming a legacy system?
I don't see that. We're adding features and enhancing Unix and doing all kinds of new things to move the ball forward. I think the Unix market [will be] around for longer than people believe. There are a lot of the big business applications and the whole supporting environment in the high-end space, and I don't see CIOs switching off that.
At the very high end in the supercomputing space, Linux is something that's really proliferating. Because the customers there - the national labs and universities - have armies of programmers who can support Linux. They don't have the same kind of business stability and service-level agreement requirements that some of our big enterprise customers do. At the high end of the business application space with the enterprise customers, Unix is still very strong. At the lower end - application servers and name servers and all the smaller platforms - Linux is really going great guns.
Will there come a point where there's no reason to run Unix because Linux will do everything Unix does for a lot lower price?
You could ask that same question about NT.
I could ask the same question about anything, but let's deal with this one?
At some point, but not anytime soon. That's a controversial statement. The Linux guys have made a lot of progress in the last couple years, and they're moving into certain parts of the enterprise. But in a lot of the mission-critical stuff, it's still a Unix world. Linux is an operating system that is positioned very well in certain market segments. The same is true with HP-UX. What our strategy has been is to make all three of these operating environments interoperate - Linux, Unix and NT.
Do you think Sun Microsystems, which is so dependent on Unix, has a viable long-term business model?
Sun's got two problems. They've got a very expensive hardware platform that they've got to get out from under. And they've got a crisis of confidence in their customers around whether they're going to be a viable company. So Solaris is losing market share to the other big Unixes and to Linux on x86, where the ROI is better.
What do you make of what The SCO Group is doing, and how do you expect it to affect the Unix market and Unix users?
It's caused a lot of noise. It's not clear to me until all the [intellectual property] stuff sorts out whether they are going to have any impact. This is far from over.
Are any of your Unix users asking you for reassurances that SCO won't be going after you next?
Not really. We have a pretty good relationship with SCO, so we haven't had the kinds of issues that some of the other guys have. For us, it hasn't been an issue.