Gordon Haff Quotations

The potential here is a lot more efficient use of hardware, better centralized management, and potentially more flexibility with users in that they're not tied to a particular desktop device. I don't believe there's a one-size-fits-all approach, we'll probably end up with a combination of thin clients and various types of fat clients.

Software patents are a mess, ... Even if you don't believe in abolishing them entirely, there have been far too many examples of patents being granted even in light of overwhelming evidence of prior art and of dusty old patents on their third owner suddenly being used for a form of blackmail.

Obviously, they have to sell lots of systems of this size to pay back what has to be a considerably larger R&D investment than they have even with Galaxy.

Certainly, the fact that you need to be using brand new hardware is a selling disadvantage out the door.

Certainly Power Everywhere, to some level, includes OEMing systems, though for the most part it's more focused on the use of the Power architecture by various partners in client devices, game consoles, whatever, than it is in computer systems, ... Power Everywhere and OpenPower are really about using collaborators ? to get Power into new types of devices.

Certainly it's at least at the same level of sophistication that Office 97 was in terms of features and capabilities, which is enough for most people, ... Microsoft seems determined to see how much money it can squeeze out of buyers, but in doing so, they're opening up an opportunity for somebody else to come in.

Business desktops are a narrow, narrow margin business; they are almost a commodity. You have to be very efficient in order to make money selling desktops.

Basically, no one ever used hardware-based IA-32 execution, so better to use the silicon for something else. Of course, basically no one uses software-based emulation either, but at least that doesn't cost chip real estate.

History says that binary translation basically doesn't work. The day may come when someone can do a good enough job with it, but that concept has been thrown out there many times in the computer industry, and it's always fallen flat on its face.

It is a question of the devil you know.

Particularly with this kind of niche, this isn't really so much about the box, it's about channels and go-to-market. It's about setting up distribution and supply partners more than it is about the right box.

People have tried a lot of special purpose processing devices over the years and, with the exceptions of graphics units and arguably floating point units, general purpose processors have always won out in the end.

Microsoft is going to be very late to the game and will need to overcome a presumption that it will favor Windows.

Microsoft has made a few unsuccessful attempts to become a key player in the supercomputer market but the company could be in a better position for success this time.

Microsoft has built its business on working with third party software vendors and wants to leverage that expertise to the supercomputing arena.

Michael Dell likes to periodically send messages to Intel that Dell does have the option of introducing AMD.

You have to do a lot of running just to stay in place in this business.

The problem is, to be at that nirvana, Apple would have to have the ISVs increase their investment on Apple hardware, at least in the transition. That is very difficult for Apple.

Those do not appear to be cherry-picked numbers of some particular benchmark on a blue moon with a tailwind blowing.

When the Pentium 4 first came out, the performance was very disappointing to a lot of people initially. It's hard to call the Pentium 4 a failure, given how many they sold, but it was certainly an evolutionary dead end.

With the benefit of 20/20 hindsight, there would not be an Itanium today.

Enterprises must either adopt multi core or stick with slower chips.

I see this as fine-tuning more than changing direction. Obviously, they're struggling financially, but in general this is a small number of people in the scope of things at Sun.

They are all comparable in performance for high-end, non-X86 architectures.

Programming Cell is relatively hard. Certainly a higher-level, more abstracted model makes programming a lot easier.

People do not typically leave companies if they are deliriously happy with their roles, but sometimes even people who are reasonably content with their jobs find better opportunities that come along.

A fast chip got faster. This certainly helps to cement their strong current position.

Dell has had a very focused strategy: Don't spend a lot on research and development, leverage the work of others, and deliver cookie-cutter servers and desktops cheaply. That lined up well with the market when saving money was in. But IT managers have to consider more than price now in making buying decisions.

Dell has had a really good run selling essentially cookie-cutter rack-mount servers. But the era of solving problems by just buying more boxes is coming to a close.

A number of companies are starting to look at supercomputing for various data analysis applications.

This isn't a new generation of servers aimed at grabbing a lot of new-to-Sun business. That's what the Galaxy server it just announced is for. But this is a good upgrade.

This isn't even really an open-source issue, although it's received a lot of publicity in that context.

This announcement of yet another no-hassle SPARC system upgrade provides great asset protection to Sun's huge base of applications and customers.

This is about reducing administrative costs more than about security.

This is a bigger story for Apple than for Intel. Apple liked Intel's roadmap better than that of Power PC, and the idea is that they will now sell a higher percentage of notebook computers.

AMD has made some gains in mobile technology, but has been only marginally successful. There is a higher level of integration with notebooks than with desktop PCs, which gives Intel an edge since they are focused on a complete mobile platform.

It's not like Dell had been making investments in Itanium and suddenly decided it wasn't going to do that and pull back its support. It had a relatively older product it wasn't promoting at all, and it really doesn't have a near-term path where it could move forward if it wanted to.

It's a local company with some useful code, engineers who obviously have skills and background and experience in the type of systems management that HP has. We don't know the terms of the deal, but it probably was not all that expensive. It looks to be a good deal for HP.

It's hard to see the Itanium Solutions Alliance accomplishing anything major that the billions already spent by HP, Intel, and the rest haven't. You don't see an x86 Solutions Alliance.

It's probably not possible to be head and shoulders above the competition in the x86 space. There just isn't that much opportunity for being compellingly different. That said, they're nice boxes.

doesn't fundamentally change UltraSparc and Sun's positioning in any way. It's the type of incremental upgrade that computer makers sort of have to deliver.

It's still not clear how visible that will be. I don't think it will be pervasive in 2006.

Sun has given a speed boost to its installed base as the company focuses on x64 systems and the forthcoming Niagara processor,

Intel woke up from its frequency delusion and is now going full-bore on dual-core.

Intel let a competitor sneak into a market where they weren't before. Still, it was hard to believe that AMD would be able to maintain the kind of performance lead that they have enjoyed over the past one to two years.

Intel has shown preliminary benchmarks that backup their general statements about performance. Based on what I have seen, there is not reason to think the products will not live up to their billing.

Intel has inconsistently put energy into storage. It hasn't had a real comprehensive strategy in the past.

In the sense of having this ongoing awareness of and feedback mechanism, it does mirror certain elements of open source development. If not openness of code and standards, there is greater visibility into what's going on. Customers have visibility more than an ability to necessarily influence Microsoft greatly.

I don't think Microsoft had much choice in the matter. The bottom line is there is not a whole lot of market to buy that product from Microsoft.

I don't think anybody knows exactly what the end result of open sourcing the chip technology is.

I assume Scott will paint it as things are back on track and the company is in good hands, but I think the reality is he would have preferred to have waited until the financial results were really there.

If developers don't make the transition to Apple and Intel, they'll probably be making the transition to Windows and Intel. Intel doesn't have a lot to lose.