Replies to post #14388 on Advanced Micro Devices Inc (AMD)

[yourbankruptcy]

Elmer, the stupid thing is that Sparc itself is just a small part of what Sun has. They are not in cpu business. Sparc is just a helper, a sideeffect of their business. The only thing holding Sun with Sparc is binary compatibility and unique instruction set that makes the whole Sun infrostructure a small universe, where many customers live.

I think Solaris and server experience are still distinctive enough to keep Sun running. Adoption of x86 instruction set is painful, but once done, will keep them going. The hardware they make is still distinctive enough to prevent succesful cloning. Noone will buy IBM servers to install Solaris on them.

What makes Opteron strongly preferable to Sun is the total control on the HT bus they have. This way they can keep Opteron as just a processing unit inside the server. The way Intel bus works, technically and legally, is that cpu becomes the core the whole computer is build around. There is nothing that can prevent Sun from having anyone to clone their computers, as they never will have more control on the cpu bus than anyone who has a license. Or even doesn't.

Well, that's my impression.

[sgolds]

yourbankruptcy, Elmer, I essentially agree with Elmer that Sun is in a death spiral. (Actually, I'm more negative on Sun than Elmer.) Yes, they are #3. Wasn't long ago that they were #1 in servers. The drop is secondary to the reasons why:

1. Sun needs premium prices to make their business model work. Use to be that they gave unique solutions to the server market, that is a hard sell now for reasons 2 & 3 below. So reason #1 is that Sun is overpriced for what you get, and server quantities don't have the manufacturing leverage of a top-to-bottom architecture that includes desktops and mobiles.

2. SPARC does not give advantages when compared to a number of other hardware platforms. Used to be that Sun had the best motherboard implementations, with wide DRAM busses and fast I/O. Everyone is doing that now and there aren't any easy ways for Sun to get ahead. The processor is lagging way behind others, also.

3. Sun was always popular with the Unix folks, IS engineers who learned the ins and outs of Unix in college. This made their software popular with their target market, folks who like to tweak the systems using the arcane Unix syntax. Now Linux gives the same arcane interface on top of a microkernel which runs on cheaper, faster hardware. More and more apps are available on Linux. Thus Solaris is being outcompeted by Linux.

Where does Sun go from here?

They could adapt Opteron as Elmer suggests. This would solve the hardware bind, and get them out of a losing manufacturing proposition. Problem: How do the differentiate themselves from every other x86-64 system out there? How do the compete with IBM? Solaris? Doesn't work, see reason #3 above. Thus I don't think that Sun can solve their problems by going to standard hardware - the problem is deeper than that.

Sun is looking a lot like Amdahl, another company that competed well with IBM for a while until their business plan became dated. Sun is still #3 because they have an installed base that does not just go away instantly. That installed base is being eroded daily by IBM, though. We are now at the start of a technology replacement cycle, and Sun is declining quickly because no one in their right mind would switch to Sun now, and current customers are ripe for the picking when they look at updating their systems.