Originally posted by rmendis
No i'm not talking about Macs not being in those markets.
Of course they are...with DESKTOPS and WORKSTATIONS.
Not as BIG IRON - mainframe class machines - superservers and superclusters. I think i said BIG IRON WITH AN APPLE LOGO.
Well, there is that little (*cough*) cluster at UCLA...
My point was that Apple already has a presence in those markets because of their utility to science. They're not used frivolously or lightly now
. The fact that Apple doesn't make big iron is actually a point in their favor, because big science is moving away from big iron, and toward Beowulf clusters and the like - except that those are a pain to set up and maintain. Enter Apple, a company known for reliable machines with tremendous ease of setup and transparent networking. They have all the credibility they need in all the right areas right now.
SGI, on the other hand, is dying. I don't know how valuable a name associated with $20,000 workstations and commercial-van-sized supercomputers is to a market that's more interested in clustering racks full of, say, Xserves. Look at the cluster named Green Destiny: It doesn't eat lots of electricity, doesn't require pulling special cabling in or any sort of environmental control beyond a stout air conditioner, it was cheap to build with out of off-the-shelf parts, etc. That's the future of high-end computing. Now, imagine the same basic idea, but with Apple's plug-and-play.
Apple underwent a soul searching mission in the late 80s early 90s Scully Era that almost destroyed Apple - he tried to turn the Apple brand into something it wasn't and something it could never be - a business brand.
Yup. I had a "Professional Macintosh" once. The 8600/200. Solidly built (and how!) but bland. That's the wrong tangent, I agree. The Xserve is much closer to the right idea: Make the function
attractive to your customer, and keep the style and flair pure Apple.
It is remarkable that Apple produced the Xserve, especially given that Steve Jobs had stated in the past that Apple would "never" produce servers. The reason is the one i mention...Apple's brand is now firmly established in the hearts and minds of consumers around the world as a premium commodity brand. And one that sells Macs.
Also, servers just aren't what they used to be - in a good way. They used to be more remote, before networking got as fast as it is. They used to be headless, or command-line-oriented; or alternately, they required expensive clients (we spent $10K on an Xterm, back in the day), and they ran a different class of operating system.
Now, because of the shift in technology, servers are easy and cheap and convenient and common, and because of OS X, an iBook
is a credible server. The whole landscape has changed. Clustering is crowding out big iron at the high end, and all this plays directly to Apple's long-cultivated strengths. (I remember a network in college that was a bunch of toaster Macs linked up via serial Appletalk to a Mac II running a client/server version of Word. It was slow, but it worked and it worked well. And this was, oh, 1988-89, I think?).
So not only will it be able to sell to govt., research/edu and media companies desktops and workstations but ALSO high end super clusters and render farms as SGI.
Ah, but desktops are fast becoming workstations - heck, some laptops
are credible workstations - and they're also being recruited for server farms and clusters, scaling as high as you please. Big iron will still be useful for other things (a cluster of Macs or Linux boxes can't hope to replace the AS/400 in the basement of an insurance company) but those markets are of much less interest to Apple.