SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Silicon Graphics, Inc. (SGI) -- Ignore unavailable to you. Want to Upgrade?


To: Esvida who wrote (3496)11/8/1997 7:31:00 PM
From: John M. Zulauf  Respond to of 14451
 
> Are you making a case for SGI the niche player?

Actually I'm saying what I think I heard in the financial analysts conference call, that until the NT products show up, sometime in the next fiscal year, that the high-server business should carry the company as the low-end UNIX workstation revenue flattens out. Given the Origin 2000's technologic advantages, new aggressive pricing, new agressive marketing in that space, management seems very serious about this.

People talk causally about the O2000's scalability, but it is a truly astounding platform. You can start off with a roughly $40K deskside module, and just by adding more modules (no system replacement) scale to a 128 processor (NCSA has one running) system. In a deskside module, you can have 8 processors and 16GB of memory (addressable as a single unified space) -- WAH! I think only SGI and DEC can has 64 bit memory addressing to allow that sort of thing. Multiply that by 16 modules and that's 128 processors and 256GB of RAM -- all by just adding modules.

It's mind boggling, when you include that the I/O and crossection bandwith grows right along with the additional modules. The architecture is scoped for 1024 processors, but I haven't any idea for the time frame to go beyond 128 processor systems. Certainly they can be clustered today, so even 128 isn't a top number (anybody know the supported cluster size?) This thing is the Borg cube of servers.

BTW, my understanding is that many thousands have been solid in the first year. When consider the expandability of the system, the "sell more modules" to the installed based should start helping out the bottom line as well.

SGI isn't going to stay a niche player from what I hear in their public statements. But the "niche" products they have should hold them through the sea change ($657M in cash can't hurt either).

Unofficially (and long),

john