For Geeks only!
>(Most UNIX administrators I've talked to are very unimpressed with
> the token TCP/IP applications that Windows95/NT supply. To get good
> interoperability with UNIX computers you still need to buy a third
> party suite.) All that extra software adds alot of expense;
> X-Windows software still sells for over $200/seat! Moreover, you
> end up with a desktop computer that runs Windows applications very
> well and has less than perfect compatability with your UNIX
> computers.
Compared with UNIX, the TCP/IP stuff by MS is simplistic but sufficient. In any case, you need a good X emulator. "X on Net" I believe it is called works flawlessly with Suns and it can be had for about $50/seat on a site license.
> Demand for the SGI servers should grow as the multimedia and
> information revolution rolls on. As we all know, multimedia files
> keep getting larger and larger; the same is true for databases.
> Even as we speak, these technologies are merging to meet the
> anticipated challenges of information on demand. Even though PC
> disk technology continues to improve and get faster. It will be a
> long time before we have terabyte disks on our desks or the
> combination of CPU Power and I/O bandwidth to search them. We will
> have to rely on clusters of servers for this task. SGI is prime to
> take the lead in this market. Their scalable O2 architecture is a
> leap above their competition and builds on their reputation for
> building the fastest multimedia file servers (as exemplified by the
> Challenge series).
I agree here. As I've said earlier, if SGI doesn't penetrate into the server market they are dead. They have the technology. The Origin 200 and 2000 are great scalable machines. Its up to marketing. DEC was in a similar situation when the Alpha first came out and the blew it. Hopefully SGI will not be another DEC. Marketing will ultimately make or break the company and I also noticed in the latest 10K that they spent a lot of money training a new sales team.
The frustration I have now is that when I have a customer with a dog application running on a Sun or HP and I recommend SGI the usual comment is "They're just a graphics company". SGI IS NOT DOING A VERY GOOD JOB OF GETTING THE WORD OUT! But thats not an entirely fair comment to make either. The other valid argument they have is that they would have to support an entirely different box for one app which is a valid point. Sun and HP have a lot of momentum that SGI has to catch up with.
> The state of the art is always changing. One million polygons/
> second might deliver realistic 3D images on a 15" inch monitor,
> but will it be good enough a two or three years from now? PC's
> are powerful enough to handling video editing for television,
> but what happens if we finally adopt HDTV? I am in a business
> that deals in virtual reality. My development group produces
> research flight simulations for NASA. Beleive it or not, SGI
> Onyx's (with Reality 3 engines) do not quite have the graphics
> horsepower to supply realistic out-the-window images for our
> simulations. We bought an Evans & Sutherland ESIG3000 for that
> purpose. PC's are on the verge of providing realistic 3D rendering
> for show, but not for interaction. It takes alot more graphics
> power to run a realistic 3D world where all the objects are
> independent, i.e. capable of both rotation and translation. The
> question then is, will there be a big enough market for this kind
> of virtual reality and are they willing to pay for high margin
> graphics servers like the Onyx's? Right now, I think its enough
> to sustain the company. In a couple of years, it may be enought to
> grow the company.
I can relate to your demanding simulation app. But that isn't SGI's bread and butter anymore. The current annual report indicates that >50% of the sales are on desktop systems.
The # of polys/sec per 1$ doubles about every 2 years. So in 2 years the boxes now doing $1M polys for say $8K (O2 included) will be doing 2M polys/sec in 1998 and 4M polys/sec in 2000. Pretty hard to imagine. |
