Silicon Investor (SI) -- The First Internet Community

STOCKTALK

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 : All About Sun Microsystems -- Ignore unavailable to you. Want to Upgrade?

To: nihil who wrote (28561)	3/5/2000 10:20:00 AM
From: rudedog	Read Replies (1) \| Respond to of 64865

nihil - re: did all of heavy computation for ILLIAC IV simulators Now you're really bringing back memories Did you know Jack Lipovsky at U of I? I tracked his work on a design for a switch fabric using Motorola 6800 8-bit microprocessors as the compute elements - a full crossbar design... I think that was maybe in '76 or '77. I used some of the concepts later in a design to preprocess digital images using 1024 6800s, one processor per pixel stream, but that was finally implemented in 2900 bit slice and fed an array processor, so it hardly counted as a "computer". The BSD version of Unix for the PDP 11/70 was the one we used to help students learn operating system design principles. The target machines were PDP8s with almost no memory - most had only 8K - and getting a kernel design which could both load and actually run something was a good exercise. We initially had 20 teletypes and 5 PDP8 target machines in 1976 - the teletypes were 33s and were handy because they had a paper tape punch right on the terminal which could then download to the pdp8. The program was so over-subscribed that it was expanded to 100 terminals and 20 PDP8s the next year, and still students were lined up around the clock. Ken Olsen hardly understood computers at all - he understood power supply design and packaging, and how to get heat out of a cabinet but hardware and software architecture was beyond him. He made a number of bad calls - and I think was as much responsible for running DEC into the ditch as anyone. He could have had the franchise for Unix, he could have had a lock on PCs, which were after all a logical extension of the pervasive low-cost computing revolution that DEC had started, he could have been a leader in networking - instead, he tried to "kill" all of those movements to preserve his proprietary lock on a revenue stream. In some ways this is like what's going on at MSFT today... DEC thought that they had driven a revolution in the way people thought about computers, when really they were just at the right place and time with an enabling product. As a result they were trying to trip up the other runners when they needed to run as fast as they could. If DEC had shifted hard over to Unix as a strong alternative to RSTS for PDP11, with real engineering support instead of the lame Ultrix skunk works effort, there would have been more momentum to do the same kind of thing on the VAX architecture, and the network infrastructure that DEC was developing for the VAX could have been more closely linked with the ARPA work, which might have enabled broader connectivity maybe 5 years earlier than what actually happened. I think that Gordon Bell's heart attack and subsequent retirement in the mid '80s removed the real engineering drive from DEC and I date their decline as an important influence from about that point. No one else had the presence to drive company direction against Olsen's more misguided impulses.