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 : LSI Corporation

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Jock Hutchinson who wrote (17135)2/22/1999 11:53:00 PM
From: Tony Viola  Read Replies (2) of 25814
 
Jock, re IBM system on a chip announcement today, need to hear more about it.

With IBM's technology,
twenty-four million "gates," or circuits -- equal to as much as eight times
the processing and two to four times the memory found on today's typical
PC -- can be packed on a single chip.


In this statement, it's hard to imagine two to four times the memory of today's PCs on a single chip, AND all the processing power they're saying. 2X even "just" 48 megabytes is 768 million bits. Each bit requires its own memory cell. So, how to I get 768 million cells in 24 million gates of space and leave room for processors? The processor part, they say as much as 8X today's is exaggerated, but in the ball park. The Pentium III probably has 5 - 8 million or so gates, so 24 million in IBM's vapor chip might get 4X a Pentium III or so. But then, where's the memory go? I forgot, besides the 768 million cells, the memory needs tons of address registers and decoding, sense amps and other ancillary logic.

One report I heard was that they could go into production in maybe 3 - 5 years. That's a long time. Anyone remember Josephson's Junctions or superconductors? They were always like 3 - 5 years out.

Another report I heard said IBM would target hand held PCs and cell phones. You'd have enough memory to store voicemail messages, lots of them! I'd think so. No mention of desktop PCs. Maybe the technology is much too slow or they'd surely go after the biggest volume product, the desktop. The reason I'm guessing too slow is there's always a speed-power tradeoff. Putting four Pentium IIIs and 96 megabytes or more memory on a chip, and, did you say you want it fast too? I don't care if their copper has negative resistance, that baby would vaporize from the heat. Three to five years down the road, maybe not.

Well, speculating much too far out here. Need more information, like speed mostly. Question about this IBM "breakthrough" went out to Paul Engel tonight, Intel post #74276 tonight. Let's see what Paul comes back with.

Thanks for asking me about it. I hadn't thought about the power/speed part until I started thinking while writing here. Sometimes the give and take here actually works. Whether the idea holds any water remains to be seen.

Tony

Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext