home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!dtix!darwin.sura.net!gatech!emory!rsiatl!exnet!dhd
- From: dhd@exnet.co.uk (Damon)
- Newsgroups: comp.arch
- Subject: Re: Re : Single Transistor Clock - ALPHA
- Message-ID: <BuFAHJ.5L0@exnet.co.uk>
- Date: 11 Sep 92 16:44:06 GMT
- References: <1992Sep8.091816.22659@rkna50.riken.go.jp> <1992Sep8.201415.8422@cs.yale.edu>
- Organization: ExNet Systems Ltd Public Access News, London, UK
- Lines: 30
-
- In article <1992Sep8.201415.8422@cs.yale.edu> ewing-martin@yale.edu writes:
- >
- >In article <1992Sep8.091816.22659@rkna50.riken.go.jp>, wong@rkna50.riken.go.jp (Wong Weng Fai) writes:
- >...
- >|>
- >|> 1. I guess whether its a single transistor or not depends on how you
- >|> count. But is this huge "thang" responsible for the infamous "oven-toaster"
- >|> effect of the 21064 ? Will pumping the clock speed higher melt the chip ;-) ?
- >|>
- >
- >In CMOS and other high-impedance technologies, the transistors are driving
- >mainly capacitive loads. Each change of state requires Q=1/2 CV**2
- >increment of energy, and the current and power drain scale pretty much
- >linearly with clock rate. (Conversely the drain goes nearly to zero at
- >low clock rate.)
- >
- >The melting point of Si is pretty high, but the thought of a CPU glowing
- >cherry-red appeals to me. (The days of 304 TLs and other big glass RF power
- >tubes are gone forever. :-)
-
- I'd have thought the possibility of increasing the diffusion speed of
- the dopants to something that would kill the chip quickly, or of
- buckling and cracking the chip, would be much higher...
-
- Damon
- --
- Damon Hart-Davis | Tel/Fax: +44 81 755 0077 |1.24|| Cheap Sun eqpt available.
- Internet: dhd@exnet.co.uk | Also: Damon@ed.ac.uk || US high-perf motor groups.
- --------------------------+----------------------++ >12 mail&news polls / day.
- Public access UNIX (Suns), news and mail for #5 per month. FIRST MONTH FREE.
-