The RTX-2000 was an of shoot of the NC4000. Even at 10MHz, they could out compute a 40MHz 80386. One execution per clock cycle plus possibly using 3 16 bit busses in a single cycle. A 4MHz NC4000 could sort 1K 16 bit values in 19.7 milliseconds. Dwight
________________________________________ From: cctalk <cctalk-boun...@classiccmp.org> on behalf of Toby Thain <t...@telegraphics.com.au> Sent: Wednesday, April 20, 2016 10:35 AM To: General Discussion: On-Topic and Off-Topic Posts Subject: Harris RTX-2000 - Re: High performance coprocessor boards of the 80s and 90s On 2016-04-20 1:28 PM, dwight wrote: > There was a Harris RTX-2000 based accelerator card around > the 80386 time period. I hadn't even heard of that chip :/ Interestingly: "The RTX 2000 is specifically designed to execute the Forth language" (https://users.ece.cmu.edu/~koopman/stack_computers/sec4_5.html) --Toby > Dwight > > > ________________________________________ > From: cctalk <cctalk-boun...@classiccmp.org> on behalf of Ali > <cct...@ibm51xx.net> > Sent: Wednesday, April 20, 2016 10:04 AM > To: 'General Discussion: On-Topic and Off-Topic Posts' > Subject: RE: High performance coprocessor boards of the 80s and 90s - was Re: > SGI ONYX > >>>> I'm changing the subject because the subject of RISC coprocessor >>>> boards has already been interesting to me; I owned the NuBus Levco >>>> Translink II (for Mac II family) with four TRAM slots for >> transputers. >>>> > > I never had much run in with these kinds of boards as they were geared > toward very specific markets. However, the Intel i860 and i960 did make it > into some consumer level boards. I have a MCA SCSI RAID Controller somewhere > that uses an i960 for coprocessing... Of course that may be a bit too new > for CCtalk.... > >