yrdbrd@bmcg.UUCP (Larry J. Huntley) (06/21/85)
>> larryh@tekcbi > David Wolpert >>...at >>a rate of approximately 10 Million Forth primitives per second (at a clock >>rate of 8 MHz). > Amazing! > (a.k.a. Unbelievable!) >>...the >>following must be taken with a killogram of NaCl.... > So, I think, should the preceding. > Unless I am missing something, even an *extremely* complex processor > could execute only 8 "Million Forth primitives per second (at a clock > rate of 8 MHz)." >David Wolpert > Hewlett-Packard Company, Instrument Systems Lab > P O Box 301 - Loveland, Colorado - 80539 (USMail) > {*!}{hp*!}hpisla!wolpert (un*x) Ahem. You *are* missing something, David. The fact is that the Novix chip can execute a subroutine return in 0 cycles. The implication is that the final instruction in a subroutine can be executed as the return is being taken and you get two instructions for the (clock) price of one. Since Forth programs are strings of subroutine calls (and subsequent returns) and the return constitutes a Forth primitive (EXIT, I think), it is possible to get 10 MFP/S from an 8-MHz machine. The Novix NC4000P also allows simultaneous access to more than one bus (it has 5) at a time so that operations like reading from one bus and writing to another in one cycle are possible. Even if you don't happen to be interested in Forth for it's own virtues, the NC4000P represents an interesting architecture. Call Novix and ask for their information package (408/996-9363); it makes interesting reading. One other thing --- the 8 Mhz engine is a gate-array prototype. A full custom version is in the works which is expected to perform an order of magnitude faster. 100 Million Forth Primitives per Second, anyone? Can you say "Like a bat out of Hell?" I knew you could. 'Bye, boys and girls... 'brd -- Larry J. Huntley Burroughs -(B)- Corporation Advanced Systems Group MS-703 10850 Via Frontera San Diego, CA 92128 (619) 485-4544 -*- "Dear Friends, It's 'ORGAN LEROY' at his organ again." -*-
henry@utzoo.UUCP (Henry Spencer) (06/24/85)
> One other thing --- the 8 Mhz engine is a gate-array prototype. A full > custom version is in the works which is expected to perform an order of > magnitude faster. 100 Million Forth Primitives per Second, anyone? > Can you say "Like a bat out of Hell?" ... Yes, and I can also say "damned expensive memory or lots of wait states"! -- Henry Spencer @ U of Toronto Zoology {allegra,ihnp4,linus,decvax}!utzoo!henry
yrdbrd@bmcg.UUCP (Larry J. Huntley) (06/27/85)
In article <5721@utzoo.UUCP> henry@utzoo.UUCP (Henry Spencer) writes: >> .... 100 Million Forth Primitives per Second, anyone? >> Can you say "Like a bat out of Hell?" ... > >Yes, and I can also say "damned expensive memory or lots of wait states"! >-- > Henry Spencer @ U of Toronto Zoology Come on, Henry; say something nice about *something* just once, huh? At least this chip isn't made by Intel, so that should please you. It does, however, use 64K segments, sort of. -Sigh- I guess you can't please some of the people all of the time. (Aside...) Hey, Forth Freaks -- let's try and develop a suitable punishment for the first idiot who builds a machine using the Novix chip that is "IBM PC Compatible." Boiling in lard? 24-pin DIPs under the finger- nails? Walking on hot glowing heat-sinks? Send in your suggestions today! { .25[ :-)] } 'brd -- Larry J. Huntley Burroughs -(B)- Corporation Advanced Systems Group MS-703 10850 Via Frontera San Diego, CA 92128 (619) 485-4544 -*- "Dear Friends, It's 'ORGAN LEROY' at his organ again." -*-
jordan@greipa.UUCP (Jordan K. Hubbard) (06/29/85)
Has anyone produced a machine based on this chip yet?
Has anyone out there built a prototype?
--
Jordan K. Hubbard
@ Genstar Rental Electronics.
Palo Alto, CA.
{pesnta, decwrl, dual, pyramid}!greipa!jordan
"ack pfffft. gag. retch. barf.. ack"
- Bill again.henry@utzoo.UUCP (Henry Spencer) (06/30/85)
> Come on, Henry; say something nice about *something* just once, huh? > At least this chip isn't made by Intel, so that should please you. It > does, however, use 64K segments, sort of. -Sigh- I guess you can't > please some of the people all of the time. Who, me, hard to please? What a ridiculous idea! This is a preposterous suggestion!! Rave, rant, froth, foam... :-):-):-) Believe it or not, there actually are machines I don't have complaints about. A few. The Berkeley RISC design is one; I admire it a lot. I was groping towards something similar myself six or seven years ago (as were others), but I never did solve some problems that the Berkeley folks have handled very nicely. I wish it was commercially available, with the necessary support goodies. -- Henry Spencer @ U of Toronto Zoology {allegra,ihnp4,linus,decvax}!utzoo!henry
toma@tekchips.UUCP (Tom Almy) (07/01/85)
In article <1718@bmcg.UUCP> yrdbrd@bmcg.UUCP (Larry J. Huntley) writes: >[...] let's try and develop a suitable punishment for >the first idiot who builds a machine using the Novix chip that is >"IBM PC Compatible." [...] Well, The Novix "Beta-Board" evaluation board requires an IBM-PC to use. So we didn't have to wait long!
magik@wlcrjs.UUCP (Ben Liberman) (07/02/85)
In article <260@greipa.UUCP> jordan@greipa.UUCP (Jordan K. Hubbard) writes: > >Has anyone produced a machine based on this chip yet? >Has anyone out there built a prototype? Novix has a monoboard called "Beta Board" at ~$3500 that uses a PC or clone as a terminal and file server. They say that a standalone is FORTHcomming :-) If anyone wants serious info on the chip or board, drop me a note and I'll post their address, phone # etc. -- ----------------------------------------- Ben Liberman {ihnp4|ihldt}!wlcrjs!magik
steve@kontron.UUCP (Steve McIntosh) (07/02/85)
> Has anyone produced a machine based on this chip yet? > Has anyone out there built a prototype? > -- Jordan K. Hubbard Novix has been marketing a demo/evaluation board that needs an IBM PC for a development and control machine. This board implements the chip with gate arrays. Last I heard, there were 32 chips in existance (one wafer worth, I guess). Novix has been good about letting people know how close the chip is to being produced in quantity. -- Steve McIntosh, Kontron electronics, Irvine CA