test-drb May 2001

test-drb@ccmp.vtda.org

273 participants
642 discussions

by mcguire＠neurotica.com

On May 4, Clint Wolff (VAX collector) wrote: > In terms I understand, what's a J90, and what's it worth? A used, operational J90 is worth anywhere from ~$8K-35K depending on configuration. Here's some quick info: Type: PVP (parallel vector processor) CPUs: 4 min, 32 max CPU architecture: YMP family binary compatible Memory: 4GB max Disk: unbelievably huge quantities possible Performance: 200MFLOPS sustained per CPU Word width: 64 bits FP capabilities: 64 bits single-precision, goes up from there Vector element depth: 64 elements Cooling: air Power: 220V split-phase, about 3-4KW depending on config OS: Unicos (SystemV variant) Languages: C, C++, Fortran77, Fortran90, others I have a relatively large J90 here. Pics of it can be seen at http://www.neurotica.com/cray/mine. I absolutely love this machine. It is, for lack of better descriptive terms, UNCOMFORTABLY fast when crunching numbers. -Dave McGuire

24 years

Back to booting that MicroVAX...

by ajp166＠bellatlantic.net

From: healyzh(a)aracnet.com <healyzh(a)aracnet.com> > >I don't think the VAXstation 3100/10 is later than a M76, and it's got that >oddball connector. I thought it was the other way around. Also either the >DECstation 2100's and/or 3100's used the same cable. The VAXstation if memory serves is later. The MV and DECserver3100s were SCSI-II and before the M76 though the difference was less than a year. Allison

24 years

How many transistors in the 6502 processor?

by mcguire＠neurotica.com

On May 4, healyzh(a)aracnet.com wrote: > > >Is anyone else having a problem with mails coming from > > >ard(a)p850ug1.demon.co.uk? They come out blank on my end. > > >Bob Mason > > > > Yes. That is Tony (Hi Tony!) and for some reason they come out in Opera 5.x > > as just an attachment that is unopenable (strange but his are the only ones). > > --Chuck > > Strange, they're just fine in 'elm'. Last I checked they were also fine > under Eudora 4.0.2/Mac. Fine in VM under xemacs here.. -Dave McGuire

24 years

Tony's mail (was Re: How many transistors in the 6502 processor?)

by fmc＠reanimators.org

knightstalkerbob(a)netscape.net (Bob Mason) wrote: > Is anyone else having a problem with mails coming from > ard(a)p850ug1.demon.co.uk? They come out blank on my end. No. But Tony's mailer is inserting a header Content-Type: text ...which isn't valid according the the MIME spec, but then there isn't a MIME-Version header either so it your mailer shouldn't be trying to interpret it as a MIME content-type. -Frank McConnell

24 years

KXT11 Variants?

by ken＠seefried.com

Can anyone tell me what the differences are between a KXT11-A, -AA, -AB * -CA? I know -A? is dual wide, and -C? is quad wide, but other than that? BTW...how would one get Micropower Pascal for one of these? Ken Seefried, CISSP

24 years

Apple II IDE interface

by celt＠chisp.net

Found this on usenet. Someone has kludged up an IDE interface for the Apple II (!). Schematics and software for the interface can be found at: http://s.guillard.free.fr Apparently the thing costs about $5 in parts to put together. - Mike

24 years

Latest find: Victor Comptometer, model 12E.523.157.C need power cord

by ethan_dicks＠yahoo.com

OK... from visiting various Comptometer sites (including a museum by Victor themselves), this unit appears to be c. 1961, and is powered by a 100-130V motor. The power cord is like nothing I've seen - three brass rods, the middle one (earth), fixed, the outer two slightly floating, about 1.25 cm long, 2-3 mm in diameter, spaced at about 1.25 cm on centers, nearly, but not entirely, in line. The outer two conductors are slightly farther apart than a modern Continental 220V plug. I can tap in power internally, I can maybe use insulated alligator clips, but mostly, I'd like to find a proper mains cord for this. From manually operating the innards, it needs cleaned and oiled, but that's to be expected. -ethan ===== Visit "The Seventh Continent" http://penguincentral.com/penguincentral.html __________________________________________________ Do You Yahoo!? Yahoo! Auctions - buy the things you want at great prices http://auctions.yahoo.com/

24 years

How many transistors in the 6502 processor?

by ajp166＠bellatlantic.net

From: jpero(a)sympatico.ca <jpero(a)sympatico.ca> > From: Brian Chase <bdc(a)world.std.com> > > And then it'd be rather fun to implement your very own 6502 using 74* > series logic chips. >Possible, but length of traces to wire all those together will keep >it to KHz range and needs few large boards that needs so much power >that hottest athlon cpu is low power by comparsion. It would eat power but it would not be that slow. The mos logic of the time was quite slow compared to the ttl of the time. I did do a 2901C version of z80 and it was much faster than the MOS version, not due to the 2901s as they were logic savers but the acutal speed bottleneck as well. I have seen photos of apple IIe prototype laid out on early pre-production logic board, one of that 40 pin chipset socket is hooked to a equally same size board filled to the 4 edges with tightly packed TTLs and few large ICs, I think it was shown in Byte as well as few other publicatons. Large board full of chips vs. 40 pin IC, same thing...Amazing! That was replaced by a 2000 or so gate array so thats were it all went. Also the proto was built of all SSI TTL making it less dense. I did some work like that in the early 80s. Logic was faster than you think. It's true to get to the sub 100ns range for serious cpus you have to get the interconnect problem out of the way. However for a fairly fast machine (4-10mhz clocks) there are many examples that are mostly ttl. For starters the PDP-8E ca. 1971ish executes instructions in 1.4uS for the base modes and that was SSI ttl on low density boards. FYI it was limited by the core cycle time not the TTL, I ran one at nearly twice speed easily with semicon memory. Allison

24 years

DEC finds and wants

by broth＠heathers.stdio.com

OK. It looks like it will take a while to sort through this stuff from my last haul but I will post some of the more interesting items(to me anyway) as I run across them. A few of the items were from Bell Atlantic(Sorbus) and deal with maintenence and diagnostics. One box I found included a TK50, RX50, and 9-track tape diagnostic software. I know there is a manual that goes with this in the pile somewhere as well. The media is labled "Pro-Cure/OSD V1.3" Anyone familiar with this? I also have found many diagnostic dongles for serial/ethernet diags. Also lots of color sales brochures on pretty much the whole VAX line. I should have my website back up in a week or so completely redesigned. I will slowly be adding pictures ands scans from many pieces in my collection. I am inviting any of the 'DEC archivers' from the list to feel free to contact me if I can fill in any of the holes in software or docs. Recent additions are a VAX 4000/200 (this weekend) and beleive it or not I think I found a CR11. If anyone can help me, I am still looking for a front panel for my DSSI microVAX 3800. Someone on the list said they had one but I can't remember who. Also I need a front panel for a PDP8a. Unfortunately it was wiped out during the trip back. Brian. -- Brian Roth - System Administrator www.webwirz.com - Old Computer Repository Preoccupation is my main occupation.....

24 years

How many transistors in the 6502 processor?

by mcguire＠neurotica.com

On May 4, Richard Erlacher wrote: > My first prototype of my all-digital clock-data separator circuitry was > wire-wrapped and operated nicely at 80 MHz. The speeds mentioned by Emanuel are > certainly realistic, and certainly not the upper end of what was done with > wire-wrapped high-speed logic. I have to agree with Richard here. Back in 1986/87, I worked on the Navier-Stokes Supercomputer project at Princeton University...the prototype processors for that machine were wire-wrapped (most of the boards by me! 8-)) and signals ran between 20-30MHz. We had no problems at all. We used twisted-pair wiring for some of the clock lines as I recall, but the vast majority of the wiring was simple point-to-point wirewrap with no special routing considerations of any kind. Don't knock wire-wrapping...it's fast, easy, works very well, and properly done and cared for, it will last for years. And if you're a weirdo like me (and I know some of you are) it can even be FUN! 8-) As an interesting side note, that supercomputer (the project is still underway from what I'm told, undergoing various refinements to keep pace with advancing technology) is built on Unibus-form-factor boards, using PDP11/24-style expansion chassis. It was a cheap and easy way to get nice, compact chassis with high-quality power supplies and very well-known (and large) board geometries. -Dave McGuire

24 years

← Newer
1
...
55
56
57
58
59
60
61
...
65
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

test-drb May 2001