shep@datacube.UUCP (09/13/86)
We have been using Xilinx Logic Cell Arrays (LCAs) in our products for a few months now. LCA technology is very desirable because of its fast turnaround and reasonable density. Because the LCAs are ram based, you can just *download* different configurations and you've changed your design. Conceptually, the gate/functional complexity is somewhere between the highest end PAL devices and your vanilla digital gate-array. The vendor sells a design tool called "Xact" which runs on the IBM-PC <projectile vomiting here> and serves as an "ok" tool to turn the designer's graphic intentions into a bitstream format suitable for loading into the chip(s). Are any other usenet hardware-types using or thinking of using LCA technology? Are there other LCA users who would *LOVE* to see the Xact tool running on a Sun workstation? We've used LCAs for glue collection, address generation, and even see them as candidates for reconfigurable systolic arrays in vector pipelines; what are other people doing with these neat widgets? - waiting for my .lca file to load... Shep Siegel UUCP: [ihnp4 | mirror]!datacube!shep Datacube Inc.; 4 Dearborn Rd.; Peabody, Ma. 01960; 617 535 6644
daveh@cbmvax.cbm.UUCP (Dave Haynie) (09/16/86)
> Nf-ID: #N:datacube:3200011:000:1199 > Nf-From: datacube.UUCP!shep Sep 13 10:32:00 1986 > > > We have been using Xilinx Logic Cell Arrays (LCAs) ... which runs on the > IBM-PC <projectile vomiting here> ... > > Are any other usenet hardware-types using or thinking of using LCA technology? > Are there other LCA users who would *LOVE* to see the Xact tool running > on a Sun workstation? We've used LCAs for glue collection, address generation, > and even see them as candidates for reconfigurable systolic arrays in vector > pipelines; what are other people doing with these neat widgets? > > - waiting for my .lca file to load... > > Shep Siegel UUCP: [ihnp4 | mirror]!datacube!shep > Datacube Inc.; 4 Dearborn Rd.; Peabody, Ma. 01960; 617 535 6644 I haven't had a chance to use them yet, but I'm quite interested in them. I actually had a design in mind about six months before Xilinx announced these LCA that was a perfect application for them. Essentially, I wanted to be able to configure the bulk of a machine's memory map on the fly, in software, replacing what's currently a custom logic array device. The main problems with the LCAs (aside from haveing to use the IBM PC) was their speed at the time and the cost. Both of these factors are getting in the right ballpark for my kinds of applications. I'd really like to be able to program these things on a Sun or Apollo, though the support may have to come from Xilinx; the last time I talked with them, they were planning to be really protective of the internals of the device and how its actually programmed. -- /\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ Dave Haynie {caip,ihnp4,allegra,seismo}!cbmvax!daveh "I gained nothing at all from Supreme Enlightenment, and for that very reason it is called Supreme Enlightenment." -Gotama Buddha These opinions are my own, though for a small fee they be yours too. \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/
shep@datacube.UUCP (09/18/86)
Dave Haynie {caip,ihnp4,allegra,seismo}!cbmvax!daveh writes...
> ...I'd really like to
>be able to program these things on a Sun or Apollo, though the support
>may have to come from Xilinx; the last time I talked with them, they were
>planning to be really protective of the internals of the device and how
>its actually programmed.
We'll you certainly have our vote for pushing Xilinx to port their
Xact tool to the Sun. I don't think they are being "really protective"
about the internals: They just haven't documented the bitstream format
down to the PIP level for the users. It's now public knowledge that
Monolithic Memories (MMI) will second source the Xilinx LCA silicon.
Perhaps between Xilinx and MMI, a groovy s/w design tool that runs on
the Sun will emerge! (FYI: Xilinx telephone is 408-559-7778)
Shep Siegel UUCP: [ihnp4 | mirror]!datacube!shep
Datacube Inc.; 4 Dearborn Rd.; Peabody, Ma. 01960; 617 535 6644
markwm@magic.UUCP (Mark W. Beckner) (09/22/86)
Dave Haynie {caip,ihnp4,allegra,seismo}!cbmvax!daveh writes...
> ...I'd really like to
>be able to program these things on a Sun or Apollo, though the support
>may have to come from Xilinx; the last time I talked with them, they were
>planning to be really protective of the internals of the device and how
>its actually programmed.
When we purchased the LCA development tools for the PC, I had
Xilinx write into the contract that our licenses could be transferred
to future implementations on Suns.
They gave me the impression that porting their tools to UNIX workstations
was in the cards since all the tools (besides SILOS, I think) are
written in C.
Mark Beckner
Bell Communications Research
.