Re:Lisp as the Maxwell Equations of Software (2012)
by GregBuchholz on Thursday February 12, 2015 @05:06PM
(#9041680)
I'm not saying it is economically competitive. (It is possible pay over $25k for a really high end FPGA) And if you are just synthesizing a general purpose synchronous CPU you definitely not going to get a lot of bang-for-the-buck, because you are going against the grain of what an FPGA can provide. In that instance you're just vetting your design until you convert it over to a mask-programmable "equivalent", or do a full-custom design. The interesting things about an FPGA would be to use its inherent parallelism, fine grain programmability, and the reprogramability to run circles around something constrained by a von Neumann bottleneck in the cache hierarchy.
As to clock speeds, here's part of the abstract to a white paper that might interest you:
"A clock rate higher than 500 MHz can be supported
on a mid-speed grade 7 series device with almost
100% of the logic slices, more than 90% of the DSP48
slices, and 70% of the block RAMs utilized. This
requires the designer to follow some rather simple
design rules that cover both algorithmic and
implementation aspects. These rules are reviewed in
the paper."
...but clock speed isn't necessarily a super interesting factor if your data bus is 2048 bits wide, with a pipline 100 stages deep, comparing to say 64 bits wide and 10 stages deep on a CPU. Again, this is not to say that anyone should try implementing a Lisp machine on an FPGA to try to take market share away from Intel.
As to clock speeds, here's part of the abstract to a white paper that might interest you:
"A clock rate higher than 500 MHz can be supported on a mid-speed grade 7 series device with almost 100% of the logic slices, more than 90% of the DSP48 slices, and 70% of the block RAMs utilized. This requires the designer to follow some rather simple design rules that cover both algorithmic and implementation aspects. These rules are reviewed in the paper."
http://www.xilinx.com/support/documentation/white_papers/wp4...
...but clock speed isn't necessarily a super interesting factor if your data bus is 2048 bits wide, with a pipline 100 stages deep, comparing to say 64 bits wide and 10 stages deep on a CPU. Again, this is not to say that anyone should try implementing a Lisp machine on an FPGA to try to take market share away from Intel.