Interesting Xilinx article on using FPGAs with LED displays

[leon_heller]

The latest issue of the Xilinx Xcell Journal has an interesting article on using FPGAs to control large LED displays:

http://cde.cerosmedia.com/1Z4bb08866af93c427.cde

See page 16.

Perhaps someone from XMOS should contact D3, if they haven't already done so.

[otitov]

thanks for pointing it out - interesting reading!

[leon_heller]

I think that several of those applications might be suited to XMOS devices.

[shawn]

The Xmos cores and FPGA's are a symbiosis.
Xilinx makes affordable Spartans and there
just like Virtex minus speed, I/O, and a few
resources. It's worth noting that the Virtex
are expensive and the minimum size Virtex
starts at 75KLut's. If you want Virtex on a
budget there is a very decent e-bay market.

If you want a fast FPGA, really!, look at the
Lattice ECP's and check see there SC's. So
so expensive, so nice, the SC's, don't even
look.

Actel's anti-fuse are very fast and affordable.
Anti-fuse is expensive but approachable under
500KSystem gates 12kluts, very low latency,
tighley coupled dense gates. Partnered with
their /Smart/Fusion/ProAsic's are quit a treat
and represent a elegant solution at a descent
scale/$/speed.

I think the Actel's are one of the best choices
because they offer the best solution for volitility.
That puts the silicon in more real world apps, now.
On the the other hand @ 150KLut's for $200/chip
Xilinx Spartan's should satisfy. Xmos can replace
most FPGA's functions and are Nonvolatile making
for instant on and instant market. FPGA's are not
a good substitute for an Xmos core. the Xmos core
is a grand model for inspired soft-cores.