[Hardware] Fully unrolled RC5 on FPGA

[Gerald Richter]

Ben Payne wrote:

> (I forget what the number is).  

127 devices including hubs iirc, it's actually 128 but the "root hub" in 
the controller eats id 0... :)

> But maybe we could build a small board that had a daisy chaining
> capability.  So the first node would be the master node and it would be
> connected via USB to the PC.  Each board would have a in and out
> connection for daisy chaining.  This would be some type of short range
> serial or parallel bus.  We'd probably have to limit the number of nodes
> in a chain to keep the bandwidth required simple.  This would allow us
> to get several nodes on a single USB connection.  

Erg, i've always been a not fan of kludges like that. They work, and 
very well for some things, but just ugh. Gimme a full length PCI card 
with a bridge chip, whatever other support logic is needed and sockets 
for like 8 chips and i'm happy. I dunno, how big are the spartan 4? 
Could we fit more spartan 3s in the space?

The reason for the sockets being so I could get the board with like 1 on 
it and get going and add chips as I could afford them. I'm more or less 
just an end user here with some odd knowledge, don't have a lot of money 
to spend on this project or a hobby bench to set external crunchers up 
on (I live in a studio apt w/o much space to spare)... But I want to 
contribute.

> I've been thinking that very small bandwidth but that depends on how
> flexible we want this platform to be.  It sounds like at the rates that
> Guerric is getting we'd need about 1 block of keys per second to a node.
> Since a block of keys is just the first key and the length this is very
> low bandwidth.  However I'm thinking that we should plan for something
> like 20-50 k bits/s, so that we have flexibility.  I'm hoping to hear
> comments on this assumption.    

Nearly every idea I like has overkill bandwidth...

Now for some off the wall - someone mentioned something about a pmc 
style mount for the chip, which someone kinda shot down that embedded is 
expensive. A cheap 3 slot VME chassis appears on ebay for ~ $30 shipped. 
For $100 more you can go to 20 slots. Those are doable price points for 
a chassis... the slot mechanics shouldn't be hard to build a board for, 
and to heck with the cover plate and insert/remove tabs (unless they are 
cheap, this is after all a hobby approach.) If I could get a set of 
boards that I could just stuff into a chassies like that with a single 
bridge to a host computer, with sockets to take spartan 4s (or whatever 
good chip, not gonna be picky here) as I can afford to add them then i'd 
be happy. I wouldn't even really care how the bridge occurred - single 
20mbps usb to the host? Cool by me. VME to PCI bridge using a slot in 
the host computer? Cool too. That's 2 cables and some software fiddling 
and I can have a hardware cracker that is really contributing.

And to be fair, yes a larger board design is more expensive and more of 
a pain, but not being familiar with that side of things, it'd be nice to 
have some enlightened commentary on the looney idea. I would love to 
have a cute little 3 slot VME chassis sitting beside my computer quietly 
nibbling away on RC5 or whatever contest we opt to do.

One of the advantages to going with programables is they could be 
reconfigured for other contests, yes? In that case, could someone fill 
in the details on the difficulties on taking such a cruncher array and 
reconfiguring it after -72 closes? :)

--Gerald