40 Computers waiting for Tablegeneration!!

[Reaper]

Hi Bitweasel, i solved the problem.

Today i got 40 "old" universitycomputers. Every has a NVIDIA 9800 GX2 in it. I bought them with 2 friends for 7000 € and we are interesting into GPU Rainbowtables. We are new in the Rainbowtable-Cryptanalysis and want to improve our skills. Maybe we can help you with your NTLM 8 full US charset if you want our help.

Reaper

[Bitweasil]

I will see what I can do to help with distributed compute I really need to get this BOINC'd...

[Salazasu]

Well...(I don't really want to steal your thunder Bitweasil )

http://www.freerainbowtables.com just brought their BOINC server back online, and they're generating tables again.
So, until Bitweasil starts a distributed project, you could contribute to this one.

Although you might get a real shocker when the power bill comes!

[Reaper]

Hi Salayasu

I am interested in GPU-Rainbowcracker and not in the CPU based rainbowcracker. I think the GPU tables are faster as the freerainbow tables and have more potential.

[Bitweasil]

I'm working on refactoring the GPU RT code to make it more usable and theoretically Windows friendly. Hopefully I will have time over this winter break to finish this and start on a BOINC setup, or at least something usable that lets people upload workunits.

[Salazasu]

Hi Salayasu

I am interested in GPU-Rainbowcracker and not in the CPU based rainbowcracker. I think the GPU tables are faster as the freerainbow tables and have more potential.

Well, the freerainbow tables are generated with CUDA. The tool for using those tables, rcracki_mt, doesn't currently support GPU-based cracking but hopefully will soon.

Though I must admit, a distributed brute-forcer would be, *ahem*, fascinating...

[Bitweasil]

Though I must admit, a distributed brute-forcer would be, *ahem*, fascinating...

Next version of the Multiforcer will support network distribution of work

[quel]

Hi Salayasu

I am interested in GPU-Rainbowcracker and not in the CPU based rainbowcracker. I think the GPU tables are faster as the freerainbow tables and have more potential.

Well, the freerainbow tables are generated with CUDA. The tool for using those tables, rcracki_mt, doesn't currently support GPU-based cracking but hopefully will soon.

Indeed after the sse2/3 items get into distrrtgen, rti2 header format is complete, then I'll be backporting the sse2/3 and cuda items into rcracki_mt. For ia32 linux and win32 I already have an improvement for the cpu distrrtgen for the reduction function that results in about 60% less work for a given WU (this isn't even sse but a precursor.) That improvement will of course make a big difference in rcracki_mt as well.