Possible use for rack mounted pci expansion chassis - Printable Version +- hashcat Forum (https://hashcat.net/forum) +-- Forum: Misc (https://hashcat.net/forum/forum-15.html) +--- Forum: Hardware (https://hashcat.net/forum/forum-13.html) +--- Thread: Possible use for rack mounted pci expansion chassis (/thread-6089.html) |
Possible use for rack mounted pci expansion chassis - sneaky_peet - 12-02-2016 I have been searching through some of the posts here looking to see if anyone has ever done a build using a PCI expansion chassis to house the cards and then a separate server to drive them. I know this type of setup is popular for deep learning and protein folding calculations etc but curious if anyone has done this or if its possible for a hashcat build. Thinking of things like these: http://www.ebay.com/itm/Dell-PowerEdge-C410x-Titanium-16-Bay-Chassis-with-4x-1400W-for-Tesla-M2070-M0290-/371516577386?hash=item56801d166a:g:np8AAOSw5IJWfHVg Or maybe two of these: http://www.ebay.com/itm/NVidia-Tesla-S1070-1U-GPU-Computing-System-4x-M1060-C1060-Rackmount-2x-HIC-Cable-/152303856237?hash=item2376046a6d:g:H14AAOSwx2dYGoc0 It would probably take some finagling but might be doable and would be much cheaper than something like the SuperMicro SuperServer or a custom shops charge. Thoughts? RE: Possible use for rack mounted pci expansion chassis - epixoip - 12-02-2016 The chassis you just linked to only work with old passive Tesla GPUs and will not work with desktop GPUs. You can find more comments about expansion chassis in this thread: https://hashcat.net/forum/thread-5064.html RE: Possible use for rack mounted pci expansion chassis - sneaky_peet - 12-02-2016 (12-02-2016, 09:13 PM)epixoip Wrote: The chassis you just linked to only work with old passive Tesla GPUs and will not work with desktop GPUs. ah thanks for the reply. Seems like the hardware to just house and power a bunch of GPUs should exist and be reasonable priced but I guess it just doesn't. RE: Possible use for rack mounted pci expansion chassis - epixoip - 12-03-2016 Hardware to house and power a bunch of GPUs is expensive to produce, and it's not exactly high-demand so few units are produced in each production run, which means the prices are higher than typical commodity hardware. |