Thank you Eric, this is very good to know! We'll probably stick with something similar to our current specifications in this case.
Best, -Will On Friday, October 6, 2023 at 1:35:27 AM UTC-4 Eric Deutsch wrote: > Hi Will, thanks for the question. There no easy answer. An SSD is good > because there’s a lot of disk I/O processing data. Comet itself is pretty > thrifty with memory, so you won’t need lots of memory for Comet. The > Prophets can use a lot of memory if you process huge experiments at once, > although they can often be broken up into pieces. Many of our cluster nodes > have only 96 GB of RAM and do fine for most datasets. I would probably > spend more on cores than on RAM. The machine you list there with 64 cores > and 128 GB is a good choice. If you can spend more than that, I think I > would increase the core count over more than 128 GB. A GPU won’t help for > Comet and the Prophets, so I would save money and go without the GPU. > > > > Regards, > > Eric > > > > > > > > *From:* spctools...@googlegroups.com <spctools...@googlegroups.com> *On > Behalf Of *Will Comstock > *Sent:* Wednesday, October 4, 2023 10:57 AM > *To:* spctools-discuss <spctools...@googlegroups.com> > *Subject:* [spctools-discuss] Optimal hardware for TPP server? > > > > Hi all, > > > > Kind of a naive tech question, but my lab is going to purchase a Linux > server to run the TPP and we're trying to decide on a hardware > configuration that would let us run Comet, PeptideProphet, PTM Prophet, > XPRESS, Libra, and a few other modules at robust speeds. > > > > We currently run the TPP locally on a Windows PC with 64 cores (AMD > Threadripper 3990X) and 128gb of RAM running on an SSD. Would a stronger > CPU or more RAM (or even a GPU) give us any significant boost in search > speeds if we're running a Docker image of the TPP on a Linux server? > > > > Thank you! > > -Will > > -- > You received this message because you are subscribed to the Google Groups > "spctools-discuss" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to spctools-discu...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/spctools-discuss/71df6d40-cadf-4009-8f91-78a41dae7e6bn%40googlegroups.com > > <https://groups.google.com/d/msgid/spctools-discuss/71df6d40-cadf-4009-8f91-78a41dae7e6bn%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- You received this message because you are subscribed to the Google Groups "spctools-discuss" group. To unsubscribe from this group and stop receiving emails from it, send an email to spctools-discuss+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/spctools-discuss/94852e3e-a0bd-4e83-8bd4-54009c2600e9n%40googlegroups.com.
