Re: [ccp4bb] buying a cluster

2018-12-04 Thread James Holton
Graeme's suggestion of a standard benchmarking dataset is a good one, but I'm not so sure a 24 GB download size is going to get a lot of hits. In fact, in some countries you have to pay for internet by the GB, and it costs as much as mobile phone data! The large size also makes it hard to separ

Re: [ccp4bb] buying a cluster

2018-12-02 Thread graeme.win...@diamond.ac.uk
Re: publishing benchmarks - great idea - expand on what James described earlier. Most programs are GHz dependent (for most “sensible” definitions of GHz (not the mega-hyper-pipeline stall prone P4 say) however I see your point that “threaded” and “optimised for vector systems (e.g. AVX512)” woul

Re: [ccp4bb] buying a cluster

2018-12-02 Thread Markus Heckmann
Hi Graeme, I suspect that this conclusions depends very closely on (i) the shape of > the problem and (ii) the extent to which the binary has been optimised for > the given platform. > > I do hope some of these info are analyzed and either published or at least put at ccp4 wiki. I am pretty sure

Re: [ccp4bb] buying a cluster

2018-12-01 Thread graeme.win...@diamond.ac.uk
HI James, Re: dissenting opinion I suspect that this conclusions depends very closely on (i) the shape of the problem and (ii) the extent to which the binary has been optimised for the given platform. I am pretty sure that there are some applications (heavily threaded, making extensive use o

Re: [ccp4bb] buying a cluster

2018-11-30 Thread James Holton
I have a dissenting opinion about computers "moving on a bit".  At least when it comes to most crystallography software. Back in the late 20th century I defined some benchmarks for common crystallographic programs with the aim of deciding which hardware to buy.  By about 2003 the champion of m

Re: [ccp4bb] buying a cluster

2018-11-26 Thread V F
Dear all, Thanks for all the off/list replies. > To be honest, how much are they paying you to take it? Can you sell it for > scrap? May be I will give it a pass. > To compare, two dual CPU servers with Skylake Gold 6148 - that is 40 cores - > will probably beat the whole lot even if you could ke

Re: [ccp4bb] buying a cluster

2018-11-24 Thread Robert Esnouf
c.uk / rob...@well.ox.ac.uk / robert.esn...@bdi.ox.ac.uk Tel: (+44)-1865-287783 (WHG); (+44)-1865-743689 (BDI)   -Original Message- > From: "V F" > To: CCP4BB@JISCMAIL.AC.UK > Date: 23/11/18 13:11 > Subject: Re: [ccp4bb] buying a cluster > > Hello Peter &g

Re: [ccp4bb] buying a cluster

2018-11-23 Thread Ian Tickle
Dear Veronica The answer depends to a large extent on your own, or your group's, mode of working. Specifically, are you interested in the fastest turn-around of individual jobs, or are you interested the highest throughput of many jobs running simultaneously, i.e. draining the batch queues as qui

Re: [ccp4bb] buying a cluster

2018-11-23 Thread George Sheldrick
Powell<193323b1e616-dmarc-requ...@jiscmail.ac.uk> Sent: 23 November 2018 11:05:13 To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] buying a cluster Hi For X-ray diffraction image processing, both XDS and DIALS can use multi cpus. For cryo-EM, Relion can run on multi-cpus, but can also mak

Re: [ccp4bb] buying a cluster

2018-11-23 Thread V F
Hello Peter Thanks for quick replies. > (1) a single large machine with 32 cpus, or: No > > (2) a system with 32 separate computers linked together in a network? > Yes Each node has HP ProLainat that contains 2x Intel Xeon Processor X5550 (quad-core, 2.66 GHz. ###

Re: [ccp4bb] buying a cluster

2018-11-23 Thread Ville Uski
On Fri, Nov 23, 2018 at 11:30:01AM +0100, V F wrote: > Which programs benefit from multi-cpu cluster? in addition to what the others have mentioned, you can also configure ccp4i2 to run jobs remotely on the cluster. Currently it requires SGE, but support for other job schedulers (such as Slurm) ca

Re: [ccp4bb] buying a cluster

2018-11-23 Thread Peter Keller
Dear Veronica, Is this: (1) a single large machine with 32 cpus, or: (2) a system with 32 separate computers linked together in a network? If (1) others have already answered, although be aware that for many applications the throughput doesn't scale linearly with the number of cpus. This mea

Re: [ccp4bb] buying a cluster

2018-11-23 Thread Nicolas Soler
Hi Veronica, Fragment-based molecular replacement programs such as ARCIMBOLDO are based on PHASER /SHELXE jobs parallelization over multiple CPUs in a single machine or in a cluster. All the best, Nicolas On 11/23/2018 11:30 AM, V F wrote: Dear all, Which programs benefit from multi-cpu c

Re: [ccp4bb] buying a cluster

2018-11-23 Thread Thomas, Jens
CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] buying a cluster Hi For X-ray diffraction image processing, both XDS and DIALS can use multi cpus. For cryo-EM, Relion can run on multi-cpus, but can also make good use of GPUs, so don't forget this as an option if you're going over to the "dark sid

Re: [ccp4bb] buying a cluster

2018-11-23 Thread Harry Powell
Hi For X-ray diffraction image processing, both XDS and DIALS can use multi cpus. For cryo-EM, Relion can run on multi-cpus, but can also make good use of GPUs, so don't forget this as an option if you're going over to the "dark side"... On 23 Nov 2018, at 10:30, V F wrote: > Dear all, > Whic

Re: [ccp4bb] buying a cluster

2018-11-23 Thread luis.fuentes-mont...@diamond.ac.uk
The first two programs that come to my mind (I know there are a lot more programs) are xia2 and ccp4 cloud. Sheers, Luiso, From: CCP4 bulletin board on behalf of V F Sent: 23 November 2018 10:30:01 To: ccp4bb Subject: [ccp4bb] buying a cluster Dear all

[ccp4bb] buying a cluster

2018-11-23 Thread V F
Dear all, Which programs benefit from multi-cpu cluster? Since the physics department is getting rid of a old 32 compute node cluster, I was hoping to find some benefit using for crystallographic work. Looking a ccp4wiki or google-fu did not help Many thanks Veronica ##