Seqanswers Leaderboard Ad

Collapse

Announcement

Collapse
No announcement yet.
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • bioinformatics workstation

    We are looking to purchase a workstation capable of performing large bioinformatics projects in gene expression prognostic signatures, run Partek and Affy genomic suites and RNA seq.

    One suggestion I found on this group message board was this:
    CPU i7-3770k OR i7-4770K 3.50 GHz Processor
    RAM 32GB, ROM 1TB
    MB GA-X79-UD5
    HDD WD Red 2 -TB
    PSU SS-650RM
    CASE 300R
    Cooling H80
    Cooling UCTB12
    DVD 24B5ST
    OS W7P SP1 64bit OR W8
    Discussion of next-gen sequencing related bioinformatics: resources, algorithms, open source efforts, etc


    However, the HP sales people suggest a different option:
    HP Z820 Workstation (ENERGY STAR) (F1K12UT)
    Windows 7 Professional 64 (available through downgrade rights from Windows 8 Pro 64)
    Processor: Intel® Xeon® E5-2637 v2 (3.5 GHz, 15 MB cache, 4 cores)
    Standard memory: 16 GB 1866 MHz DDR3 ECC Unbuffered RAM (2 X 8 GB)
    Memory slots: 16 DIMM
    Internal drive bays: Four 3.5"
    External drive bays: Three 5.25" Internal drive: 1 TB 7200 rpm SATA
    Storage controller: Integrated SATA 6.0 Gb/s
    Optical drive: SATA SuperMulti DVD+/-RW
    Graphics: NVIDIA Quadro K4000 (3 GB)

    I have no experience in choosing a workstation in the 3K - 5K price range, I have only used laptops for my stats.

    Can anyone please suggest which might be the better option?

    Much obliged,
    Serban

  • #2
    The only obvious benefit of the HP machine is that it has more cache (the graphics card is nice, but you don't really need that). I would want the RAM amount in the first machine, you'll inevitably use it. Ditch windows, it'll be way to restrictive if you want to get serious.

    You might want to go ahead and just get more cores (e.g. an i7-4960X, which is sort of an updated version of what I have in my workstation). Always max the RAM, you'll need it, and whatever you do, don't forget to come up with a way to backup your data!

    Comment


    • #3
      Thank you Ryan. There is a huge difference in price between the two machines, the first one is about half the cost of the HP workstation. Any idea why, since they both do pretty much the same thing?

      Comment


      • #4
        Both the Xeon processor and the Quadro graphics cards will cost a bit more. The better CPU is fine, you'll use it, but you're unlikely to need much more than a basic graphics card (get a discrete one, just so it doesn't use a much CPU or RAM) given what you're doing.

        -Devon

        Comment


        • #5
          Also the HP has ECC ram which is a helluva lot more expensive per GB (but looks expandable up to 128Gb which may be useful down the track!)

          Comment


          • #6
            If this is an official purchase then HP (with added RAM) would be ideal choice.

            If you enjoy running after individual vendors when components fail (and some will over time) first choice would be fine.

            As Devon suggested downgrade the video card and up the RAM to as much as you can. If you need to run partek, affy suite then you will have to keep windows available on one of the disks and perhaps dual boot.

            Comment


            • #7
              Try to get as much ram as possible especially when you want to open many windows in Partek.

              Comment


              • #8
                I've got 64GB memory on my single-processor (E5-2620) "desktop" computer which I built just over a year ago, cost about $3,000 NZD with 3TB HDD + 120GB SSD. It works well for all the bioinformatics work I do, which in some cases has involved de-novo assembly of transcriptomes and small genomes. You'll be able to get a better-spec computer for cheaper now.

                If you're planning on doing any de-novo assembly, more than 32GB memory would be a very good idea -- the older assemblers can require over 100GB for human genome-sized assembly. That's about the only reason for high memory. Otherwise just go for thread count rather than processor speed -- you should get pretty good performance for bioinformatics work using most high-end desktop computers these days.

                Comment


                • #9
                  Thank you everyone! Much appreciated. This is an institutional rather than personal purchase so there is more flexibility with the funds.

                  It looks like the HP with whatever warranties apply would suit the institution better, as GenoMax suggested. I will downgrade the graphics card as Devon suggested, since there is no additional benefit with the more expensive version. Also, I will get estimates for 64 up to 128 GB of RAM and stay with the i7-4960X. GenoMax, what do you mean by dual boot?

                  Thank you!

                  Comment


                  • #10
                    Originally posted by Serban View Post
                    what do you mean by dual boot?
                    Use one (or more than one separate operating systems, but not at the same time) http://en.wikipedia.org/wiki/Multi-booting on one computer.

                    If you are willing to put up virtual machines then you could get parallels/VMware(both not free) /virtualbox (free) and run two OS's at the same time.

                    Comment


                    • #11
                      I've got both of these options on my desk at the moment, an i7 3930 on consumer gigabyte motherboard with 32gb RAM and a Dell workstation with dual E5 2680's and 256gb RAM, first machine cost about $3k (built myself) and the second cost $9k after institutional discounts.

                      I am the kind of guy that doesn't mind messing around inside computers, but have had a lot of trouble getting 64GB of RAM to work well in the homemade machine, and ended up dropping it back to 32G. I think part of the problem is that the RAM sets for consumer boards tend to come in max 32GB packs tested to play together nicely so it is a bit pot-luck whether you can find 8x8GB DIMMs that work well, even when underclocked a bit. For a while I couldn't figure out what was going on because I was using a version of memtest shipped with ubuntu that had a known error and spent a couple of weeks trying to figure out why all the analyses were segfaulting.

                      For single threaded tasks the "gaming machine" is faster, but for threaded analyses and anything that requires a lot of RAM the dell machine leaves the other in the dust.

                      My advice would be to go for the Xeon's with ECC RAM and a warranty, show HP the specs of the other machine, ask for a discount, or see if another company will give you an institutional discount, it probably is better to not have to worry about things going wrong!

                      Comment


                      • #12
                        Thank you James, it's definitely better to go with HP for the reasons you mentioned. It's over my head and not my task really to figure out which components are compatible. I will try to configure one of HP's workstations, probably HP Z820 to get at least 64 GB RAM, dual OS and upgrade to a i7-4960X (Turbo 4GHz) 6 core processor.

                        What do you think of this option?

                        Comment


                        • #13
                          I'd be inclined to go for a dual xeon machine in that price range, you will get ECC RAM, which seems to be generally more stable because it is designed for servers/enterprise, and it will be much easier to upgrade the RAM later, 16 DIMM slots would allow you to go up to 256GB in the future if you had 16GB sticks, or 128GB if you had 8GB sticks.

                          Comment


                          • #14
                            Thank you James!

                            Comment

                            Latest Articles

                            Collapse

                            • seqadmin
                              Current Approaches to Protein Sequencing
                              by seqadmin


                              Proteins are often described as the workhorses of the cell, and identifying their sequences is key to understanding their role in biological processes and disease. Currently, the most common technique used to determine protein sequences is mass spectrometry. While still a valuable tool, mass spectrometry faces several limitations and requires a highly experienced scientist familiar with the equipment to operate it. Additionally, other proteomic methods, like affinity assays, are constrained...
                              04-04-2024, 04:25 PM
                            • seqadmin
                              Strategies for Sequencing Challenging Samples
                              by seqadmin


                              Despite advancements in sequencing platforms and related sample preparation technologies, certain sample types continue to present significant challenges that can compromise sequencing results. Pedro Echave, Senior Manager of the Global Business Segment at Revvity, explained that the success of a sequencing experiment ultimately depends on the amount and integrity of the nucleic acid template (RNA or DNA) obtained from a sample. “The better the quality of the nucleic acid isolated...
                              03-22-2024, 06:39 AM

                            ad_right_rmr

                            Collapse

                            News

                            Collapse

                            Topics Statistics Last Post
                            Started by seqadmin, 04-11-2024, 12:08 PM
                            0 responses
                            18 views
                            0 likes
                            Last Post seqadmin  
                            Started by seqadmin, 04-10-2024, 10:19 PM
                            0 responses
                            22 views
                            0 likes
                            Last Post seqadmin  
                            Started by seqadmin, 04-10-2024, 09:21 AM
                            0 responses
                            16 views
                            0 likes
                            Last Post seqadmin  
                            Started by seqadmin, 04-04-2024, 09:00 AM
                            0 responses
                            47 views
                            0 likes
                            Last Post seqadmin  
                            Working...
                            X