Discussions

Expand all | Collapse all

Thank you for the open source version

  • 1.  Thank you for the open source version

    Posted 08-03-2019 15:47
      |   view attached
    Thank you OmniSci for making an open source version available. I prepared a Demo & a presentation to convince my upper management to invest in OmniSci. I've attached the presentation to this message. The Demo is available at:

    http://34.83.129.0:8080/

    (Looks better if viewed on the Web browser. I haven't spent a lot of time to make it Mobile friendly.)

    I know OmniSci Immerse allows me to create many fancy reports, graphs etc. I will continue to develop those. Also I plan to talk about 'Immerse'.

    If you think I've mis-interpreted something, please let me know.

    Thanks again for making OmniSci open source.

    - Ajay
    #Uncategorized
    #Core
    #General
    #Examples

    Attachment(s)

    pptx
    GPU Databases.pptx   493K 1 version


  • 2.  RE: Thank you for the open source version

    Posted 08-05-2019 02:01
    Edited by Candido Dessanti 08-05-2019 09:50
    Hi,

    thanks for trying out and writing such a nice presentation about GPUs database.

    I would correct a few things

    1) AMX EPY 32 Cores  (a mistyping) AMD EPYC ;)
    2) Kinetica has a BI TOOL called Reveal heavily based on Apache SuperSet (IMHO it's a little chunky and not so easy to use)
    3) If you want to do some DV you can try omnisci/mapd-charting it's an opensource JS library with most of Immerse's objects
    4) GPUs are generally faster, and it's because they have more cores than GPU, but being two different kinds of cores, better to not give the idea that 5000 GPU cores are 156x faster than 32 CPU cores ;)
    5) TPUs are intended for different workloads than analytics; they have very low precision FPs and very rigid.




  • 3.  RE: Thank you for the open source version

    Posted 08-05-2019 07:44
    Thanks for the feedback @Candido Dessanti. I will incorporate all these changes in the presentation. Will also try out mapd-charting. ​


  • 4.  RE: Thank you for the open source version

    Posted 08-05-2019 10:01
    I forgot to mention the name of Kinetica BI Tool...il called Reveal ;(


  • 5.  RE: Thank you for the open source version

    Posted 08-05-2019 11:04
    Hi @AJAY CHITRE

    You can refer to my blog https://www.omnisci.com/blog/analyze-oil-and-gas-big-data-with-omniSci-visual-apps in which I explain how to use mapd-charting to create charts for an oil & gas dataset.

    Regards,
    Veda​


  • 6.  RE: Thank you for the open source version

    Posted 08-05-2019 21:42
    Thanks @Veda Shankar for the link. I will try out the mapd-charting.​