What computer system should I use with FLO-2D???

Home Forums Announcements What computer system should I use with FLO-2D???

This topic contains 3 replies, has 2 voices, and was last updated by  Karen 7 months ago.

Viewing 4 posts - 1 through 4 (of 4 total)
  • Author
    Posts
  • #602

    Karen
    Keymaster

    Updated – 4/20/2016

    FLO-2D will run on Windows 7 to 10 with a 64-bit OS and in any previous Windows operating system. We recommend you to purchase a 64-bit system with quad core or octa core processors. Processor speed is a very important consideration and we recommend an Intel processor with at least 3.2 Ghz. The Intel(R) Core(TM) processor i7 and i9 have also shown remarkably faster run-time speeds over less efficient processors as the AMD processors. Solid State drives (SSD) are also an important consideration because booting up a system and launching programs will take less time. Also, a SSD increase the performance of a computer system by how often the drive is accessed.

    For very large projects, gaming computers have proven to be effective at decreasing the total simulation time, they are typically built to maximize the performance so they have the best computing and graphical processors. For simulations with a number of elements that exceed the 250,000 grid elements and with large hydrographs (water volume), we strongly suggest you go with a gaming computer.

    I use a gaming computer with the following specs:
    • Operating system: Windows 10 Pro, 64-bit operating system
    • Processor: Intel(R) Core i7-4820KCPU quad-core @ 3.70 GHz
    • Memory RAM: 16.0 GB
    • Graphics: NVIDIA Graphic Card GeForce GTX 660, this is a pretty powerful graphic card.

    A simulation with around 2 million elements with all surface components turned ON runs in about 24 hours in my PC.

    A typical configuration that I recommend for demanding FLO-2D simulations will be:

    • Processor: Intel(R) Core(TM) i7 4 to 12 cores. 3.2Ghz minimum
    • Memory RAM: 16GB/32GB
    • Graphics: Nvidia GeForce GTX

    • This topic was modified 1 week, 2 days ago by  Karen.
    #1057

    Karen
    Keymaster

    I use a gaming computer with the following specs:

    • Operating system: Windows 10, 64-bit operating system, x64-based processor

    • Processor: Intel(R) Core i7-4820KCPU quad-core @ 3.70 GHz

    • Memory RAM: 16.0 GB

    • Graphics: NVIDIA Graphic Card GeForce GTX 660, this is a pretty powerful graphic card.

    A simulation with around 2 million elements with all surface components turned ON runs in about 24 hours in my PC.

    A typical configuration that I recommend for demanding FLO-2D simulations will be:

    • Processor: Intel(R) Core(TM) i7 quad-core

    • Memory RAM: 16GB/32GB

    • Graphics: Nvidia GeForce GTX

    #1387

    Anonymous

    Hi, continuing with the current publication, I wanted to ask, with the new processor technologies, what would be the best recommendation now?

    We found in the market AMD Ryzen Threadripper with 16 cores at 4.0 Ghz, or, Intel i9 with 10 cores at 4.3 Ghz. Without discounting the previous lines like the Ryzen 3, 5 and 7 of AMD or the i3, i5 and i7 of Intel. I am interested in testing models of more than one million grid elements with high hydrographic values, but additional, I need to save processing time, what hardware configuration would be the most advisable?

    Thank you very much for your attention.

    Cordially,
    L. David

    #1392

    Karen
    Keymaster

    Hi David,

    I haven’t tested anything higher than an i7 or a Xeon with 12 cores. I don’t have anything higher than 3.xx Ghz. But since the open mp code should break up the same regardless of the number of cores, I think having 16 cores vs 12 cores should still result in better processing times when you get over 500,000 cells. Having faster processor is one of the most sensitive parameters so getting over 4 Ghz should result in some excellent processing times.

    Also remember optimization. If you have something taking a long time to run, watch the optimization video. Sometimes editing a few channel nodes or fp nodes can make a lot of difference in the timestep control. Jim’s video is pretty helpful in that regard.

Viewing 4 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic.