What hardware do you recommend for COMSOL Multiphysics?

Solution Number: 866
Title: What hardware do you recommend for COMSOL Multiphysics?
Platform: All Platforms
Applies to: All Products
Versions: All versions
Categories: Installation

Problem Description

I am going to buy a new dedicated computer for running COMSOL Multiphysics. What hardware do you recommend? What about CPU speed and amount of RAM?


With regards to good COMSOL performance, the overall most important parameter for a modern computational machine is that you have enough physical RAM memory. COMSOL solvers use all available processor cores in parallel by default in shared-memory processing. See details in knowledgebase solution 1096. Therefore, also processor parameters that affect the parallelization efficiency, like memory bandwidth and number of memory channels, are important for good COMSOL performance. Those numbers are more important than the CPU clock rate nowadays.

In this solution we will first discuss which architecture parameters to look at. Then we will translate that into specifications you typically will find in vendor price lists and quotes and give some general recipes how to select hardware.

Parameters that Affect Performance

  • A 64-bit operating system is highly recommended.
  • For a given hardware, the choice between Windows, MacOS, or Linux will not significantly affect performance.
  • At least 4-8 GB physical RAM per core of the computer is recommended.
  • Dual-socket or if possible four-socket nodes are recommended.
  • A CPU with as fast a memory bus as possible is beneficial (QPI or HyperTransport bus). The numbers are often measured in GigaTransfers/Seconds (GT/s) and can be found on manufacturers' webpages or Wikipedia (see example below).
  • As high memory bandwidth as possible is beneficial.
  • If you can find information about number of memory channels, it should be as high as possible. This information can be more difficult to find.

Selecting Hardware

Since the hardware and hardware driver market is a moving target, it is a challenge to specify a COMSOL-optimal CPU type and amount of RAM. This will rapidly change over time. Also, there is a trade-off between amount of physical RAM memory and cost. Past a certain amount of RAM, cost starts to grow exponentially. For different budgets, you will find different optimal choices.

A rule of thumb is to pick the highest CPU model number in a series. Below are specifications taken from couple of vendor quotes for two different Intel-based platforms.

  1. "Intel Xeon E5620 2.4 GHz, 24 GB RAM, 2 Processors, 1066 MHz Max mem"
  2. "Intel Xeon X5650 2.66 GHz, 48 GB RAM, 2 Processors, 1333 MHz Max mem"

Our intuition tells us that the higher model number (X5650) should be better, and more RAM must be better too. Now is a good time to go to a detailed spec chart to do some comparisons. This Wikipedia page gives some interesting information.

  1. Both quotes are 2-processor systems, which is the same thing as Dual-socket, so we are fine here on both quotes.
  2. The Wikipedia link above tells us that the E5620 in quote A has 4 cores. Dual sockets makes it 8 cores in total and thus 24GB/(4*2) = 3 GB/core. The X5650 in quote B has 6 cores, and the memory per core counts to 48GB/(6*2) = 4 GB/core. Quote A does not match the minimum recommendation of 4GB/core.
    QUOTE B wins
  3. The Max mem statement is the same as the Memory column on the Wikipedia page which is a measure of how fast the memory is (memory bandwidth). Quote B is 25% faster than quote A.
    QUOTE B wins.
  4. Quote B has 12 cores whereas A has 8 cores (see point 1. above). This together with point 3. above gives us reason to believe that the spec in quote B is at least 50% faster only based on these numbers. Note that the clock rate (2.4 and 2.66 GHz) is probably unimportant here. The memory speed is the bottleneck.
    QUOTE B wins.
  5. If you look in the chipset column on the Wikipedia page, you can see that quote A has QPI 5.86 GT/s whereas B has slightly more favorable, 6.4 GT/s. The column also gives us the interesting piece of information that if you go higher up in model series than X5650, you get no additional QPI or memory speed. So you probably don't get much additional performance for the extra money you would spend.
    QUOTE B wins.

Of course quote B will be more expensive. This type of comparison can be done with AMD processors as well.

In summary, below are three generic examples of computers specs in increasing order of price and performance

  1. Budget alternative, less recommended for high fidelity COMSOL modeling:
    Dual core processor, 4 GB RAM, 64-bit OS, 512 MB Graphics card. Roughly 1-1.5 kUSD.
  2. Mid-range alternative, good for many COMSOL applications:
    Dual quad-core processor, 32 GB RAM, 64-bit OS, 1GB Graphics card. Roughly 3-4 kUSD.
  3. High-end alternative, for large-scale COMSOL models:
    Dual 6-core processor, 96 GB of RAM. Roughly 10-12 kUSD.


We recommend modern AMD FirePro or NVIDIA based dedicated graphics cards. A list of tested graphics cards can be found on the system requirements page.


General-purpose computing on graphics processing units (GPGPU, also referred to as GPGP and less often GP²) is currently not supported by COMSOL.

Hard Drives

Hard drive speed does not affect COMSOL performance as long as the problem fits into main memory. If main RAM memory is insufficient, the hard drive swap will affect performance severly, regardless of which hard disk you have.

See Also

Selecting hardware for a compute cluster, solution 1116.
COMSOL and Multithreading, solution 1096.


Poor | Excellent
Document quality?


COMSOL makes every reasonable effort to verify the information you view on this page. Resources and documents are provided for your information only, and COMSOL makes no explicit or implied claims to their validity. COMSOL does not assume any legal liability for the accuracy of the data disclosed. Any trademarks referenced in this document are the property of their respective owners. Consult your product manuals for complete trademark details.