0

I'm operating on a large amount of data with python pandas and it is depended on RAM only. My computer only has 8GB of RAM but I have a 1TB SSD.

Is there a way with windows 10 to increase my RAM with my hard disk? I read about virtual memory and a the paging file. This is already enabled but doesn't seem to have any effect on what the computer considers RAM.

Any suggestions are appreciated.

  • 2
    The virtual memory is does not increase the RAM as displayed by Windows. Instead, if the RAM is 'full' Windows will write some data from the RAM into the virtual memory in order to free RAM. Windows does this automatically (data from the RAM, which is least frequently used is moved into the virtual memory). Thus, you cannot force Windows to directly use the virtual memory. – daniel.neumann Jul 9 '16 at 9:16
  • So there is no way to increase RAM with hard disk? This seems like an essential OS problem then. – Nickpick Jul 9 '16 at 9:16
  • 1
    No. Hard drives aren't a replacement for more memory but maybe this answer would be helpful to you. – Vinayak Jul 9 '16 at 9:18
  • If you have a data set of 14 GB which you load into your RAM, than first the RAM will be used. When it is 'full', the data will be loaded into the virtual memory. Thus, you increase your working memory. But you do not increase your RAM. At least it is possible to load data sets larger than your RAM. – daniel.neumann Jul 9 '16 at 9:26
  • 1
    Even a SSD is still multiple orders of magnitude slower than RAM, especially when it comes to latency. It’s just not suitable to be used as “RAM”, even if it were possible. – Daniel B Jul 9 '16 at 9:36
1

This is not possible to do in current computer architectures.

Since you are using python pandas to work in large amounts of data, you could try to follow this thread: https://stackoverflow.com/questions/14262433/large-data-work-flows-using-pandas

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.