My data is too big to fit in GPU RAM. Can I use MapD?


#1

For info on how MapD uses both host memory and GPU memory, see:

If your hot data (i.e. active columns and partitions in compressed format) can’t fit in the GPU memory of one node of MapD, you can go distributed on the enterprise edition:
https://www.mapd.com/docs/latest/mapd-core-guide/distributed/

Also, although not its primary use case, MapD can run in host memory and CPU-only. See “–cpu” flag here:
https://www.mapd.com/docs/latest/mapd-core-guide/configuration/?highlight=cpu