Comment on 1.1B Taxi Rides Using OmniSciDB and a MacBook ProparentComments−shaklee36yThe biggest loss for omnisci was the in-memory limits. The highest end GPUs have 32GB, while you can find CPU servers with multiple TB. As soon as you spill out of that you take a big performance hit.
Comments
The biggest loss for omnisci was the in-memory limits. The highest end GPUs have 32GB, while you can find CPU servers with multiple TB. As soon as you spill out of that you take a big performance hit.