Terality and Pandas
Last updated
Last updated
Because Pandas doesn't scale well, we have built Terality to enable data scientists and engineers to scale Pandas just by changing the import line in their notebook or IDE.
Terality enables to run Pandas code 10x faster, even on 100 of GBs of data, in a full serverless way. No infrastructure setup is needed.
Pandas (locally) | Terality | |
Syntax | Pandas | Pandas |
Ideal dataset's size | Less than 1GB | 1GB to 100GBs |
Parallelization on multiple cores for lightning speed | No | Yes (10-100x faster than Pandas) |
Memory available | Local Memory | Virtually unlimited, fully managed |
Read our Benchmark Blog article https://www.terality.com/post/benchmark-terality-vs-pandas-10x
