Consider GPU Acceleration / GSST

[alamb]

I am adding these here to capture the usecase / requirement, not that I plan to make any of these changes myself

@lwwmanning says:

I think it's worth experimenting with FSST12, and also integrating in the extra data required for GSST (https://www.vldb.org/2025/Workshops/VLDB-Workshops-2025/ADMS/ADMS25-01.pdf)

The paper is "High Throughput GPU-Accelerated FSST String Compression"

TLDR is that paper seems to suggest there are several techniques / additional details required to use FSST on GPUs