🚨🚨 Feature Request

• A new implementation (Improvement, Extension)

General plan for contributors to track progress on benchmarks:

• README as specified by [FEATURE] Improve README for benchmarks #446 @haiyangdeperci
• variable number of workers in all benchmarks
• standardization of the benchmarks (e.g. merging Tiledb & zarr-related files into one for sequential access, parametrization across all the benchmarks) @DebadityaPal
• implementing write access benchmarks @DebadityaPal
• create a benchmark from a TileDB open source use case to verify the results @DebadityaPal
• direct result parsing into the README
• logging old results
• quantifying uncertainty of measurements (±) @DebadityaPal
• memory benchmarks ([FEATURE] Benchmarking memory #516) @KrishnaChaitanya1
• extending storage for all the benchmarks ([FEATURE] Extend benchmarks to all storage destinations #520) @madhucharan
• improve to_pytorch and to_tensorflow performance (The training model using to_pytorch and to_tensorflow is slow  #314)
• continuous delivery of benchmarks
• network utilization benchmarks @benchislett

If your feature will improve HUB

These implementations would ensure Hub's performance is well-measured and improved over time.