sbert-benchmark-hardware-specs
IN premise — summaries/2026/08/24/reimers-2019-sentence-bert-s6-ablation-study.md
Created 2026-08-25T02:58:29+00:00
SBERT speed benchmarks were run on Intel i7-5820K @ 3.30GHz, Nvidia Tesla V100, CUDA 9.2, cuDNN; SBERT and InferSent use PyTorch, USE uses TensorFlow.
Summary
These speed comparisons were measured on a specific 2016-era workstation (i7-5820K, Tesla V100 GPU) using CUDA 9.2, so the numbers reflect that particular machine rather than modern hardware. An important caveat is that SBERT and InferSent run on PyTorch while USE runs on TensorFlow, meaning the speed gap partly reflects framework differences, not just the models themselves.