Glossary3 sections1 min read
What is TFLOPS?
TFLOPS means trillions of floating-point operations per second. For AI GPUs, the headline figure is Tensor Core throughput in a given precision, quoted dense or with structured sparsity.
Why it matters when renting a GPU
Always compare TFLOPS in the same precision and the same convention. EverXCloud GPU pages show dense figures (without sparsity).
TFLOPS predict compute-bound work such as training at large batch sizes; memory-bound inference depends more on bandwidth.
Worked example
The H100 SXM reaches 989 dense BF16 TFLOPS and the A100 SXM 80GB 312: about 3.2 times more peak compute, while its bandwidth is only 1.6 times higher.
Frequently asked questions
- What does sparsity mean in TFLOPS figures?
- Some figures assume 2:4 structured sparsity, which doubles the headline number. EverXCloud shows dense values so GPUs compare fairly.