Aven Cloud

Rent H200

141 GB HBM3e

Hopper with a massive 141 GB of HBM3e — fits larger models and longer context in memory, ideal for big-model inference and memory-bound training.

Specifications

Memory
141 GB HBM3e
Memory bandwidth
~4.8 TB/s
Form factor
SXM (NVLink)
FP16/BF16
~990 TFLOPS

Best for

Large-model inferenceLong-context LLMsMemory-bound training