Nvidia shrinks LLM memory 20x without changing model weights

Nvidia's KV Cache Transform Coding (KVTC) compresses LLM key-value cache by 20x without model changes, cutting GPU memory costs and time-to-first-token by up to 8x for multi-turn AI applications.

Nvidia BlueField-4 STX adds a context memory layer to storage to close the agentic AI throughput gap

Nvidia's BlueField-4 STX reference architecture inserts a dedicated context memory layer between GPUs and traditional storage, claiming 5x token throughput and 4x energy efficiency for agentic AI ...

New memory architecture targets AI inference bottlenecks

Lightbits Labs Ltd. today is introducing a new architecture aimed at addressing one of the most stubborn bottlenecks in large-scale artificial intelligence inference: the growing mismatch between the ...
TweakTown

Meta's huge 16,384 NVIDIA H100 AI GPU cluster: HBM3 memory crashed half of Llama 3 training

Meta released a new study detailing its Llama 3 405B model training, which took 54 days with the 16,384 NVIDIA H100 AI GPU cluster. During that time, 419 unexpected component failures occurred, with ...
TweakTown

NVIDIA's next-gen China-specific H30 AI GPU rumored with GDDR memory, instead of HBM

TL;DR: NVIDIA is reportedly developing a China-specific H30 AI GPU using GDDR memory instead of restricted HBM due to US export controls. This shift may limit performance compared to previous models, ...
CRN

Nvidia’s H200 GPU To One-Up H100 With 141GB Of HBM3e As Memory Race Heats Up

The H200 features 141GB of HBM3e and a 4.8 TB/s memory bandwidth, a substantial step up from Nvidia’s flagship H100 data center GPU. ‘The integration of faster and more extensive memory will ...