cuda – Jacar

View options

View Cards

Mode Paginated

Per page Per batch 48

Filters

What gpu-unlock unlocks on GeForce RTX 30 and 40 cards and how much AI speed you gain

gpu-unlock is a Linux kernel module that, before the NVIDIA driver loads, lifts the cap that halves tensor core throughput for FP16 and BF16 with FP32 accumulation on GeForce RTX 30 cards. On an RTX 3090 Ti, BF16 goes from 87.5 to 157.4 TFLOPS; in LLMs the prefill gains, not generation.

October 26, 2026 13 min 5

Artificial Intelligence

TensorRT-LLM: Extreme Acceleration on NVIDIA GPUs for LLMs

TensorRT-LLM is the NVIDIA inference engine that compiles each model into a binary optimized for the exact GPU and batch size it will serve. It uses hand-written CUDA kernels and native FP8 quantization on H100. Against vLLM it can run 2 to 3 times faster in the best case, at the cost of a 30 to 90 minute build.

November 4, 2024 7 min 278