Terry Kong

Trillion-parameter agentic RL refits are now 12-40x faster, shrinking an 87.5-minute full checkpoint transfer to 150 seconds by sending only bit-exact weight deltas.

Achieving six times the inference throughput of current LLMs while maintaining accuracy, Nemotron 3 Ultra redefines performance benchmarks for agentic reasoning tasks.

Speculative decoding, typically used post-RL, can be integrated directly into RL training loops to accelerate LLM rollout generation by up to 2.5x.

Multimodal models can now achieve state-of-the-art performance in real-world tasks like document understanding and audio-video comprehension with significantly reduced inference latency thanks to novel token-reduction techniques.