News & Updates

Paiton's new Qwen3.8 3-bit weights approach 16-bit task scores on two R9700 cards. Native MTP prefix caching resumes a 200K prompt in 0.68 seconds. Results, limits and setup.

NVIDIA DSX POD brings Vera Rubin AI factories to modular data centers. ElioVP explains site readiness, cooling and the thinking behind ModFlex.

Paiton runs Qwen3.8 Flash Next in 4-bit on two R9700 cards: 96.8% GSM8K, 96.3% HumanEval, 189.8 tokens/s and an optional RAM prefix cache. Accuracy, speed and setup.

Paiton runs Qwen3.8 Flash Next with 3-bit experts on two R9700 cards: 216.3 tokens/s, 565.1 aggregate tokens/s and a separate 200K mode. Benchmarks, quality and setup.

Qwen3.8 27B on one R9700: 19.9% faster weighted decode, a new 4-bit cache, 200K context and optional vision. Benchmarks and current Paiton setup.