Stories tagged: inference

Stories tagged: inference.

A tinted 1944 photo of a control room full of unlabeled dials, beside a screenshot of Fireworks' X post, under the headline "1,500 What?" and a yellow bar reading "The release never names the unit"

Fireworks Ultrafast: 1,500 What? Europe Capacity Waits for 2027

Fireworks Ultrafast claims "speeds exceeding 1,500" with no unit in the release; new European Nvidia capacity starts in 2027. What ships now, what waits.

2026-11-04 · 5 min read

A blue-tinted NVIDIA DGX B200 system flanked by two server racks, under the headline Speed is not throughput

Models

LlamaIndex OCRia-1 Inference: What the 10x Speedup Measures

LlamaIndex says OCRia-1 runs nearly 10x faster on one B200. The technical report shows that's single-stream speed; pages per second rise about 4.7x.

2026-11-03 · 5 min read

Products

NVIDIA AICR for Dynamo: What 'Validated' Means on Kubernetes

NVIDIA's AICR recipe pins 17 Helm charts for Dynamo 1.4.2 on Kubernetes. What "validated" covers, what it leaves to you, and one alpha release in the stack.

2026-11-03 · 7 min read