Full archive

Benchmark: Infrastructure AI News

Search reporting on AI markets, models, products, companies, policy, and culture.

Showing 1-6 of 6Results matching "B300"

Today

2 stories

1 hour agoInfrastructure

Benchmark Infrastructure

InferenceX Ranks B300 First for Qwen3.5 at 15,815 Tokens per Second

The live ranking gives operators a matched speed-and-cost comparison for a long single-turn chat request, not a verdict on every production inference pattern.

2 min

18 hours agoInfrastructure

Benchmark Infrastructure

InferenceX Ranks GB200 First for MiniMax M3, While MI355X Has Lower Token Cost

The live AgentX result gives operators a comparable speed-and-price snapshot at one responsiveness target, but the ranking can shift with serving software and does not settle the wider configuration trade-offs.

3 min

Earlier this week

1 story

Sep 2, 2026Infrastructure

Benchmark Infrastructure

InferenceX Puts GB300 First for MiniMax M2.7, but Only on a Fixed Chat Test

The new ranking gives infrastructure teams a comparable speed and cost snapshot, but its single-turn design does not test the multi-step agent work MiniMax highlights for M2.7.

2 min

Last week

2 stories

Aug 25, 2026Infrastructure

Benchmark Infrastructure

OpenAI’s Jalapeño Claims Up to 1.9x Efficiency Gain, but Leaves Nvidia Training Intact

The Broadcom co-developed inference processor is headed for deployment by year-end. Its published gains are meaningful, but they exclude model training and Nvidia’s Vera Rubin platform.

3 min

Aug 24, 2026Infrastructure

Benchmark Infrastructure

NVIDIA Says Vera Rubin Delivers 30x More Agentic Throughput per Megawatt, Pending Review

The preliminary comparison shifts attention from fixed prompt benchmarks toward the power required to keep long, tool-using AI sessions responsive.

3 min

August 2026

1 story

Aug 18, 2026Infrastructure

Benchmark Infrastructure

AgentX Replays Claude Code Sessions to Test AI Serving Systems

The replay dataset is designed to measure infrastructure behind agentic coding workloads, not whether a model writes better code. Its value depends on whether synthetic traces preserve the traffic patterns operators need to serve.

3 min