Lambda

MLPerf, the AI industry's standard speed test, adds its first test of AI agents as a record 30 organisations take part

22 September 2026 · 5 min

MLCommons published MLPerf Inference v6.1 with a record 30 submitting organisations and, for the first time, a test of AI agents doing multi-step work such as coding.

Lambda borrows just over $1 billion to build more GPU computers

1 October 2026 · L3 Iron and cloud · $1.008bn; 6.78% fixed; to 2033 · Confirmed

Lambda, a cloud company that rents out Nvidia computer chips for AI work, has closed a $1.008 billion loan to pay for more of the computers it buys and leases out. The loan carries a fixed interest rate of 6.78 percent and is due in 2033, and is backed by the GPU servers themselves plus payments Lambda already has coming from two large customers it hasn't named. Borrowing against already-signed customer payments, rather than raising money by selling shares, is how several AI-cloud companies are funding their buildout right now; it means the computers get built, but it also means Lambda now owes over a billion dollars whatever happens to AI demand later.

Source

Lambda

Lambda plans an Oklahoma data centre it expects to pay about $500 million in taxes over ten years

27 September 2026 · L1 Power and land · est. $500m in taxes over 10 years; up to 1,000 construction and 100 permanent jobs · Confirmed

Lambda, a company that rents out AI computing, announced on 27 September that it plans to build a data centre at MidAmerica Industrial Park in Mayes County, Oklahoma. It says it is expected to pay an estimated $500 million in taxes over the next ten years at current rates, and that the project is expected to create up to 1,000 construction jobs and up to 100 permanent ones. Citing Oklahoma's Data Center Customer Ratepayer Protection Act, Lambda says it will pay 100 percent of the site's energy costs, and it plans closed-loop cooling to limit water use. The announcement gives no figure for how much electricity the site will use.

Source

Lambda

The industry-standard AI speed test adds its first test of AI agents, and a record 30 organisations take part

16 September 2026 · L3 Iron and cloud · 30 orgs (record); DeepSeek R1 best per chip 5.7x in a year · Confirmed

MLCommons published results for MLPerf Inference v6.1 on 16 September, its industry-standard, peer-reviewed benchmark of how fast AI systems run, with a record 30 organisations submitting, including Nvidia, Microsoft Azure, Oracle, CoreWeave, Nebius and Lambda. For the first time it includes a test of AI agents: the Edge Agentic Inference test, which measures multi-turn work such as agentic coding, where each step depends on the ones before. MLCommons says the best per-chip result on its visual-language test improved 2.99 times in six months, and on the DeepSeek R1 reasoning test 5.7 times in a year. For people using AI tools, MLCommons says gains like these could let the same systems serve more users.

Source

Read the story: MLPerf, the AI industry's standard speed test, adds its first test of AI agents as a record 30 organisations take part · Nvidia · Lambda · Oracle · Microsoft Azure · CoreWeave · Nebius