Urgent.News

What's breaking now, across thousands of outlets.

AI

OpenAI data center chief Chris Malone is out, the latest in a string of executive exits

An OpenAI spokesperson said the company "recently reorganized our infrastructure organization to support the scale and pace of our work."

Nvidia's latest results sent Wall Street into a frenzy, demonstrating the enduring enthusiasm for AI and its leading chip designer. The company surpassed expectations for its second quarter and increased its third-quarter earnings forecast. Nvidia's data center revenue grew by 117% year-over-year to $89 billion, and the company projected third-quarter revenue of $108 billion.

Investors cheered the news, yet the stock initially dipped after-hours, only to recover after CEO Jensen Huang emphasized that AI had reached a pivotal moment. CEO Huang's remarks spurred a 4% jump in shares. Seth Hickle, chief investment officer at Mindset Wealth Management, praised the report as "amazing." Nvidia's shares have surged over 1,700% over the past four years, but this year, other chip companies are expected to benefit from the AI data center construction boom.

The company's exposure to potential circular deals, where AI goods producers fund other ecosystem players, has raised concerns about inflated demand and market volatility. Nvidia's maximum gross exposure under all land, power, and shell guarantees is $3.5 billion, a small fraction of its quarterly revenue. Analysts believe the positive outlook for the AI sector remains solid despite the stock's initial underperformance.

Written by urgent.news from CNA - Business's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Read the original at cnbc.com →

More in AI

The sm_120 shared-memory cliff: why FP8 KV cache crashes vLLM on workstation Blackwell

If you run vLLM with --kv-cache-dtype fp8 on a DeepSeek-family (MLA) model and your GPU is a GB10, an RTX PRO 6000, or any workstation or consumer Blackwell card, there is a decent chance the engine…

  • vLLM crashes on Blackwell cards due to fp8 KV cache exceeding shared memory limit
  • Recommended solution: compute tile's shared-memory requirement vs device's sharedmemoryperblockoptin

More from Tuesday 25 August →