DigestAI news desk

Cut through the AI noise.

Hardware & Compute3 min read

Amazon reportedly pursues $7B deal for Decart to accelerate AI on Trainium chips

Amazon is in talks to acquire Decart for around $7 billion, according to a report by Search: NVIDIA. Decart’s core value lies in its Optimization Stack, software that speeds up AI workloads on Nvidia GPUs, Alphabet TPUs, and Amazon’s Trainium chips without requiring model retraining. This could help Amazon shift AI workloads from Nvidia’s CUDA-dependent ecosystem to its own hardware, a…

1 source

Key points

  • Amazon reportedly offers $7B for Decart, a 3-year-old AI startup with ~100 employees and undisclosed revenue
  • Decart’s software optimizes AI workloads on Nvidia GPUs, Alphabet TPUs, and Amazon Trainium chips without model retraining
  • Anthropic reportedly passed on a $6B bid in September 2026 after due diligence

Earlier, Anthropic reportedly considered a $6 billion bid but abandoned it after due diligence. Decart’s valuation has surged from $3.1 billion in its last round, with Nvidia as an investor. The reported price—75% higher than the May round—reflects its potential to accelerate AI adoption on Trainium, though no terms or regulatory filings have been disclosed.

The story so far

2 episodes →
  1. Amazon reportedly pursues $7B deal for Decart to accelerate AI on Trainium chipsthis story
Full story from 247wallst.com · by Omor Ibne Ehsan · via Search: NVIDIAOpen source ↗

Amazon Wants to Pay $7 Billion for a 3-Year-Old AI Startup That Makes AI Run Faster on Nvidia’s Chips and Its Own

247wallst.com · 13 October 2026

Loading the full article…

This text was published by 247wallst.com and written by Omor Ibne Ehsan. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗

Topics · follow one to build your own front page

The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.

Comments

via GitHub Discussions

More in Hardware & Compute

All →

Related stories