openai ropes in samsung, sk hynix to source memory chips for stargate

OpenAI Expands AI Infrastructure with Korean Tech Giants

OpenAI is aggressively pursuing the expansion of its computational resources to support its artificial intelligence initiatives.

The creator of ChatGPT announced on Wednesday that it has reached agreements with leading memory chip manufacturers, Samsung Electronics and SK Hynix.

Strategic Partnerships for Stargate

These agreements focus on the production of DRAM wafers specifically for the Stargate AI infrastructure project, alongside the construction of new data centers within South Korea.

The letters of intent were formalized following a meeting held in Seoul.

Key attendees included OpenAI CEO Sam Altman, South Korea’s President Lee Jae-myung, Samsung Electronics’ Executive Chairman Jay Y. Lee, and SK Chairman Chey Tae-won.

Increased Memory Chip Production

Samsung and SK Hynix are planning to significantly increase their manufacturing output.

The goal is to produce up to 900,000 high-bandwidth DRAM memory chips each month.

These chips will be utilized in both the Stargate project and supporting AI data centers.

SK Group highlighted that this production level represents more than double the current industry-wide capacity for high-bandwidth memory chips.

The Stargate Project: A Massive Investment

Stargate is a large-scale infrastructure project spearheaded by OpenAI, Oracle, and SoftBank.

The project aims to invest $500 billion in the development of data centers dedicated to AI research and deployment within the United States.

Recent Investments in AI Compute Capacity

Wednesday’s announcements follow a period of substantial investment in AI compute capabilities.

OpenAI has been at the forefront of this activity.

Just weeks prior, Nvidia committed to investing up to $100 billion in OpenAI.

This investment will grant OpenAI access to over 10 gigawatts of compute capacity through Nvidia’s AI training systems.

Subsequently, OpenAI announced plans to construct five data centers in collaboration with SoftBank and Oracle for the Stargate project.

This expansion aims to elevate OpenAI’s total compute capacity to 7 gigawatts.

Oracle's Commitment

Earlier in September, Oracle agreed to provide OpenAI with $300 billion worth of compute capacity over a five-year period.

Expanding into Korea

OpenAI is also collaborating with the Korean Ministry of Science and ICT to identify potential locations for AI data centers outside of Seoul.

A separate agreement has been established with SK Telecom to facilitate the construction of an AI data center.

Furthermore, OpenAI has signed additional agreements with various Samsung subsidiaries to explore opportunities for building further data centers within the country.

Integration of OpenAI Technologies

As part of these agreements, both Samsung and SK Group will integrate ChatGPT Enterprise and OpenAI APIs into their respective operations.