Nvidia has officially approved Samsung’s advanced 12-layer HBM3E memory chips. This long-awaited decision was confirmed in a report from News1. It paves the way for Samsung to supply its high-bandwidth memory for Nvidia’s flagship AI accelerators.
This approval marks a significant breakthrough for Samsung after a challenging 18-month qualification process. The company can now compete directly with SK Hynix and Micron in the lucrative AI chip market.
A Breakthrough Deal Forged at the Highest Level
According to the report, Nvidia CEO Jensen Huang sent a formal letter of intent to Samsung Electronics Chairman Jay Y. Lee. This communication confirmed the successful approval of the 12-layer HBM3E chips. The two companies are now finalizing purchase quantities, pricing, and delivery schedules.
Samsung’s main rivals have been supplying HBM3E to Nvidia for over a year. The South Korean tech giant faced repeated setbacks as its earlier chip designs reportedly struggled with heat-related issues. This necessitated a complete redesign of the memory stacks in the first half of this year to meet Nvidia’s stringent quality standards.
Shifting the Competitive Landscape for AI Memory
This approval is a strategic victory for Samsung, even with the industry’s impending move to next-generation HBM4. Gaining entry into Nvidia’s supply chain establishes a critical foothold. It positions Samsung to potentially supply HBM4 chips next year, securing its future in the high-stakes AI market.
Chairman Jay Y. Lee’s direct involvement was reportedly instrumental in finalizing the deal. His extended business trip to the United States in August included high-level meetings with Nvidia’s CEO and other partners. This top-down effort underscores the strategic importance of this partnership for Samsung’s semiconductor division.
The approval of Samsung’s HBM3E chips by Nvidia injects new competition into the AI memory market. This deal is expected to have a multi-billion dollar impact and strengthens the global AI supply chain for the foreseeable future.
Info at your fingertips
What are HBM3E chips used for?
HBM3E chips are a type of high-bandwidth memory. They are essential for powering advanced AI accelerators and data centers, enabling faster processing of complex AI models.
Why did it take so long for Nvidia to approve Samsung’s chips?
Samsung’s initial HBM3E designs faced technical hurdles, specifically related to heat management. The company had to completely redesign its chips to pass Nvidia’s rigorous quality and performance tests.
Who were Nvidia’s previous HBM3E suppliers?
Before approving Samsung, Nvidia sourced its HBM3E chips exclusively from SK Hynix and Micron. These two companies have been supplying the chips for over a year.
What does this mean for the AI industry?
Samsung’s entry into Nvidia’s supply chain creates a more competitive market. It diversifies sources for critical AI components, which can lead to better supply stability and potentially influence future pricing.
What is the next step after HBM3E?
The semiconductor industry is already developing the next standard, HBM4. Samsung’s current approval with Nvidia positions it favorably to become a supplier for these future memory chips.
Trusted Sources
News1, Reuters
Get the latest News first — Follow us on Google News, Twitter, Facebook, Telegram , subscribe to our YouTube channel and Read Breaking News. For any inquiries, contact: [email protected]