NVIDIA working 'as fast as it can' to get Samsung HBM3E memory certified for its new AI GPUs

NVIDIA is "working as fast as it can" to certify Samsung's new AI memory chips -- HBM3E -- with CEO Jensen Huang himself telling Bloomberg TV.

2

Jensen talked with the press after attending a ceremony at the Hong Kong University of Science and Technology, where the NVIDIA CEO was given an honorary doctorate in engineering. Jensen said that NVIDIA is looking at ways to purchase 8-layer and 12-layer HBM3E memory chips from Samsung, while as it stands, the company is acquiring most of its HBM chips used on its AI GPUs, from South Korean rival SK hynix.

Earlier this week during NVIDIA's Q3 2024 post-earnings call with analysts, the CEO didn't mention Samsung when talking about the company's major suppliers -- naming SK hynix and Micron -- but now, the company is "working as fast as it can" to get Samsung HBM3E certified.

• Read more: Samsung's new 8-layer HBM3E memory chips pass NVIDIA tests, deal soon
• Read more: Samsung to cull up to 30% of its global workforce: admin, sales and marketing
• Read more: Samsung showcases AI and auto tech: HBM3E, GDDR7, OLED, more
• Read more: Samsung denies claims its HBM3E memory failed NVIDIA's quality tests for AI GPUs

Samsung said that it was making "significant progress" on supplying HBM3E memory chips to major clients including NVIDIA, during its Q3 2024 earnings call on October 31. Samsung Electronics Executive Vice President Kim Jae-june said: "We have made significant process by completing critical steps in the quality testing with key clients, which should enable expanded sales in the fourth quarter" adding that Samsung plans to "release improved HBM3E chips based on the timeline of next-generation GPUs that [their] major clients are working on".

SK hynix has been dominating the HBM market for a while now, with fellow South Korean giant Samsung falling considerably behind. Samsung has had major changes in its leadership within its semiconductor business, its internal Samsung Foundry business, and more... so let's hope that NVIDIA speeding up the process gets Samsung into the HBM3E game earlier, rather than later.

NVIDIA stepping in and helping makes sense, as the company needs to feed itself with as many HBM memory chips as possible, as it's selling every single AI GPU it is getting fabbed from TSMC and made by its key manufacturing and production partners. Missing (more) of its required HBM memory chips is not good, so getting the third major HBM manufacturer onboard will help facilitate NVIDIA's continued AI GPU growth into 2025 and beyond.