More

    Samsung Races to Close AI Chip Gap with Nvidia Amid Setbacks

    Samsung Electronics (SSNLF.US) is making a determined push to reclaim its position in the competitive AI chip market, aiming to close the gap with industry leader Nvidia (NVDA.US). Following a series of development challenges, Samsung is now advancing in the artificial intelligence sector, especially with the recent approval of its high-bandwidth memory (HBM) chip, HBM3, by Nvidia.

    This breakthrough comes after Samsung faced significant hurdles, including development missteps and falling behind SK Hynix, its primary competitor. Historically, Samsung, a dominant force in the memory chip market due to its scale and engineering prowess, has struggled to keep pace with SK Hynix’s rapid progress. In response to these challenges, Samsung replaced the head of its semiconductor division in May, a move highlighting the company’s commitment to revitalizing its HBM segment.

    Jim McGregor, an analyst at Tirias Research, noted the unusual nature of Samsung’s current position. “We’ve never seen Samsung in this position before,” he stated. “The industry and Nvidia need Samsung more than anyone, but they need Samsung to go all out.” Samsung has not disclosed specific details about its partnerships but confirmed that it is working closely with clients and that its testing processes are progressing positively.

    The HBM market is anticipated to expand from $4 billion last year to $71 billion by 2027, according to Morgan Stanley. With AI demand surging, Samsung’s swift progress in securing Nvidia’s approval is critical for capturing a share of this growing market. Analysts Shawn Kim and Duan Liu from Morgan Stanley highlighted Samsung as a top pick, projecting that its HBM market share could grow by at least 10% by 2025, potentially boosting its revenue by about $4 billion.

    Samsung’s upcoming second-quarter earnings report may shed light on its HBM strategy. However, the company’s ability to meet Nvidia’s requirements by November remains uncertain, with some sources suggesting that delays could push the timeline to 2025.

    Samsung’s struggles come at a time when the company is also dealing with the legal battles of Executive Chairman Jay Y. Lee and has not prioritized HBM development. The surge in AI demand, spurred by the release of ChatGPT by OpenAI in late 2022, has intensified competition in the AI chip market.

    SK Hynix, in contrast, is poised to benefit from the demand surge, having already positioned itself strongly in the HBM market. The company has achieved significant progress and is now accelerating its production of HBM3E chips. SK Hynix plans to mass-produce next-generation 12-layer HBM3E chips this quarter, with shipments expected to commence in the fourth quarter.

    Samsung, under the leadership of Jun Young-hyun, is also making strides. The company has developed its own 12-layer HBM3E technology and is seeking Nvidia’s approval for both its 12-layer and 8-layer HBM3E chips. Analysts suggest that while Samsung is catching up, it faces a crucial test in establishing itself as a strong secondary supplier to Nvidia in a market projected to be worth $71 billion by 2027.

    Despite the challenges, Samsung’s extensive resources and production capabilities position it well to address market shortages once it meets Nvidia’s approval criteria. As the AI chip market continues to evolve, Samsung’s ability to adapt and innovate will be crucial in determining its future success in this rapidly growing sector.

    Related topics:

    What Is Emotion Classification NLP?

    How to Training Nlp Models?

    What Is Supervised Learning

    Recent Articles

    TAGS

    Related Stories