Loading…
Loading…
Goldman Sachs has made a bold prediction following the launch of NVIDIA Corp’s NVDA latest artificial intelligence chips. The investment bank expects significant growth for the memory chips used in AI systems, which could potentially benefit three major memory chip manufacturers.
What Happened: Goldman Sachs is anticipating a substantial expansion in the total addressable market for high-bandwidth memory (HBM) chips, driven by the demand for Nvidia’s new Blackwell GPU chips. The bank predicts the market to grow tenfold to $23 billion by 2026, up from $2.3 billion in 2022, reported CNBC.
The three primary beneficiaries of this burgeoning HBM market, as per Goldman Sachs, are SK Hynix, Samsung Electronics Co SSNLF, and Micron Technology MU. The bank also expects these stocks to gain a boost from the market expansion.
Goldman Sachs analysts, led by Giuni Lee, pointed out that the increasing demand for AI servers and the higher memory chip density per GPU are the key factors driving their optimistic outlook. They also highlighted the challenges faced by the major suppliers, such as larger chip sizes and lower production yields for HBM compared to conventional DRAM memory chips, which are likely to keep the supply tight in the near future.
Why It Matters: The launch of Nvidia’s new AI chips comes at a time when the company’s stock has been the center of attention. The stock has been on a tear since the start of 2023, fueled by the explosion of interest in artificial intelligence sparked by OpenAI‘s ChatGPT. However, concerns have been raised about the sustainability of this growth, with some suggesting that it might be fueled by reckless speculation.
Despite these concerns, investor optimism has hit a two-year high, with the onset of the Great Rotation potentially underway. The future of technology, particularly the significance of computing power, has been a topic of interest, with some experts suggesting that “compute” will be the “currency of the future” instead of fiat or cryptocurrency.
Image Via Shuterstock
Engineered by Benzinga Neuro, Edited by Kaustubh Bagalkote
The GPT-4-based Benzinga Neuro content generation system exploits the extensive Benzinga Ecosystem, including native data, APIs, and more to create comprehensive and timely stories for you. Learn more.
Loading…
Loading…