Nvidia’s Blackwell Chips Face Manufacturing Hurdles, But Analysts Remain Bullish

Nvidia, the leading player in artificial intelligence chips, is pushing forward with its new Blackwell line of processors. These chips are significantly larger and more complex than their predecessors, boasting 2.6 times more transistors, and are designed to maintain Nvidia’s dominance in the AI market. However, this ambitious leap in technology has brought about significant manufacturing challenges, impacting Nvidia’s profit margins and leading to a $908 million provision.

The Wall Street Journal reports that a major flaw in the Blackwell chip can render it unusable, significantly affecting manufacturing yields. The primary challenge, as explained by G. Dan Hutcheson, vice chair of TechInsights, lies in ensuring the cohesive functionality of these complex chips. Low yields in individual components can quickly lead to overall failure.

Despite these challenges, analysts remain optimistic about Nvidia’s future. Wedbush analyst Matt Bryson reiterated an Outperform rating on Nvidia with a price target of $138, citing the company’s strong second-quarter results, driven by data center sales. Bryson also noted that Nvidia expects Blackwell shipments to ramp up in the fourth quarter, generating several billion dollars in revenue for the year.

Rosenblatt analyst Hans Mosesmann also maintained a Buy rating with a price target of $200. He highlighted Nvidia’s robust performance in generative AI and networking, attributing it to the Hopper chip line. Although acknowledging the impact of Blackwell’s start-up ramp, Mosesmann emphasized that the revenue ramp for the January quarter remains unchanged.

Both analysts underscored the importance of Nvidia’s GPU roadmap, driven by accelerated computing, which offers significant return on investment compared to general-purpose computing, and the burgeoning generative AI market. Despite the challenges posed by Blackwell, both analysts remain confident in Nvidia’s long-term prospects, fueled by the strong demand for its chips in the data center and the rapidly expanding generative AI space.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top