NVIDIA has officially addressed rumors regarding a potential delay in the release of its new AI chip series, Blackwell. The company acknowledged that demand for its Hopper chips remains exceptionally high and confirmed that Blackwell samples have already begun extensive testing. NVIDIA also announced plans to ramp up production in the second half of the year but refrained from making further comments on the rumors. So, will chip deliveries be delayed?
NVIDIA Responds to AI Chip Crisis: Blackwell Samples Undergoing Extensive Testing, Production to Increase
According to reports from The Information, design flaws in the Blackwell AI chips could potentially delay their release by at least three months. This situation has raised concerns among major customers like Meta, Google, and Microsoft. The report suggests that NVIDIA has informed these key clients about potential delays, which could pose significant issues for these companies, given the critical importance of AI projects.
In response to these concerns, NVIDIA placed an additional order for 4nm chips from TSMC in July, aiming to increase GPU chip production for the Blackwell platform by 25%. This move is expected to accelerate the production of Blackwell chips, which have been described as “the world’s most powerful AI chip” and are intended for use in AI servers.
However, large-scale production of Blackwell chips is now expected to be delayed until early 2025. This delay could challenge NVIDIA’s efforts to maintain its leadership in the AI chip market. Nevertheless, the company is working intensively to resolve production issues and meet customer demands.
NVIDIA is collaborating closely with TSMC to overcome the challenges in the design and manufacturing processes of the AI chips. The Blackwell chips are anticipated to bring significant innovation to AI and machine learning applications, with the company claiming that these chips will redefine industry standards through their high performance and energy efficiency.
The Blackwell platform represents the latest technological advancement for NVIDIA’s AI servers and data centers. However, production delays and design flaws may complicate the achievement of these goals. What do you think? Could this delay cause major disruptions in AI projects? Share your thoughts in the comments below.