According to analysts' predictions, after the delay in the release of the B200 chip, the price may be set between $0.03 million and $0.04 million. The quarterly shipments as of January 2025 are expected to reach about 32,500 units, contributing $3 billion in revenue, accounting for 9% of the total expected revenue for the quarter.
What is the revenue impact of NVIDIA's delayed shipment of B200?
According to the technology media, The Information, NVIDIA has informed its customers that the new Blackwell B200 chip will be delayed for three months or longer before being shipped in bulk, and it may be delayed until Q1 of next year.
Blackwell chips were originally scheduled to start mass production in October 2024, and a delay that pushes it back to April 2025 will directly impact NVIDIA's quarterly revenue. In response to this, NVIDIA informed the media that the strong demand for the Hopper chip and the production plan for the Blackwell chip have not changed.
A spokesperson for NVIDIA said,
"As we said before, the demand for Hopper is very strong. Large-scale Blackwell sampling work has been completed, and production in the second half of the year will proceed as planned. In addition, we do not comment on rumors."
This means that even if the B200 is delayed to market, the impact on NVIDIA's revenue may not be particularly significant because the market demand for the Hopper chip is still strong.
Currently, three types of Blackwell GPU architecture chips are known: B100, B200, and GB200 super chips. If the B200 is delayed, according to Mizuho analyst Vijay Rakesh, its price may be set between USD 0.03 million and USD 0.04 million, while the GB200 may be priced between USD 0.05 million and USD 0.07 million.
UBS Group analyst Timothy Arcuri predicts that by the quarter ending January 2025, shipments of the B200 and GB200 will reach approximately 32,500 and 43,400, respectively, accounting for only 7% of all chip shipments in that quarter, most of which will still be Hopper chips.
Combining the estimates of the two analysts, B200 product revenue contribution for the quarter ending January next year will reach USD 3 billion, accounting for 9% of the total expected revenue for that quarter (USD 34.5 billion).
For NVIDIA's customers, the chip delay may exacerbate their concerns about the timing of deliveries. According to reports, second-quarter capital expenditures of technology giants including Alphabet, Amazon, Microsoft, and Meta totaled nearly USD 60 billion, a year-on-year increase of two-thirds, and a significant portion of these expenditures flowed to NVIDIA.
The Information also reported that Meta has now placed an order worth at least USD 10 billion, and Microsoft has increased its order volume by 20% in recent weeks and plans to prepare 55,000-65,000 GB200 chips for OpenAI by Q1 2025.
In addition, design flaws will also affect production and delivery of NVIDIA NVLink server racks, as companies engaged in server work must wait for new chip samples before finalizing server rack designs.
However, some observers pointed out that NVIDIA's small fluctuations in sales and revenue are unlikely to change its popularity on Wall Street. As of Friday, NVIDIA's stock price closed at USD 107, with a price-to-earnings ratio of 40 based on this year's expected earnings and a projected P/E ratio of 30 for next year, which is still high.