How did Blackwell recover from its reported 2024 shipping delay?
Blackwell’s recovery was driven by a production correction rather than the delay simply disappearing. Nvidia first described the platform as in full production, but reports then said volume shipments of its most advanced parts had moved to early 2025. Nvidia subsequently disclosed a GPU-mask change to improve yields and targeted a fourth-quarter ramp; by November it again described Blackwell as in full production. The company then reported billions of dollars in first-quarter Blackwell sales, moved NVL72 systems into full-scale production, ramped Blackwell Ultra, and ultimately said Blackwell sales and cloud GPUs were sold out.
Trace the chain
The production reset
- 1B200 and GB200Announcement
Blackwell platform announced at GTC 2024 with B200 and GB200
Nvidia announced the Blackwell architecture in San Jose, listing the B200 Tensor Core GPU and the GB200 Grace Blackwell Superchip, which pairs two B200 GPUs with a Grace CPU. The release claims up to 25x lower cost and energy for LLM inference compared with H100.
Nvidia Investor Relations ·
Packed with 208 billion transistors, Blackwell-architecture GPUs are manufactured
Why it matters here. The new Blackwell products gave Nvidia a platform it next described as being in full production.
- 2Initial production claimFinancial
NVIDIA says Blackwell is in full production
NVIDIA said the Blackwell platform was in full production. It said the platform formed the foundation for trillion-parameter-scale generative AI.
NVIDIA ·
We are poised for our next wave of growth. The Blackwell platform is in full production and forms the foundation for trillion-parameter-scale generative AI.
Why it matters here. This production claim set the benchmark that the next shipment-delay report challenged.
- 4Mask changeResponse
Nvidia discloses a Blackwell mask change to improve yields on its Q2 FY2025 call
On the earnings call, CFO Colette Kress said Nvidia had changed the Blackwell GPU mask to improve production yields and that the production ramp would begin in the fourth quarter. Jensen Huang said Nvidia meant to ship product, not merely start production, in Q4. Nvidia also said Blackwell demand was well above supply.
The Register ·
We executed a change to the Blackwell GPU mask to improve production yields.
Why it matters here. The yield-improving mask change and planned fourth-quarter ramp set up Nvidia's subsequent full-production statement.
- 5Full production againFinancial
Nvidia says Blackwell is in full production; data center revenue hits $30.8 billion
In its third-quarter fiscal 2025 results, Nvidia described Blackwell as being in full production and reported record data center revenue of $30.8 billion, up 17% from the prior quarter.
Nvidia Newsroom ·
Third-quarter revenue was a record $30.8 billion, up 17% from the previous quarter
Why it matters here. Restored full-production status led into the first reported quarter of Blackwell sales.
- 6First Blackwell salesFinancial
First quarter of Blackwell sales: Nvidia cites billions in revenue, data center at $35.6 billion
Fourth-quarter fiscal 2025 results reported data center revenue of $35.6 billion. Nvidia said Blackwell AI supercomputers achieved billions of dollars in sales in their first quarter, without giving an exact figure in the release.
Nvidia Newsroom ·
Fourth-quarter revenue was a record $35.6 billion, up 16% from the previous quarter and up 93% from a year ago.
Why it matters here. The first reported Blackwell revenue established a commercial base before Nvidia introduced Blackwell Ultra.
Scale-out
- 7Blackwell UltraAnnouncement
Blackwell Ultra announced at GTC 2025 for the second half of 2025
Nvidia announced Blackwell Ultra and the GB300 NVL72 rack, claiming 1.5x the AI performance of the GB200 NVL72. Partner products were expected from the second half of 2025.
Nvidia Investor Relations ·
Blackwell Ultra-based products are expected to be available from partners starting from the second half of 2025.
Why it matters here. The GB300 NVL72 announcement set up the system-level production scale reported next.
- 8NVL72 at scaleShipment
NVIDIA says Blackwell NVL72 is in full-scale production
NVIDIA said its Blackwell NVL72 AI supercomputer was in full-scale production. It said production covered system makers and cloud service providers.
NVIDIA ·
Our breakthrough Blackwell NVL72 AI supercomputer — a ‘thinking machine’ designed for reasoning— is now in full-scale production across system makers and cloud service providers.
Why it matters here. Full-scale NVL72 production was followed by Nvidia's statement that Blackwell Ultra production was ramping at full speed.
- 9Ultra rampsFinancial
NVIDIA says Blackwell Ultra production is ramping
NVIDIA said production of Blackwell Ultra was ramping at full speed. It identified Blackwell as its AI platform.
NVIDIA ·
Blackwell is the AI platform the world has been waiting for, delivering an exceptional generational leap — production of Blackwell Ultra is ramping at full speed, and demand is extraordinary.
Why it matters here. The full-speed Ultra ramp preceded Nvidia's report of Blackwell sales and sold-out cloud GPUs.
- 10Cloud GPUs sold outFinancial
NVIDIA says Blackwell sales and cloud GPUs are sold out
NVIDIA reported Blackwell sales. It said cloud GPUs were sold out.
NVIDIA ·
Blackwell sales are off the charts, and cloud GPUs are sold out.
Why it matters here. The reported sellout is the chain's operating outcome: Blackwell progressed from a disputed shipment schedule to active sales and constrained cloud supply.
What pushes back
- 3Delay report
Blackwell shipments reported delayed to early 2025; Nvidia says production is on track
Citing The Information, The Register reported that Nvidia had told Microsoft of delays to the most advanced Blackwell parts, with volume shipments pushed to the first quarter of 2025 and TSMC's CoWoS packaging suspected as the cause. Nvidia's statement did not confirm a delay: it said sampling had started and production was on track to ramp in the second half.
The Register ·
As we've stated before, Hopper demand is very strong, broad Blackwell sampling has started, and production is on track to ramp in 2H. Beyond that, we don't comment on rumors.
What it pushes against. The reported shift of volume shipments to early 2025 pushes against a smooth recovery and made a manufacturing response necessary.
Timelines in this chain
- Nvidia Data Center GPUs: Hopper, Blackwell and Vera Rubin
- HBM Supply Race: SK hynix, Samsung and Micron from HBM3E to HBM4E
- How Nvidia Came to Own AI Compute: From CUDA to the H100 Shortage
- US Export Controls on Nvidia AI Chips to China: A100 to H20 to H200
- AMD Instinct: From MI300X to Helios and the Gigawatt Deals
A chain is written by the editor from entries already published on the timelines above, each of which was read against its source before publication. The chain changes when those entries do. Data and terms.