Saturday, August 3rd 2024
Design Issues May Postpone Launch of NVIDIA's Advanced Blackwell AI Chips
NVIDIA may face delays in releasing its newest artificial intelligence chips due to design issues, according to anonymous sources involved in chip and server hardware production cited by The Information. The delay could extend to three months or more, potentially affecting major customers such as Meta, Google, and Microsoft. An unnamed Microsoft employee and another source claim that NVIDIA has already informed Microsoft about delays affecting the most advanced models in the Blackwell AI chip series. As a result, significant shipments are not expected until the first quarter of 2025.
When approached for comment, an NVIDIA spokesperson did not address communications with customers regarding the delay but stated that "production is on track to ramp" later this year. The Information reports that Microsoft, Google, Amazon Web Services, and Meta declined to comment on the matter, while Taiwan Semiconductor Manufacturing Company (TSMC) did not respond to inquiries.Update 1:
Update 2:
SemiAnalysis's Dylan Patel reports in a message on Twitter (now known as X) that Blackwell supply will be considerably lower than anticipated in Q4 2024 and H1 2025. This shortage stems from TSMC's transition from CoWoS-S to CoWoS-L technology, required for NVIDIA's advanced Blackwell chips. Currently, TSMC's AP3 packaging facility is dedicated to CoWoS-S production, while initial CoWoS-L capacity is being installed in the AP6 facility.Additionally, NVIDIA appears to be prioritizing production of GB200 NVL72 units over NVL36 units. The GB200 NVL36 configuration features 36 GPUs in a single rack with 18 individual GB200 compute nodes. In contrast, the NVL72 design incorporates 72 GPUs, either in a single rack with 18 double GB200 compute nodes or spread across two racks, each containing 18 single nodes.
Source:
Bloomberg
When approached for comment, an NVIDIA spokesperson did not address communications with customers regarding the delay but stated that "production is on track to ramp" later this year. The Information reports that Microsoft, Google, Amazon Web Services, and Meta declined to comment on the matter, while Taiwan Semiconductor Manufacturing Company (TSMC) did not respond to inquiries.Update 1:
The production issue was discovered by manufacturer TSMC, and involves the processor die that connects two Blackwell GPUs on a GB200." — via Data Center DynamicsNVIDIA needs to redesign its chip, requiring a new TSMC production test before mass production. Rumors say they're considering a single-GPU version to expedite delivery. The delay leaves TSMC production lines idle temporarily.
Update 2:
SemiAnalysis's Dylan Patel reports in a message on Twitter (now known as X) that Blackwell supply will be considerably lower than anticipated in Q4 2024 and H1 2025. This shortage stems from TSMC's transition from CoWoS-S to CoWoS-L technology, required for NVIDIA's advanced Blackwell chips. Currently, TSMC's AP3 packaging facility is dedicated to CoWoS-S production, while initial CoWoS-L capacity is being installed in the AP6 facility.Additionally, NVIDIA appears to be prioritizing production of GB200 NVL72 units over NVL36 units. The GB200 NVL36 configuration features 36 GPUs in a single rack with 18 individual GB200 compute nodes. In contrast, the NVL72 design incorporates 72 GPUs, either in a single rack with 18 double GB200 compute nodes or spread across two racks, each containing 18 single nodes.
105 Comments on Design Issues May Postpone Launch of NVIDIA's Advanced Blackwell AI Chips
Hardware prices in my country never go down, so there is no point for me to wait really
I'm excited to see how 9950X3D vs Arrow Lake will turn out this October
Another point he makes is the cost. When you buy an $1200+ card, you expect to max the settings in the game. You shouldn't expect it to give you 35fps in return. The 4080 is too slow for the money, plus some people save for a long time to afford such a card and will expect it to be in their system for 3-5 years. It's an investment, and then to be told you $1200 1 year old card is not only slow in 2025 games, but that amazing new monitor you've been waiting for won't work with you card because it's only DP1.4...
Not everyone uses their computer the same as you or me, you have to respect that.
Kinda like saying Usain Bolt is slow for only running 10m/s, he should be running at 20m/s because some dude who have no idea about human limits thought so ;)
We are obviously not like you, Mr Hardware expert, playing Pong on a 12" 320x180 CGA display at 10,000,000FPS with an overclocked 4090.
I'm playing all the lastest games at 4K144hz actually, gonna need alot more horsepower than 4090 that 5090 might just be barely enough
Just to note, a GPU to me is 500 GBP max. If I'm paying more, I'm expecting A LOT more!