Nvidia CEO Jensen Huang responded to investor concerns regarding Blackwell GPU delays, reassuring that their full production has resumed and shipments will commence in Q4 of this year. The statement was made at the Goldman Sachs Tech Conference, intending to calm worries about the production timeline. In fact, this is the second time in as many weeks that Huang has needed to ease concerns by claiming Blackwell is coming.
Addressing the Manufacturing Hiccups
Manufacturing issues that necessitated a mask change led to production delays, as disclosed during Nvidia's Q2 earnings call. Huang acknowledged these setbacks and their effects on stakeholders, asserting that Nvidia has ramped up production to satisfy strong demand.
Initially presented at Nvidia's GTC conference, the Blackwell architecture is set to significantly outperform its H100-class predecessor, offering 2.5 to 5 times the performance and doubling both memory capacity and bandwidth. While a mid-2024 release was initially planned, the schedule has now shifted to Q4.
Market Reactions and Influences
Investor unease has been partly driven by share price volatility, which coincides with a Department of Justice subpoena and a patent lawsuit from DPU vendor Xockets. Huang attributed the increasing demand for Blackwell GPUs to the rising interest in generative AI technologies like ChatGPT, which are becoming critical for various industries.
During the conference, Huang addressed ROI concerns for the high-cost GPU systems that power the AI surge. He highlighted that performance gains from these GPUs make the elevated infrastructure costs worthwhile, pointing out that technologies like the Spark data processing engine could see a 20-fold speed increase with GPU acceleration, rendering infrastructure expenses more manageable.
Huang remains positive about the ROI from AI technologies, suggesting that for every dollar invested in Nvidia's infrastructure, service providers can earn $5 in rentals. Nonetheless, the debate continues over the long-term viability of dedicated AI accelerators.
Investments in Supply Chain and Production
Nvidia has significantly invested in its supply chain to ensure the efficient rollout of Blackwell GPUs. Huang emphasized collaborating with partners to secure essential components amid a global semiconductor shortage affecting multiple industries, including automotive and electronics.
Huang criticized the inefficiencies of traditional air-cooled datacenters and highlighted Nvidia's SuperPODs, which offer high-density computing power in a compact form. He advocates for smarter datacenter designs to slash costs, replacing thousands of conventional nodes with more powerful, compact systems.
