Quick View: Are NVIDIA investors missing the woods for the trees?
Portfolio Manager Richard Clode shares the main insights from NVIDIA’s latest quarterly earnings call, following the US tariffs announcements and H20 chip ban.

6 minute read
Key takeaways:
- Surging AI token generation is enabling new use cases from AI applications and driving greater intelligence, providing credence to the debate around the justification of huge amounts of AI capital expenditure.
- NVIDIA’s Blackwell rack design is bringing next generation AI infrastructure into the AI reasoning world, aiming to meet ever increasing demand requirements for machine learning, inferencing and reasoning models.
- Innovation in reasoning models is driving a pivot in AI infrastructure demand from training clusters to inferencing to support an inflection in token generation, creating a longer, larger and more sustainable runway for tech companies’ growth.
NVIDIA’s pre-announced largest inventory write down in history post the H20 chip China ban following President Trump’s tariffs announcement in April, as well as ongoing manufacturing issues for the Blackwell racks led to another results announcement preceded with trepidation.
 Tokens are (really) the new oil
Every nation now sees AI as core to the next industrial revolution, a new industry that produces intelligence….I don’t know any company, industry, country who thinks that intelligence is optional.
                                                          Jensen Huang, NVIDIA Founder and CEO
AI tokens (the units of data processed by AI models during training and inference) are the new oil, given the recent AI chip deals announced by President Trump and Jensen Huang in Saudi Arabia and the UAE. Microsoft also disclosed it had 100 trillion tokens generated in Q1 25, up 5x year-on-year, and 50 trillion tokens processed in March. Google, meanwhile, at their I/O event last week revealed it is now processing over 480 trillion tokens per month, up 50x year-on-year.
NVIDIA asserted that token generation has surged tenfold over the past year. This is a very important inflection for a number of reasons. The market has been debating the use cases of AI, with token generation being a key metric of usage as well as intelligence. This surge also highlights the misconceptions from the ‘DeepSeek moment’ earlier this year when the market overly focused on the lower AI capex required to train the model, and ignored the significantly greater compute intensity of the reasoning model DeepSeek released. According to NVIDIA reasoning models can be 100x to 1000x more compute intensive than a single shot chatbot query given the model is ‘thinking’, exploring multiple pathways to solve a more complex query and checking its answer. That also generates exponentially more tokens, while also enabling new use cases from that greater intelligence, be it agentic AI for consumers or enterprises, or physical AI-powered autonomous vehicles and humanoids. AI innovation is proliferating AI use cases providing credible return on investment and more sustainable AI infrastructure spending due to that return. This is supporting the expanding usage of AI via inferencing rather than just throwing more compute at training up ever larger frontier AI models.
Blackwell built for a reasoning world
The Blackwell NVL72 AI supercomputer is ‘a thinking machine’ designed for reasoning.
                                                                          Jensen Huang
In the same way the Hopper chip was specifically designed to train large language transformer models, NVIDIA specifically designed its Blackwell chip to meet the performance demands of next generation reasoning models. Interacting with a reasoning model and ensuring a reasonable quality of service, as well as avoiding prohibitive cost or power consumption required NVIDIA to design at the rack level for the first time ─ the NVL72 weighs in at two tonnes with 1.2 million components. The result is 40x the inferencing performance of Hopper, bringing next generation AI infrastructure into the AI reasoning world, answering many questions from last year around the relative merits of ASICs vs GPUs.
The ramp up of Blackwell has been far from serene with a re-spin of the underlying chip at TSMC late last year, plus recent board and yield issues with the GB200 racks. However, these issues appear to be finally in the rear-view mirror, allowing the market to look forward to a potential significant ramp up in rack supply through the rest of the year. NVIDIA talked about hyperscalers now deploying on average nearly 1,000 NVL 72 racks per week, or 72,000 Blackwell GPUs per week. Blackwell Ultra GB300 racks are also now sampling and with the same physical footprint and rack design as the GB200, this could result in a much smoother supply ramp from H2 2025 to meet burgeoning inferencing demand with around 50% greater performance.
Geopolitics and deglobalisation implications
The question is not whether China will have AI, it already does. The question is whether one of the world’s largest AI markets will run on American platforms.
                                                                                 Jensen Huang
AI is at the front line of geopolitics and superpower supremacy. Following the H20 ban, we appear close to an end game for NVIDIA’s datacentre business in China, but Jensen Huang continued to make an impassioned plea that these restrictions are misguided. This is because they will not stop China from having AI or AI chips, they already have both. A superior strategy would be to ensure one of the world’s largest AI markets with half the world’s AI developers is built on US platforms and infrastructure.
NVIDIA continues to explore options to supply China with a further degraded AI performance chip, but it remains to be seen whether that will be allowed or whether those chips can be competitive. On a more positive note, the rescinding of the Biden administration’s AI Diffusion Rule that was meant to go into effect on 15 May enabled the recent deals in Saudi Arabia and the UAE. NVIDIA also clarified that while Singapore contributes a significant portion of its sales, the vast majority of that AI compute ends up with US-based customers, refuting recent reports of diversion of chips to circumvent export restrictions.
Finally, NVIDIA confirmed that in the space of a year it will be producing AI chips from TSMC fabs in Arizona, and assembling AI supercomputers in factories in Texas. NVIDIA’s support of President Trump’s aim of building manufacturing capacity in the US continues the deglobalisation of supply chains we have witnessed post-pandemic in the face of geopolitics, trade wars and demographics.  Â
Our goal, from chip to supercomputer, built in America within a year.                                                                                                                      Jensen Huang
Missing the woods for the trees
While the markets typically look to NVIDIA’s earnings releases as a bellwether for the tech sector and geopolitical signals given its dominant role in the AI wave, we believe the short-term focus risks missing the wood for the trees. As AI continues to innovate with reasoning models, agentic and physical AI are creating compelling use cases and monetisation models. This is driving a pivot in AI infrastructure demand from training clusters to inferencing to support an inflection in token generation as reasoning models and their usage proliferate, thus creating a longer, larger and more sustainable runway for growth in the tech sector and the global economy.
Sources: NVIDIA 1st Quarter FY26 Financial Results earnings call transcript, 28 May 2025. NVIDIA blog; NVIDIA Newsroom.
AI Diffusion Rule: the rule planned to cap the export of essential US AI technology to limit its spread outside the US. The argument being that the rule would impinge America’s ambition for AI global supremacy and be a tailwind to global competitors to fill the void.
Agentic AI: Uses sophisticated reasoning and iterative planning to autonomously solve complex, multi-step problems. Vast amounts of data from multiple data sources and third-party applications are used to independently analyse challenges, develop strategies and execute tasks.
ASICs vs GPUs: ASICs are custom-designed semiconductors built to perform specific tasks giving them certain advantages versus GPUs for AI, for example being more cost-effective than GPUs.
Capex: capital expenditure refers to company spending to acquire or upgrade physical assets such as buildings, machinery, equipment, technology etc. to maintain or improve operations and foster future growth.
DeepSeek: A Chinese AI startup and developer of open-source advanced large language models (LLMs) such as DeepSeek-V3 – a key rival, and less expensive option compared to OpenAI’s ChatGPT and Google’s Gemini.
Graphics Processing Unit (GPU): A unit that performs complex mathematical and geometric calculations necessary for graphics rendering, of particular use in high-end gaming, content creation and machine learning.
Hyperscalers: companies that provide infrastructure for cloud, networking, and internet services at scale. Examples include Google Cloud, Microsoft Azure, Facebook Infrastructure, Alibaba Cloud, and Amazon Web Services.
H20 chip ban: on 9 April 2025 the US government informed NVIDIA that it required a license to export its H20 chips to China, resulting in the company writing down US$5.5 billion of revenue in its fiscal first quarter 2026.
Inferencing: Refers to artificial intelligence processing. Whereas machine learning and deep learning applies to training neural networks, AI inference applies knowledge from a trained neural network model and uses it to infer a result.
NVIDIA Hopper and Blackwell chips: Hopper launched in 2022 and introduced powerful AI capabilities excelling in Large Language Models and scientific computing. Blackwell released in 2024, is designed for the more complex demands of generative AI and large-scale simulations.
Rack: rack mounted servers improve data centre space utilisation, designed for diverse computing infrastructure needs and comprehensive server management. They adhere to uniform standards, allowing them to be conveniently stacked in a one metal enclosure or casing.
Reasoning: making use of available information to generate predictions, make inferences and draw conclusions. It involves representing data in a form that a machine can process and understand, then applying logic to arrive at a decision.
Token: AI tokens are the fundamental building blocks of input and output that Large Language Models (LLMs) use. These units of data are processed by AI models during training and inference, enabling prediction, generation and reasoning.