Weekly: US, China in two separate tech races
9.5 min read.
Highlights
US-China tech race. Two pieces on the US-China tech competition by Jake Sullivan in Foreign Affairs and Kyle Chan in Brookings/Congressional Testimony. They both make similar arguments, namely that the US and China are in fact engaged in two different races, with the US barrelling towards an innovation ‘finish line’ that does not exist, while China has a greater emphasis on diffusion and application of new technologies to the real world. Sullivan, Biden’s National Security Advisor and currently a professor at Harvard, frames his argument in prescriptive, grand strategy terms, while Kyle Chan, of Brookings, offers a nuanced portrait of China’s recent technological advancements.
On chip controls. Alasdair Phillips-Robins and Noah Tan of the Carnegie Endowment pitch a nuanced take on well-worn debate over chip exports, arguing that chip controls should be adjusted relative to China’s compute capabilities. Calculating this would be a function of both quantity and quality of chips.
ISSCC write-up. SemiAnalysis offers their signature deep write-up of ISSCC, a major semiconductor conference. They dive deep into Samsung’s latest technical paper on their HBM4, deeming it competitive with peers.
Thanks for reading.
Table of Contents
Jake Sullivan, “The Tech High Ground: What It Will Take to Gain the Advantage Over China,” Foreign Affairs, 04/15/2026.
Kyle Chan, “Competing AI strategies for the US and China,” Brookings, 04/16/2026.
Alasdair Phillips-Robins and Noah Tan, “The Right Way to Sell Chips to China,” AI Frontiers, 04/13/2026.
Afzal Ahmad, Gerald Wong, Daniel Nishball, et al., “ISSCC 2026: NVIDIA & Broadcom CPO, HBM4 & LPDDR6, TSMC Active LSI, Logic-Based SRAM, UCIe-S and More,” SemiAnalysis, 04/16/2026.
1.
Jake Sullivan, “The Tech High Ground: What It Will Take to Gain the Advantage Over China,” Foreign Affairs, 04/15/2026.
The countries that prevail in great-power rivalries are those that adapt. Athens and Sparta and their allies constantly innovated so their navies could outperform one another. During the Cold War, the United States and the Soviet Union spent nearly two decades engaged in a space race. Now, technology is the central front in U.S.-Chinese competition and in the broader contest to shape the world, and the United States must adapt again. This rivalry is playing out across frontier sectors including semiconductors, artificial intelligence, biotechnology, and clean energy. To prevail, Washington needs a clear definition of success and a clear and consistent strategy for how to achieve it.
For decades, U.S. policy toward China rested on a quiet but powerful assumption: Beijing was essentially running the same race as the United States, just a few steps behind. China was seen as a copycat—adept at imitation, lagging on innovation, and ultimately dependent on access to Western technology. The American lead was assumed to be durable, perhaps even self-sustaining.
That assumption has not been borne out. China has moved beyond simply chasing American innovation. It is pursuing a different theory of power: one that places production, scale, and control of critical inputs at the center of its national strategy. As the United States has focused more narrowly on maintaining its lead in innovation breakthroughs, confident that these would cascade naturally into economic, military, and soft power, China has focused on the cascade—aiming to translate technological advances into applied capabilities across its economy and national security enterprise. In other words, while the United States has been running one race, China has been running another. Although this shift unfolded gradually, its consequences are now impossible to ignore. In sector after sector, China has built or is building dominant positions in many of the foundational layers that underpin the modern economy.
Americans tend to see the competition as a race to a finish line—a contest to see which country reaches the next exciting innovation first. But that framing is misleading and counterproductive. This contest has no end date. Success will not manifest as a single moment of triumph with one side declaring victory. Nor will it come from running fast in a single lane. Instead, this competition will extend indefinitely, across a wide variety of sectors. It is no longer enough to be the first to discover new advancements if others are faster at deploying them, or to lead in design if the inputs and capacity vital to production sit beyond the control of the United States or its allies. Washington’s goal must be to establish all these forms of advantage at once.
The point of this contest is not simply to “beat” China. If the United States pulls ahead of China on some relative metric but fails to deliver on making its people more secure or creating greater opportunity for them, then it will have failed—full stop. Success will require fostering a techno-industrial base that drives continuous innovation, rapidly adapting the U.S. military to deter major conflicts, and spreading American digital infrastructure and standards, all while remaining open to cooperation with China over shared interests to prevent a race to the bottom that leaves the whole world worse off.
Securing these objectives must be the central task of American statecraft in the twenty-first century. Doing so will require changes in mindset that transcend partisan boundaries and persist across multiple administrations. But locking in those changes now is urgent, because technological power is translating directly and rapidly into geopolitical power to a degree the world hasn’t seen in years. And for the first time in a long time, the United States is facing a genuine peer competitor.
2.
Kyle Chan, “Competing AI strategies for the US and China,” Brookings, 04/16/2026.
China is pursuing a full-stack approach to AI development, from chips and compute infrastructure to foundation models and applications. The goal of Chinese policymakers is not to achieve AGI, or “artificial general intelligence,” but to leverage AI as a powerful, general-purpose technology that will turbocharge a wide range of sectors and services. National programs, such as China’s 2017 New Generation AI Development Plan and its more recent “AI Plus” Initiative, are geared toward integrating AI into manufacturing, health care, drug discovery, scientific research, education, and government services. China’s military and security agencies have long sought to use AI to improve both defensive and offensive capabilities.
China’s top AI models continue to lag behind American frontier models by several months or more. American AI models maintain a clear lead in overall performance across a wide range of industry benchmarks, from math and reasoning to code generation and long-horizon agentic tasks. Chinese AI labs, particularly startups, are constrained by access to compute, due to a combination of U.S. export controls on advanced AI chips and limited capital resources. Alibaba, one of China’s largest AI players, plans to invest over $53 billion in AI over three years. In contrast, Microsoft spent approximately $80 billion on AI capital expenditures in 2025 alone. America’s main hyperscalers—Alphabet, Amazon, Meta, and Microsoft—have plans to spend a total of $650 billion just this year. American data centers are reaching gigawatt scales and deploying hundreds of thousands of AI accelerators. If everything boiled down to compute and the race to AGI, the United States would hold the decisive advantage.
But China is pursuing a different approach to AI. While some Chinese AI companies, such as DeepSeek and Alibaba, also talk about trying to achieve AGI, Chinese policymakers and China’s AI industry as a whole are more focused on running several different AI races. They are more focused on making progress in model efficiency, AI adoption, and the integration of AI into the physical world. China’s focus on these dimensions of AI development is the result of several factors, including industry constraints—particularly access to large-scale compute and capital—as well as Beijing’s policy priorities. In addition, China is aggressively pursuing semiconductor self-sufficiency, seeking to localize nearly every major segment of the semiconductor supply chain in the face of U.S.-led export controls.
3.
Alasdair Phillips-Robins and Noah Tan, “The Right Way to Sell Chips to China,” AI Frontiers, 04/13/2026.
Last December, President Trump announced that the United States would allow Nvidia to sell its powerful H200 AI processors to customers in China. Officials in the Trump administration have long argued that the best way to win the AI race is to promote the export of US technology around the world, not to restrict it. Selling H200s, the administration claims, will boost the market share of US chip-makers while preserving the US hardware lead.
Critics warn of national security risks to selling advanced chips, but the administration appears committed to its course. Taking a pro-export framework as a given, policymakers can balance between market share and national security by managing a quantity that currently doesn’t receive enough attention: the United States’ relative compute advantage. Total AI compute for a country is calculated as its stock of AI processors weighted by the effectiveness of each processor; the US currently enjoys a roughly 10-to-1 compute advantage over China. That advantage matters because more compute can support more domestic R&D, more customers served by American AI products, and more ability for the US government to influence the development and use of AI.
Relative compute advantage is about the quantity of chips as much as the quality. Administration officials emphasize that the H200 has been superseded by Nvidia’s more powerful Blackwell generation, but many of the world’s largest AI supercomputers still use H200s. With enough of them, Chinese developers may be able to train and deploy AI models that are competitive with US models. Chinese companies have reportedly placed orders for more than 2 million H200s already.
Whereas Biden-era export controls attempted to make the US compute advantage as large as possible, an export-friendly framework could instead focus on maintaining a fixed, favorable compute advantage. To do so, policymakers should peg the quality of exported chips to the performance of China’s domestically manufactured alternatives, while capping quantities of those US chip exports.
4.
Afzal Ahmad, Gerald Wong, Daniel Nishball, et al., “ISSCC 2026: NVIDIA & Broadcom CPO, HBM4 & LPDDR6, TSMC Active LSI, Logic-Based SRAM, UCIe-S and More,” SemiAnalysis, 04/16/2026.
There are three major semiconductor conferences each year, IEDM, VLSI and finally ISSCC. We have covered the former two in great detail over the past few years. Today, we finally complete the trinity with our roundup on ISSCC 2026.
Compared to IEDM and VLSI, ISSCC has a much bigger focus on integration and circuits. Almost every paper comes with some form of circuit diagram, together with clear measurements and data.
In past years, ISSCC findings have been hit or miss when it comes to industry impact. This year was different, a significant number of papers and presentations were directly relevant to market trends. Topics covered range from the latest advancements in HBM4, LPDDR6, GDDR7, and NAND, to co-packaged optics, advanced die-to-die interfaces, and advanced processors from the likes of MediaTek, AMD, Nvidia, and Microsoft.
In this roundup, we will cover major categories such as Memory, Optical Networking, High-Speed Electrical Interconnects, Processors.
Memory
One key theme that caught our attention at this year’s ISSCC was memory, including Samsung HBM4, Samsung and SK Hynix LPDDR6, and SK Hynix GDDR7. Other than DRAM, logic-based SRAM and MRAM also piqued our interest.
Samsung was the only one among the top three memory vendors to present a technical paper on HBM4. Before ISSCC, we noted in our Accelerator & HBM model that Samsung had made great improvements in their HBM4 generation over their HBM3E. The data presented at ISSCC confirmed our analysis, with Samsung posting best-in-class performance - we have also detailed this development months ago, in a model update note.
The technical details presented at ISSCC, combined with industry chatter we have gathered, clearly demonstrate that Samsung’s HBM4 is competitive with its peers. Notably, it can meet the pin speed required for Rubin while staying below 1V. While Samsung still lags SK Hynix in terms of reliability and stability, the company has made meaningful progress in closing the gap on the technology front and could challenge SK Hynix’s dominance in HBM. Their 1c-based HBM4 paired with an SF4 logic base die appears to deliver stronger performance in pin speed.
–