BREAKING
US DRAFTS SWEEPING NEW CHIP EXPORT CONTROLS — MARCH 5 2026 — BLOOMBERG CHINA FIVE-YEAR PLAN TARGETS AI + QUANTUM + HUMANOID ROBOTS — NPC SESSION MARCH 5 TRUMP H200 EXPORT FLIP: DENY → ALLOW → TARIFF 25% → NEW CONTROLS DRAFTED FOXCONN REVENUE +21.6% ON NVIDIA SERVER DEMAND — FIRST 2 MONTHS 2026 IBM: 2026 MARKS FIRST QUANTUM OUTPERFORMANCE OF CLASSICAL COMPUTER CHINESE FIRMS ORDERED $14B IN H200 CHIPS — BYTEDANCE LEADS MINIMAX M2.5 RIVALS CLAUDE OPUS 4.6 AT FRACTION OF COST — MARCH 2026 CHINA COMPUTE CAPACITY: US HOLDS 10X LEAD — WITHOUT H200 EXPORTS US DRAFTS SWEEPING NEW CHIP EXPORT CONTROLS — MARCH 5 2026 — BLOOMBERG CHINA FIVE-YEAR PLAN TARGETS AI + QUANTUM + HUMANOID ROBOTS — NPC SESSION MARCH 5 TRUMP H200 EXPORT FLIP: DENY → ALLOW → TARIFF 25% → NEW CONTROLS DRAFTED FOXCONN REVENUE +21.6% ON NVIDIA SERVER DEMAND — FIRST 2 MONTHS 2026 IBM: 2026 MARKS FIRST QUANTUM OUTPERFORMANCE OF CLASSICAL COMPUTER CHINESE FIRMS ORDERED $14B IN H200 CHIPS — BYTEDANCE LEADS MINIMAX M2.5 RIVALS CLAUDE OPUS 4.6 AT FRACTION OF COST — MARCH 2026 CHINA COMPUTE CAPACITY: US HOLDS 10X LEAD — WITHOUT H200 EXPORTS
INTEL BRIEF · MARCH 6, 2026 · LIVE
Breaking · Published Today · Sourced from Reuters, TechCrunch, Bloomberg, CSIS

The Compute War America Just Sold
The Ammunition.
Three years of chip export controls. One policy reversal. And China just placed a $14 billion order.

The US spent three years building a wall around its AI advantage. Then unlocked the gate — at 25% markup. Now, one day later, the same administration is drafting rules that would require US government approval to ship AI chips anywhere outside America. This is what's actually happening.

The most important resource in the global AI race isn't talent. It isn't data. It isn't even the models. It's the hardware to run them — and for the past three years, the United States has controlled who could have it. On January 14th, 2026, that policy broke in half. Yesterday, March 5th, it started rebuilding. Here's what you need to understand about the compute war that's defining who builds the future.

§01 The Logic of the Wall

The original export control strategy had clean logic: AI capability scales with compute. If you can limit China's access to high-performance AI chips — specifically NVIDIA's H100 and H200 GPUs — you limit their ability to train competitive frontier models. No chips, no training runs. No training runs, no GPT-4 competitor. The gap stays open.

The strategy was built on a specific bet about AI development: that raw compute scaling was the primary driver of capability gains, and that whoever had the most chips would win the capability race. If that bet is right, cutting off China's chip supply is as strategic as an oil embargo. You are not blocking a product — you are blocking the fuel.[1]

China wasn't helpless. Huawei built domestic alternatives. But those alternatives operated at 60–70% of H200 capability — and could only be produced in the hundreds of thousands, not millions. NVIDIA produces millions. The gap was real, and it was growing.

US · Compute Advantage
10×
US compute capacity lead over China if no H200 exports. Source: BISI, Feb 2026.
China · H200 Orders
$14B
Value of H200 orders placed by Chinese firms, led by ByteDance. Source: BISI, Feb 2026.
Huawei Gap
60%
Huawei's best domestic chip operates at 60–70% of H200 capability. Production limited to hundreds of thousands vs NVIDIA's millions.
Trump Tariff
25%
Revenue cut imposed on H200 exports to China. Effective Jan 15, 2026. The US gets a toll on its competitor's compute supply.

§02 The Reversal

On January 13, 2026, the US Department of Commerce's Bureau of Industry and Security quietly changed one word in its export license review policy. The word was "denial." The policy had previously operated under "presumption of denial" for advanced AI chips destined for China. BIS changed it to "case-by-case review." One day later, Trump announced approval for NVIDIA to sell H200 chips to China — with a 25% revenue tariff.

Former National Security Adviser Jake Sullivan's response: "This decision is nuts."[2]

Oct 2022
Biden Administration — First Controls
BIS restricts H100 and A100 exports to China. Strategic logic: deny compute, deny capability. Scaling laws make this a plausible theory of victory.
Jul 2025
Trump allows H20 exports at 15% revenue share
The H20 is a scaled-down, export-compliant chip. Less powerful, but significant. First signal that the policy is softening. Then reverses — announces ban of H20 exports. Then reverses again.
Oct 30 2025
Trump-Xi summit in Busan
Tariff truce reached. Export controls conspicuously untouched. Within days, Trump signals willingness to allow H200 exports — then reverses within 4 days.
Dec 2025
Biden's AI Diffusion Rule — rescinded
The incoming Trump administration rescinds Biden's tiered global chip framework before it took effect. The comprehensive architecture for managing global chip diffusion: gone.
Jan 13-15 2026
The Reversal — "Presumption of Denial" → "Case-by-Case"
BIS formalizes H200 exports with 25% tariff. Chinese firms immediately begin orders. ByteDance leads $14 billion in expected purchases. The wall drops.
Mar 5 2026 · Yesterday
Bloomberg: New sweeping controls drafted
US regulators allegedly draft rules requiring government approval to ship AI chips anywhere outside the US — including allies. The pendulum swings back. Again.
US regulators have drafted rules that would require US government approval to ship AI chips anywhere outside the United States — a sweeping expansion that would give Washington significantly more control over companies like AMD and Nvidia. The proposal would represent the most extensive government involvement in chip exports in history. A spokesperson for the US Department of Commerce said only that the agency is "committed to promoting secure exports of the American tech stack" and that "there are ongoing internal government discussions about formalizing that approach."
SOURCE: BLOOMBERG, MARCH 5 2026 · VIA TECHCRUNCH

§03 The Bet Under Everything

Every policy decision in this story hinges on a single unanswered question: How close is superintelligence?

This sounds like a philosophical question. It is actually a strategic one. If transformative AI — the kind that could provide decisive geopolitical advantage — is five years away, then blocking China's compute access is an existential strategic move. The chips aren't products. They are weapons. You don't sell weapons to the adversary you're racing.

But if AI scaling has hit a wall — if the era of "more compute = smarter models" is over, replaced by post-training techniques, efficiency gains, and application-layer innovation — then the calculus inverts completely. Strict controls cut NVIDIA out of the world's largest chip market while barely slowing Chinese development, which has already shown it can produce competitive models on domestic hardware. You damaged your own company for minimal strategic gain.[3]

Every decision about AI chip export controls comes down to one question: "Will AI progress accelerate toward superintelligence, or plateau?" Those two worlds require opposite policies — and nobody knows which world we're in.

§04 The Case For and Against

Argument For Loosening Controls
Sell. Win On Revenue.
Scaling has plateaued. The next wave of AI advantage comes from post-training, applications, and execution — not raw compute. China builds competitive models with domestic hardware already.
If China can't buy H200s from NVIDIA, they buy from Huawei. NVIDIA loses the market. Huawei gets funded. Chinese domestic chip capability accelerates. Lose-lose.
Selling H200s ensures the next generation of Chinese AI runs on US hardware with US architecture. That's a form of soft dependency. And the US collects 25% at the door.
The next-gen Blackwell chips stay restricted. H200 is already being surpassed. Selling yesterday's weapon to maintain the hardware dependency is rational industrial strategy.
Argument Against — Keep the Wall Up
Don't Arm The Competition.
H200 is six times more powerful than any US chip currently available in China. The capability gap is real regardless of domestic alternatives. You are closing a 10x advantage by 50-60%.
Chinese companies purchase chips for applications that ultimately serve military infrastructure. The civilian/military distinction is not meaningful in the Chinese tech ecosystem.
The AI Overwatch Act — passed January 22 — introduces permanent uncertainty. Licenses can be revoked by Congress at any time. You sold the ammunition to someone who may be told to send it back.
Even if scaling has slowed, nobody knows when it might accelerate again. A technical breakthrough on the other side of a 10x compute infusion is a different threat than the same breakthrough on Huawei hardware.

§05 China's Move

China's response to the H200 opening has been characteristically opaque. Chinese customs authorities have been instructed to block imports of H200 chips. National technology companies have been warned against purchasing them "unless necessary." The official line is digital sovereignty — the last thing China wants is its AI infrastructure built on hardware with a 25% American tax and a potential revocation clause baked in by an act of Congress.[4]

But ByteDance and others are reportedly ordering anyway. The companies are doing what companies do: acquiring the best hardware available while the window is open, regardless of what the government posture says.

Meanwhile, on March 5th — the same day Bloomberg reported the new sweeping export control draft — China's National People's Congress opened its session with a five-year plan that reads like a military procurement list for the tech cold war: AI throughout the economy, quantum computing dominance, 6G, embodied AI for humanoid robots, nuclear fusion breakthroughs, reusable heavy rockets, scalable quantum computers, a space-earth quantum communication network. "Seize the commanding heights of science and technological development." Decisive breakthroughs in key core technologies.

That language does not read like a country that thinks it is losing.

$500B
Stargate infrastructure commitment over 5 years — US response
2M+
H200 chips ordered by Chinese firms following the January reversal
21%
Foxconn revenue growth first 2 months of 2026 — on NVIDIA server demand alone

§06 What This Means If You're Building on AI

The instinct when reading geopolitical tech news is to treat it as background noise — the stuff happening in the boardrooms and Congresses that eventually trickles down to affect the tools you use. That instinct is wrong here. This is not background noise. This is the infrastructure story.

The compute war determines inference costs. It determines which companies can train which models. It determines whether the frontier keeps moving, and at whose pace. Every decision you're making about which models to deploy, what APIs to call, what architecture to build on — those decisions sit downstream of where this hardware ends up.

The models are becoming commodities. Industry analysts are increasingly saying that large language models have become difficult to distinguish from each other in practice — differentiated only by price and ease of use. If the models are commodities, the war is at the infrastructure layer. Compute, energy, hardware access. That's the actual battlefield. And it is moving fast enough that a sweeping new control framework could be announced before this post is two days old.

The models are becoming commodities. The war is at the infrastructure layer now — compute, energy, who has the hardware and who doesn't. Every API you call sits downstream of that fight.

§07 The Part Nobody Has an Answer For

The honest summary of this situation is that the people making these policy decisions are operating on bet-the-future uncertainty about a question that has no answer: how fast is AI actually progressing, and toward what?

If the Brookings analysis is right — that scaling has hit a wall, that the next breakthroughs are application-layer, that China can get to competitive capability anyway — then the H200 reversal is defensible. Take the 25% revenue cut. Keep NVIDIA in the Chinese market. Win on architecture dependency.

If the China hawks are right — that compute still matters decisively, that a 10x gap is worth preserving at all costs, that a technical breakthrough on the other side of $14 billion in H200s is an existential risk — then the reversal is one of the most strategically damaging decisions in recent tech history.

The fact that we get sweeping new controls proposed the very next day, after the same administration just opened the gate, suggests that the people closest to this don't know either. They are managing a technology whose implications they cannot fully model, with policy tools that were designed for a world where the technology changed slowly enough to deliberate about.

It's March 6, 2026. The situation will have changed before the week is out. Watch the hardware. Everything else is downstream of it.

Sources
[1]BISI Analysis (Feb 2026): "Trump Reverses US AI Chip Export Policy to China" — Carlotta Kozlowskyj. Primary source for compute gap figures and H200 order volumes.
[2]Atlantic Council (Dec 2025): "Why Exporting Advanced Chips to China Endangers US AI Leadership" — Jake Sullivan quoted in NYT: "This decision is nuts."
[3]Brookings (Feb 2026): "If Superintelligence Isn't Imminent, The Trump Administration May Be Right to Loosen Advanced Chip Export Controls." Core framing for the scaling plateau argument.
[4]BISI: China's official posture vs. actual corporate behavior — customs block alongside $14B in ByteDance-led orders. The contradiction is real and documented.
[5]TechCrunch (March 5, 2026): "US Reportedly Considering Sweeping New Chip Export Controls." Bloomberg sourced. Published yesterday.
[6]Reuters/Laurie Chen (March 5, 2026): "China vows to accelerate technological self-reliance, AI push." Five-year plan language quoted directly from NPC session materials.
J
Justin Erickson — PropTechUSA.ai
Watching the infrastructure · March 6, 2026
More from PropTechUSA.ai Research
// The AI infrastructure layer, live
The Consilium

Ten agents building on the hardware this post is about. Ask it something about the compute war.

Open The Consilium