Ian Hunt, export controls specialist at Clear Border, headshot portrait

Ian Hunt

Trade Control Consultant

Executive summary

A technology supply chain needs optionality at both ends. Can you still get what you need to build a product, and can you still sell the product where you expected? Resilience depends on viable commercial options: knowing which alternatives exist, what they cost, and how quickly they can be activated.

Key insights

  1. Build optionality at both ends of the supply chain. Alternative suppliers and manufacturing locations protect your ability to build. Alternative products, customers, and markets protect your ability to sell.
  2. Measure your optionality gap. An alternative in theory has very limited value if qualification, licensing, redesign or production transfer takes months.
  3. Track your time-to-alternative. When tariffs or market access change, know how quickly the financial impact can be quantified – and how quickly supply, production, inventory, or revenue can be redirected.

International tech businesses operate between two movable constraints.

Upstream, there are tariffs, origin, suppliers, and manufacturing footprints – all of which determine whether products can still be built competitively.

Downstream are export controls, sanctions, licences, and destination-country restrictions. These influence whether those products can still reach their intended customers.

Geopolitics can alter either side of the equation… sometimes, both at once. Therefore, technology supply chains need optionality at both ends.

“Optionality you can’t activate quickly is not much of an option.”

Why this matters

Policy moves faster than technology supply chains can. Suppliers need qualifying, factories require tooling, products take time to redesign, and revenue can’t always migrate between markets overnight. The ability to identify and activate alternatives has commercial value – even before disruption occurs.

Governance advisory on international trade →

Routes to market aren’t set in stone

In April 2025, the US government introduced a new licence requirement for NVIDIA’s H20 AI chips destined for China. The fallout happened rapidly: NVIDIA recorded a $4.5 billion charge associated with excess H20 inventory and purchase commitments, and was unable to ship an additional $2.5 billion of H20 revenue during the quarter.

Then market access shifted again.

US licences reopened limited routes for H20 products, generating around $60 million in revenue. By February 2026, NVIDIA had also received licences permitting small quantities of its more powerful H200 chips to reach specified Chinese customers. Those chips first had to undergo inspection in the US, exposing them to a 25% tariff on importation into the country.

Then… the destination itself became the constraint.

Chinese government restrictions prevented NVIDIA from selling all the H200 products covered by its US licences. By July 2026, the company had incurred another $400 million charge associated with excess H200 inventory and purchase obligations. Shipments made under the H200 licensing programme represented less than 1% of its latest quarterly Data Center revenue.

Note that the technology remained viable throughout. It was the executable route between product and customer that kept changing.

“It’s a ‘shifting sands’ situation. Yesterday’s growth market can become tomorrow’s stranded inventory.”

NVIDIA is not an isolated case. AMD incurred approximately $440 million in net inventory and related charges during 2025, following restrictions affecting its MI308 products. Its filings now explicitly state that China sales depend on:

  • Customer demand
  • Chinese import controls
  • Its ability to secure US licences

Your plan B needs a lead time

Another supplier, another manufacturing location, another market. Perhaps even another product configuration designed around new regulatory thresholds. Tech businesses usually have alternatives – at least, in theory.

The strategic weakness lies in assuming that the existence of an alternative makes it immediately usable.

The optionality gap is the distance between having an alternative on paper, and being able to use it commercially.

“A low-cost supply chain with no room to move could become a very expensive supply chain overnight.”

Leadership teams must measure time-to-alternative.

  • Do we have another supplier?
  • How many days until it can supply us?
  • Do we have another market?
  • How much revenue can migrate there? How quickly?
  • Could we redesign the product?
  • How long until it’s approved, manufacturable, and competitive?

A supplier that requires nine months to qualify provides very different resilience from one capable of taking volume next month. The same applies to manufacturing capacity, alternative markets, and product redesign.

Markets can reopen, too

Optionality also matters when restrictions disappear.

On 29 May 2025, chip-design software and semiconductor IP company Synopsys received new US export restrictions relating to China. The company suspended its financial guidance while assessing the potential impact.

On 2 July – little more than a month later – those restrictions were rescinded with immediate effect. Synopsys began restoring Chinese customers’ access to the affected products.

In just over a month, the commercial position had moved from accessible, to restricted, to accessible again.

Therefore, commercial resilience means more than beating a hasty retreat when markets close. Firms need the operational intelligence to recognise when conditions change again – and to decide whether returning makes sense.

How quickly can your firm move?

“Policy can change quickly. Supply chains cannot.”

CFOs, procurement teams, supply chain leaders, and commercial decision-makers need to know:

  • How quickly can we quantify the margin impact of a new tariff?
  • How long would an alternative supplier take to qualify?
  • Which revenues depend disproportionately on one market or destination?
  • How quickly could inventory or production be redirected?
  • Could products be redesigned if access restrictions changed?
  • Where is our largest optionality gap?

For a cross-border tech firm, resilience is measured in options. The strongest businesses know which options are real, what they cost, and how quickly they can activate them. Yesterday’s market can become tomorrow’s restriction faster than a supply chain can be rebuilt.

Borders For the Boardroom: Episode 11 | Navigating the geopolitical turbulence

Christopher Salmon examines the importance of geopolitical context in relation to trade, sanctions, and tariffs.

Listen now on Spotify →

Listen now on Apple →

Other interesting reads

Strategy & Horizon Scannning, Thought Leadership

Reflections on my visit to Shanghai

On a recent visit to China to catch up with clients and partners, Chris reflects on how the balance of power in global trade is shifting, and what the Xi, Trump and Putin meeting means for firms caught between competing powers.
Reflections on my visit to Shanghai
Governance, Thought Leadership

Microsoft 365 outage: when trade compliance risk lives in your email inbox

The November Microsoft 365 outage exposed a vulnerability in international trade governance: critical compliance decisions may depend on communication systems that can become unavailable when businesses need them.
Microsoft 365 outage: when trade compliance risk lives in your email inbox
Export Controls, Thought Leadership

Nvidia Export Control Allegations: National Security Risks and Compliance Lessons

US investigators are examining whether Nvidia AI chips reached restricted Chinese users. See what it means for export controls and download our checklist.
Nvidia Export Control Allegations: National Security Risks and Compliance Lessons