AI chip demand is contributing to ongoing FPGA shortage

Competition for advanced processors, memory, and supporting infrastructure is intensifying under sustained AI demand. While memory has taken center stage in the news with its hyper-fueled price shocks, the FPGA shortage has been quietly expanding to new highs. Lead times are stretching out as hyperscalers and AI infrastructure companies secure whatever inventory is available.  

At the same time, the U.S. government is expanding export restrictions to control the flow of advanced AI chips and related technologies. This move comes a week after U.S.-China trade talks, which saw the allowance for some Nvidia chips to be sold to Chinese companies.  

AI infrastructure demand continues to accelerate

Memory shortages have dominated the headlines in 2026, but the supply pressure driving FPGA lead times to 52 weeks and beyond has the same culprit to thank: data center demand. As hyperscalers race to build out AI compute infrastructure, they’ve received priority allocation across categories thanks to their massive budgets and influence. Downstream, the effects are being felt by every industry that relies on the same components.  

The FPGA market structure makes it particularly vulnerable to this tidal wave of demand. With AMD’s acquisition of Xilinx, supply is highly concentrated in a single provider. When a data center customer with greater purchasing power and longer-horizon commitments moves to the front of the line, everyone else waits. Right now, with nowhere else to turn, those wait times are over a year.

The CPU market has more participants than either the FPGA or the memory sector’s three-player oligopoly, but that breadth hasn’t equaled relief as the largest buyers are still locking up allocation across multiple vendors.  

Sever-grade CPUs have followed a similar pattern to FPGA. Most notably, Intel has shifted Xeon allocation toward hyperscale cloud customers, leaving other buyers fighting to secure inventory. Intel has also delayed its next-gen Diamond Rapids CPU until mid-2027, elongating the upgrade cycle for industrial buyers and adding pressure to short supply or current-gen CPUs.  

As a result, prices for some configurations have skyrocketed in recent months. Though this is just one data point of many, it’s a sign that wherever there is demand from AI infrastructure, other buyers must negotiate around it.  

Tech companies have seen this trend and are making moves accordingly. Supply chain diversification has suddenly become a hot topic as buyers hedge against supplier concentration risk. This goes for firms as massive as Apple, which is currently courting additional manufacturers for its in-house chips after more than a decade of exclusive partnership with TSMC.  

Investment in custom AI chips and inference-optimized processors is also accelerating as companies seek to compete on AI operating costs. Organizations that built their first-wave infrastructure on generic silicon are now pivoting to internal chip programs specifically designed for efficiency within their workflows.  

To say this is a difficult market, regardless of industry, is an understatement. As data center demand sucks away inventory at a record pace, procurement teams must be able to anticipate where and when components are available and react quickly.  

Sourceability’s global sourcing expertise, franchise partnerships, and Datalynq market intelligence platform give you the visibility you need to stay ahead of tightening conditions and keep your supply chain intact.  

U.S. expands efforts to restrict AI chip exports

The regulatory environment governing advanced chip exports has shifted once again as the U.S. aims to block the export of certain high-end AI chips.  

According to TrendForce, the U.S. Department of Commerce’s Bureau of Industry and Security (BIS) is now moving to close a loophole that allowed Nvidia’s Blackwell and Rubin and AMD’s MI350X processors to reach overseas subsidiaries of Chinese companies.  

Earlier attempts to restrict China's access to advanced AI silicon focused on direct shipments, leaving indirect routes largely unaddressed. Overseas subsidiaries, primarily in Southeast Asia, operated as effective pass-throughs that allowed parent firms to bypass licensing frameworks.

By refusing to enforce the Biden administration’s AI Diffusion Rule last year, Washington effectively left open a pathway for AI chips to reach Chinese entities—whether directly or through third-party buyers. Officials estimate that hundreds of thousands of chips have already traded hands as a result of the loophole the BIS is now trying to close.  

Notably, the new guidance focuses on top-tier chips from the likes of Nvidia and AMD. According to TrendForce, “Lower-tier chips supplied under existing licenses may continue to be sold under current terms, while products already shipped will remain with customers.”  

Exporters will now have to look beyond their direct customers and verify each buyer’s ultimate parent company before doing business. While the new guidance adds a fresh layer of complexity for distributors, foundries themselves won’t have extra verification work to do.  

That exemption is noteworthy since, without upstream verification requirements, the indirect routing problem isn’t solved. The new guidance simply targets one workaround without blocking the rest.  

Regardless, each successive round of export controls narrows the set of viable procurement pathways and adds procurement complexity for any organization operating across multiple jurisdictions. Multinational buyers must be proactive in ensuring their supply chains remain in compliance even while dealing with an ever-expanding web of regulations.  

For global supply chains, the cumulative effect is fragmentation. Markets that once leveraged efficient cross-border collaboration are now reassessing as doing business across regions is becoming less fruitful. This goes for procurement, too, as strategies built for a stable, open trading environment must rapidly diversify lest they collapse entirely.  

Sourceability’s expertise and global reach makes it possible to navigate this tangled web. Through comprehensive supply chain visibility and diversified sourcing solutions, we help customers proactively reduce exposure to regulatory disruptions before the next round of guidance forces a reactive response.