Wolfspeed Is Turning to AI to Fix Its Silicon Carbide Problem
After slashing debt in bankruptcy, the chipmaker is embedding AI across its 200mm fabs to stabilize margins and scale production
Wolfspeed exited Chapter 11 in September 2025 after reducing its total debt by roughly 70 percent under a restructuring support agreement, according to reporting on the bankruptcy process.
The balance sheet reset came as the company continued scaling silicon carbide production. In the second quarter of fiscal 2026, Wolfspeed reported approximately $168 million in revenue. Non-GAAP gross margin was negative 34 percent, and GAAP gross margin was negative 46 percent. Adjusted EBITDA for the quarter was negative $82 million. The company ended the period with about $1.3 billion in cash and short-term investments.
Wolfspeed is expanding 200mm silicon carbide production at its Mohawk Valley facility in New York. In September 2025, the company announced the commercial launch of its 200mm silicon carbide materials portfolio, positioning the site as central to its scale strategy.
On February 11, 2026, Wolfspeed said it had unified factory, supply chain, and enterprise data on Snowflake’s AI Data Cloud and was deploying artificial intelligence across manufacturing, quality, supply chain, finance, and analytics functions. The company introduced WolfGPT, an internal generative AI platform built on Snowflake Intelligence, to analyze manufacturing performance, predict potential issues, and accelerate workforce training.
Scaling SiC at Negative Margins
Silicon carbide fabrication requires high capital investment, specialized tooling, and tight defect control. During production ramps, variability in yield, inspection accuracy, and equipment uptime influences cost per wafer and gross margin.
Wolfspeed’s transition to 200mm wafers is designed to improve long-term manufacturing efficiency. Ramp phases typically involve incremental qualification of processes and customers, which can constrain utilization in early stages. With non-GAAP gross margins at negative 34 percent in the most recent quarter, manufacturing performance remains central to financial recovery.
Competitive capacity is increasing. Infineon Technologies has invested in 200mm silicon carbide production at its Kulim facility in Malaysia. STMicroelectronics has also announced silicon carbide expansion initiatives in Europe, including 200mm manufacturing projects in Catania.
In fabrication environments with high fixed costs, improvements in yield stability and cycle time affect how quickly those costs are absorbed. Semiconductor manufacturers have increasingly deployed machine learning and advanced analytics in inspection and process control systems. Intel Corporation has published case materials describing the use of AI and computer vision to improve defect detection speed and accuracy in manufacturing operations.
Improving defect detection and reducing process variability can increase the number of usable devices per wafer and shorten production cycles. In a 200mm silicon carbide ramp operating with negative margins, those operational shifts influence cost structure and margin trajectory.
AI Integration Across the Enterprise
Wolfspeed’s February announcement outlined a broader integration effort beyond individual inspection tools. The company consolidated operational and enterprise data into a governed cloud environment and deployed specialized AI agents across manufacturing, quality, supply chain, finance, and corporate analytics.
The company says WolfGPT is designed to provide contextual insights from manufacturing data, helping teams surface potential issues earlier in the production flow. The company also cited workforce readiness as a priority, stating that AI tools can accelerate training in complex chip fabrication environments.
Semiconductor manufacturing roles often require extended training and domain expertise. Industry workforce policy briefs highlight ongoing shortages of experienced technicians and engineers in advanced fabrication facilities. Tools that centralize institutional knowledge and provide structured access to operational data can influence onboarding time and reduce decision latency during shifts.
The restructuring reduced Wolfspeed’s debt burden and interest expense. Margin recovery, however, depends on manufacturing performance as 200mm capacity scales. In the most recent quarter, the company also reported sequential growth in AI datacenter revenue, reflecting demand linked to high-performance computing infrastructure.
Wolfspeed’s integration of AI across factory and enterprise systems shows a focus on operational coordination during a period of financial and manufacturing transition. The measurable effects on yield stability, cycle time, and cost per wafer will shape the company’s margin recovery as global silicon carbide capacity continues to expand.