In a move that signals a significant pivot in operational strategy, Andes Technology has confirmed the closure of its Austin, Texas-based subsidiary, Condor Computing. The decision marks the end of a specialized design center tasked with high-performance processor development, with the company opting to fold the proprietary "Cuzco" processor intellectual property (IP) back into its core global operations. This restructuring is the latest development in a series of cost-cutting measures designed to align the semiconductor IP provider with the realities of a shifting market landscape.
The Core Facts: A Subsidiary Folded
Last week’s announcement regarding Condor Computing serves as a milestone in Andes Technology’s ongoing efforts to streamline its business model. The Austin-based team, which had been the focal point of the company’s aggressive pursuit of high-performance RISC-V architectures, has been largely scaled back. While a select group of engineers and essential personnel will remain with the parent company to shepherd the continued development of the Cuzco IP, the independent structure of Condor Computing has been dissolved.
This consolidation is not an isolated event but rather a component of a broader, company-wide mandate to reduce operational expenses by an estimated 10% to 20%. Headcount figures reflect this tightening; from a peak of approximately 480 employees, the company’s global workforce has been reduced to roughly 400.
Chronology of a Strategic Pivot
The seeds of this restructuring were sown long before the recent shuttering of the Austin facility. In an exclusive interview with EE Times in April 2026, Frankwell Lin, CEO and Chairman of Andes Technology, provided a candid assessment of the company’s trajectory.
According to Lin, the company had entered a period of "hyper-expansion" between 2022 and 2025. During this time, headcount and operational costs grew at a rate that ultimately outpaced revenue growth. While Andes experienced a commendable 30% revenue increase in 2024, that growth decelerated to 7% in 2025, triggering a "push for consolidation" from the board of directors.
Lin’s April comments foreshadowed the current state of affairs: “Our organization extended too large in the last three-four years, and the revenue growth was not able to keep up with that kind of burn rate,” he explained. He noted that the company had to act decisively to preserve its long-term viability, citing the need for leaner operations in the face of pricing erosion, particularly within the competitive Chinese semiconductor market.
Furthermore, Lin highlighted that the integration of agentic AI tools into the design workflow had fundamentally changed the company’s labor requirements. By leveraging AI-assisted EDA (Electronic Design Automation) tools, Andes determined it could maintain its current scale of business with fewer human resources, effectively increasing the output per engineer.
Financial Context and Market Performance
Despite the internal turbulence and headcount reductions, Andes Technology maintains a robust financial profile. By July 2026, the company reported cumulative revenue of NT$1,013 million (approximately $31.81 million), representing a significant 56% increase compared to the same period in 2025. This growth trajectory, juxtaposed with the cost-cutting measures, paints a picture of a company transitioning from a "growth-at-all-costs" phase to a "profitable-growth" phase.
The company’s market reach remains extensive. In June 2026, Andes announced a monumental milestone: over 20 billion cumulative shipments of SoCs (System-on-Chips) powered by Andes CPU IP. Their technology is currently embedded in high-profile enterprise and consumer applications, ranging from Meta’s MTIA (Meta Training and Inference Accelerator)—used for recommendation engines—to d-Matrix’s Raptor chiplets for data center AI inference. Furthermore, mass-market adoption remains strong, with Andes IP serving as the engine for Phison SSD controllers and Allwinner Linux-driven camera SoCs.
Official Responses and the "Quiet Period"
Following the reports of the Condor Computing closure, EE Times sought formal clarification from Andes regarding whether the closure was directly linked to the previously announced 10–20% operational expense reduction.
Andes declined to provide specific commentary on the subsidiary’s closure, citing a "quiet period" mandated by an upcoming capital raise involving the issuance of 6.8 million shares. However, the company did issue a broader statement on the matter:
“While we remain committed to our long-term vision on broader computing, we align our organization to best serve the current market momentum on AI, automotive, and embedded systems. All Condor IP has been brought back to Andes for further development, and we continue to work with key Condor personnel.”
This statement underscores a shift in focus: rather than maintaining high-cost, specialized satellite offices for long-term experimental architectures, the company is prioritizing immediate market relevance in the automotive and AI sectors.
Implications: The Future of High-Performance RISC-V
Industry analysts view the closure of Condor Computing as a cautionary tale of the challenges inherent in RISC-V innovation. David Harold, a senior analyst at Jon Peddie Research, noted that while the strategic intent behind the Condor initiative was sound, the execution phase proved difficult.
"Three years is a relatively short window to determine the viability of such a complex, high-performance architecture," Harold noted in his report. "We suspect the strategy was good, but the end results were not up to spec, and a key customer remained elusive."
However, the consensus is that the technology itself—specifically the patented microarchitecture featuring deep out-of-order execution and unique instruction scheduling—will not be abandoned. Instead, it will be integrated into the company’s mainstream AX66 series and future iterations.
Why the Shift Matters
The closure of Condor Computing carries several implications for the broader semiconductor industry:
Consolidation of RISC-V Efforts: As the RISC-V ecosystem matures, companies are moving away from fragmented, specialized R&D teams toward more centralized, highly integrated design workflows.
The Impact of AI on Engineering Efficiency: The adoption of agentic AI in chip design is no longer just a trend; it is a business imperative. Andes Technology is at the forefront of this shift, demonstrating that companies can sustain high-level output while reducing headcount through the automation of complex design tasks.
Sustainability over "Moonshots": The era of aggressive, venture-style expansion for IP providers is being tempered by the reality of global market volatility. Companies are increasingly prioritizing products that align with immediate, high-demand sectors like automotive autonomous systems and data center AI.
Conclusion
Andes Technology is navigating a critical juncture. By absorbing the Condor Computing initiative, the company is demonstrating a pragmatic approach to its R&D portfolio. While the closure of the Austin design center may be viewed as a setback for its long-term "moonshot" processor designs, the move provides the fiscal breathing room necessary to capitalize on its massive success in the 20-billion-unit-shipment market.
As CEO Frankwell Lin indicated in his earlier remarks, the company’s commitment to high-performance RISC-V remains unshakable. The current restructuring is not a retreat, but rather a reorganization—an attempt to ensure that the intellectual capital developed in Austin is not lost, but instead redirected toward the products that are already driving the company’s significant revenue growth in 2026. For Andes, the future is no longer about expanding the footprint of its offices, but about maximizing the efficiency of its silicon, ensuring that it remains the architecture of choice for the next generation of AI and automotive systems.
