The U.S. Defense Logistics Agency has awarded a Phase II Small Business Innovation Research contract to Phoenix Semiconductor, an Austin-based startup specializing in the resurrection of obsolete microelectronic components. The contract, valued at up to $1.2 million over 18 months, tasks the company with recreating exact functional replicas of legacy integrated circuits that are no longer in production but remain critical to weapons systems, radars, and encrypted communications gear—many of which run on curated Windows platforms. This effort addresses a growing readiness gap where shelf-life limitations and single-source supplier failures threaten to strand billions of dollars in military hardware.

Legacy chips built on 180nm, 130nm, or even older process nodes do not benefit from Moore’s Law scaling; instead, they become increasingly difficult to source as original foundries shift to cutting-edge nodes. When a 30-year-old air defense radar requires a specific DSP chip to process incoming threats, and that chip’s last production run was in 1997, the military faces unplanned downtime that can extend for months. Phoenix Semiconductor’s approach uses a combination of delayering, reverse-engineering, and modern wafer-fab partnerships to produce drop-in replacements that match the original component’s electrical, thermal, and timing characteristics exactly—a requirement that rules out simple emulation.

Phase II SBIR contracts are designed to fund prototype development and validation. For Phoenix Semiconductor, that means delivering qualification samples of a specific IC—sources close to the award indicate a radiation-hardened memory controller originally fabricated by a now-defunct supplier for the AN/TPQ-53 counterfire radar—along with full test reports and a pathway to Defense Logistics Agency stock-number qualification. The company must also submit a lifecycle sustainment plan showing how the resurrected part will remain available for the next 20 years, a timescale that aligns with the operational life of the host system, which often runs Windows Embedded Standard 7 or Windows 10 IoT Enterprise.

The Obsolescence Crisis in Defense Electronics

Defense programs are notoriously long-lived. The B-52 bomber, for example, has been in service since 1955 and is expected to fly into the 2050s. The IT backbone of many such platforms is built around ruggedized COTS (commercial off-the-shelf) computing modules running Windows. When a motherboard fails, the depot simply swaps it. But when a proprietary ASIC on that board fails, the entire subsystem is dead. The Defense Microelectronics Activity estimates that $750 million worth of military electronics are currently awaiting replacement chips that have no active source.

Traditional mitigation strategies—buying lifetime inventories upfront or using government-owned foundries—have proven inadequate. A “lifetime buy” of 10,000 units might last a decade, but if the platform’s retirement is pushed because a replacement program is delayed, the stock runs dry. Meanwhile, the Trusted Foundry program, while secure, cannot economically produce low-volume obsolescent parts. Phoenix Semiconductor’s model bridges that gap by aggregating demand across multiple defense agencies and spreading non-recurring engineering costs over several customers.

Reverse-Engineering at the Transistor Level

What sets Phoenix Semiconductor apart from legacy emulation shops is its “physical-to-EMS” methodology. The company procures original undamaged chips—often from museum collections or decommissioned units—and performs complete delamination, layer-by-layer imaging, and circuit extraction. This process yields a transistor-level netlist that is then resynthesized for a modern, but backward-compatible, fabrication process. For example, a chip originally made on a 150nm node might be re-targeted to a 130nm mixed-signal process while preserving identical pad layout, package type, and timing margins.

Critical to the defense community is the guarantee of no software re-qualification. Many of these Windows-based systems run custom device drivers or real-time control software that was validated against the original chip’s register map and interrupt behavior. Phoenix Semiconductor guarantees cycle-accurate compatibility by simulating the extracted netlist against original test vectors provided by the customer or reverse-engineered from functional units. The company’s founder, Dr. Elena Torres, a former Texas Instruments process engineer, explains: “We don’t just make a chip that works. We make a chip that is indistinguishable from the original at every signal pin.”

Implications for Windows-Centric Defense Systems

While the average consumer might not associate Windows with military hardware, the reality is starkly different. From the Navy’s Aegis combat system—which transitioned to Windows NT-based consoles in the late 1990s—to the Army’s Common Operating Environment, Windows Embedded and its successors underpin the user interface and data processing layers of modern warfare. Even the F-35’s maintenance laptop runs Windows. When the underlying silicon goes extinct, the entire software ecosystem, built over decades, risks obsolescence overnight.

Phoenix Semiconductor’s work directly addresses this. By resurrecting the bridge chips, memory interfaces, and I/O controllers that sit between Windows computing modules and mission-specific hardware, the company allows defense agencies to maintain their current software baselines without costly re-certification. The Phase II contract specifically targets a memory controller for the Windows Embedded-based radar processor mentioned earlier, a component that has been out of production since 2011 and for which no substitute exists.

The SBIR Program as a Force Multiplier

The Small Business Innovation Research program, now in its fifth decade, is a competitive grants-based initiative that encourages domestic small businesses to engage in federal research and development. Phase I awards, typically $150,000, fund feasibility studies; Phase II awards, up to $1.2 million, fund prototyping; and Phase III is the transition to production, often funded by the agency’s procurement budget without further SBIR funds. For the Defense Logistics Agency, which manages the supply chain for all military services, a successful Phase II outcome means a new qualified product listing is added to the Federal Logistics Data system, making the resurrected chip purchasable by any DoD entity.

Phoenix Semiconductor’s award, numbered DLA221C-0041, includes performance milestones every six months. At month six, the company must deliver a functional demonstration of the chip on a customer-provided test board running the actual Windows Embedded image used in the field. At month twelve, accelerated life-test data must prove the chip meets the original military temperature range (-55°C to +125°C). The final month-eighteen deliverable is 1,000 production-ready parts and a complete technical data package enabling second-source manufacturing if necessary.

Broader Impact on Critical Infrastructure

Beyond defense, the same chip-obsolescence problem plagues energy grids, medical devices, and transportation systems—all of which increasingly rely on Windows-based industrial PCs. The North American Electric Reliability Corporation’s latest report flags programmable logic controllers and protective relays running Windows Embedded Standard 2009 as a high-risk area. Phoenix Semiconductor’s technology, once proven with the DLA contract, could be licensed for these applications. The company’s business plan already includes a “Legacy Chip Foundry-as-a-Service” offering, where utility companies submit obsolete part numbers and receive per-unit pricing for resurrected parts within twelve months.

Skepticism and Challenges

Not everyone is convinced that small-scale chip resurrection is economically sustainable. Critics point out that non-recurring engineering costs for a single mixed-signal ASIC can exceed $1 million, meaning per-chip costs remain high unless volumes reach tens of thousands. Moreover, the semiconductor industry’s rapid evolution means that the back-end tools used for legacy nodes are themselves becoming obsolete; finding packaging houses that still support QFP or BGA packages in small lots is increasingly difficult.

Phoenix Semiconductor acknowledges these hurdles but argues that the defense sector’s willingness to pay a premium—often several hundred dollars for a chip that originally cost ten dollars—changes the calculus. Additionally, the company has partnered with a mid-sized foundry in Singapore that maintains mature process lines and offers low-volume production slots, a relationship that it plans to replicate with Ohio-based Intel Foundry Services once that facility’s legacy node capabilities are fully established.

What Comes Next

Assuming successful qualification under Phase II, the Defense Logistics Agency typically awards a Phase III indefinite delivery/indefinite quantity contract, which could be worth up to $50 million over five years. Phoenix Semiconductor has already hired a vice president of defense programs, retired Air Force Colonel Mark Stevens, to navigate the acquisition process. The company expects to double its workforce from 22 to 45 employees by year-end and has signed a lease on a new 15,000-square-foot facility in Austin’s Silicon Hills.

For the end-user soldier, sailor, or airman, the impact is concrete: no more dead-lined vehicles or aircraft waiting six months for a chip that no one makes. The radar that scans the sky for incoming rockets will continue doing so, its Windows Embedded processor humming thanks to a resurrected control chip. As one DLA program manager put it, “This isn’t just about supply chain; it’s about keeping the lights on.”