Casting Replacement Becomes a Defense Procurement Goal
A Defense Innovation Unit contract targeting airborne casting replacement. An 11-fold simulation speedup that compresses months of CFD work into seven minutes. Two signals from the same week in March 2026 converge on a single conclusion: the 'Digital Foundry'—the systemic use of metal AM to replace legacy castings at production scale—has moved from industry aspiration to funded government program. Driven by an Other Transaction Agreement awarded to Nikon AM Synergy under the Foundry for Operational Readiness and Global Effects (FORGE) program, and supported by Sintavia's record simulation throughput using NVIDIA's Blackwell architecture, the primary bottleneck in the aerospace supply chain—procurement lead times for flight-critical cast components that routinely stretch beyond a year for lower-volume parts—is now an explicit policy problem with funded solutions. Parallel moves in India and from Renishaw's quality automation work signal that the infrastructure for a decentralized, high-velocity defense industrial base is taking shape.
DIU Awards Nikon the FORGE Contract for Serial Casting Replacement
In March 2026, Nikon AM Synergy secured an Other Transaction Agreement (OTA) contract from the U.S. Defense Innovation Unit (DIU) under the Foundry for Operational Readiness and Global Effects (FORGE) program. The contract targets serial production of metal components for airborne systems, using Laser Powder Bed Fusion (LPBF) to replace traditionally cast parts. Production is centered at Nikon's Long Beach, California facility using Nikon SLM Solutions hardware. The award carries a one-year base period with a six-month option; the financial value was not disclosed (DIU / Company PR).
The FORGE program is specifically designed to address manufacturing bottlenecks for metal parts in high-performance aeronautical systems—parts that have historically been produced by casting and are subject to long procurement queues when volume demand is low or tooling is obsolete. By awarding this contract, the DoD is formally signaling that LPBF has cleared the threshold for flight-critical, high-stress applications at the program level, not just the component level.

Prior Art & Evolution: This builds on flight-certification work pioneered by GE Aviation's fuel nozzle (circa 2015) and F-35 engine brackets (2019–2022). What distinguishes 2026 is the shift from part-level optimization—where a single component was redesigned for AM—to category-level replacement. FORGE does not seek better parts; it seeks to eliminate the dependency on fixed-tooling foundries entirely. Larger build volumes and multi-laser systems now provide the throughput needed to compete with the unit economics of mid-volume casting runs.
Sintavia Cuts CFD Cycle Time From Months to Two Weeks
The certification bottleneck for casting replacement has historically been simulation. Complex geometries—internal cooling channels for hypersonic components, conformal heat exchangers—require massive computational resources to validate thermal and mechanical performance before a part can be printed, let alone certified. Sintavia's integration of the NVIDIA RTX PRO 6000 Blackwell workstation GPU directly addresses this.
Sintavia reported an 11-fold increase in simulation speed versus a 24-core CPU baseline: a 30-million-cell conjugate heat transfer CFD model (run in Siemens Simcenter STAR-CCM+) that previously took 88 minutes on CPU now completes in 7 minutes on a single Blackwell GPU. The design-to-simulation cycle for aerospace heat exchangers has compressed from months to two weeks. The resulting heat exchanger design demonstrated a 30% reduction in weight and 20% improvement in thermal efficiency, validated by CT scanning and in-house testing (Sintavia / BusinessWire, March 2026).

The strategic implication extends beyond Sintavia's own production: by removing the CAD-to-print bottleneck, this class of GPU acceleration changes the economics of the certification loop across the industry. As previously noted in our analysis of Physical AI (March 14), the competitive advantage in AM is migrating from the hardware's ability to print to the software's ability to reason through physics-driven constraints in near-real-time.
Global Signals: Localization and Quality Automation
The Digital Foundry model is not limited to US defense. On March 20, Phillips Machine Tools India partnered with 2OneLab to localize LPBF hardware production in Pune, India—mirroring the DIU's goal of supply chain sovereignty for the aerospace and healthcare sectors.
Renishaw's R&D investment into neural network-based in-process quality control (announced March 18) addresses a secondary barrier: process repeatability at production volumes. By monitoring build integrity at the nanometer scale in real-time, Renishaw aims to reduce post-process inspection overhead—currently a significant cost center in certified AM production (Analyst Estimate). Combined with Scantech's AM-CELL C automated metrology system, these developments collectively point toward an automated production line requiring minimal human intervention from design to final inspection.
The Certification Gap Remains the Critical Path
The technical and computational barriers to casting replacement are demonstrably falling. The regulatory barrier is not. The transition to a Digital Foundry assumes the FAA and DoD will eventually accept simulation-as-qualification for flight-critical parts. Today, physical coupon testing remains the gold standard, adding months to production timelines regardless of print speed.
What to Watch: The primary risk is latent defects in large-format LPBF builds that escape AI-driven in-situ monitoring. A high-profile failure in a flight-certified part would create significant headwinds for casting replacement programs broadly. Industry stakeholders should track the DIU's initial FORGE qualification reports expected in Q4 2026, along with upcoming hot-fire tests of the LEAP 71/HBD 200kN aerospike. Success at these milestones will determine whether the Digital Foundry becomes a new global standard or remains a specialized solution for defense-urgent applications.
