September 29, 2026 — In a strategic move to address the surging computational demands of modern edge AI and industrial automation, Efinix® Inc., a pioneer in high-performance FPGA architecture, today announced the launch of the Sapphire™ RV64 SoC. This new 64-bit RISC-V processor core represents a significant architectural leap for the company, designed to provide embedded systems engineers with a scalable, high-performance platform that maintains the low-power efficiency characteristic of FPGA-based designs.
The introduction of the RV64 is not merely an incremental update; it is a fundamental expansion of the Efinix Sapphire suite. By bridging the gap between resource-constrained 32-bit microcontrollers and high-performance 64-bit application processors, Efinix is positioning itself at the center of the next wave of intelligent, edge-deployed hardware.
Main Facts: Breaking the 32-bit Barrier
The Sapphire RV64 is a 64-bit RISC-V processor core engineered to overcome the bottlenecks frequently encountered by developers working on complex embedded systems. As applications in machine vision, automotive, and industrial robotics grow in sophistication, the limitations of 32-bit architectures—specifically regarding addressable memory, cache depth, and I/O management—have become a primary constraint on innovation.
Key Technical Enhancements:
- 64-bit Architecture: Facilitates the handling of larger data sets and more complex software stacks, including full 64-bit embedded Linux distributions.
- Expanded Memory Support: Enables developers to move beyond the memory constraints of smaller cores, allowing for multi-gigabyte memory footprints essential for modern AI models.
- Advanced Debugging: Enhanced debug capabilities provide better visibility into code execution, significantly shortening the development and troubleshooting lifecycle.
- FPGA-Native Scalability: Built upon the same user-configurable philosophy as the existing RV32, the RV64 ensures that current Efinix customers can migrate their designs with minimal re-architecting.
- Custom Instruction Capability: Retains the hallmark of the Efinix ecosystem, allowing users to implement custom hardware accelerators directly in the FPGA fabric, effectively creating a software-defined, hardware-accelerated pipeline.
The RV64 is immediately available for deployment across Efinix’s comprehensive device portfolio, including all Titanium and Topaz FPGAs, as well as Trion devices (T20 and above), ensuring that customers have a path from small-scale microcontrollers to high-performance compute nodes.
Chronology: The Evolution of the Sapphire Suite
The journey to the Sapphire RV64 began with Efinix’s initial foray into RISC-V integration, which sought to simplify the design process for FPGA-based systems.
- Early Phase: Efinix established its footprint by introducing the 32-bit Sapphire RV32. The goal was to provide a "soft-core" processor that could be easily integrated into an FPGA fabric, allowing designers to offload specific tasks from a standard CPU to custom hardware logic.
- The "Software-Defined" Shift: As the market trend shifted toward edge AI, Efinix focused on refining the Efinity® Integrated Development Environment (IDE). By integrating the RISC-V Embedded Software IDE, the company provided a unified workflow—from RTL to bitstream—that removed the friction traditionally associated with hardware-software co-design.
- Market Pressure: Over the last 24 months, feedback from key clients in industrial and automotive sectors indicated that while the RV32 was ideal for control-plane tasks, the data-plane requirements of edge AI—such as neural network inferencing and high-speed data processing—required the addressable space and performance of a 64-bit architecture.
- Current Milestone: The September 2026 launch of the RV64 fulfills this market requirement, completing the transition from a niche FPGA-based controller to a robust platform capable of handling the heavy lifting of modern embedded Linux workloads.
Supporting Data: Why 64-bit Matters for Edge AI
The transition to 64-bit computing in embedded environments is driven by the sheer scale of modern AI model parameters. A 32-bit system is inherently limited to addressing 4GB of memory. While this was sufficient for legacy embedded systems, modern computer vision models and large-language model (LLM) edge variants often require significantly higher memory throughput and larger addressing spaces for efficient inference.
Performance Indicators:
- Memory Addressing: The 64-bit architecture removes the 4GB wall, allowing for real-time processing of high-resolution sensor data that would otherwise require multiple context-switching operations on 32-bit systems.
- Instruction Efficiency: By utilizing the RISC-V ISA, the RV64 achieves a superior instructions-per-clock (IPC) count compared to proprietary 32-bit cores, particularly when handling complex branch-heavy algorithms common in AI logic.
- Low Power Footprint: By maintaining the FPGA-native approach, Efinix ensures that power consumption scales linearly with workload. Unlike general-purpose CPUs that remain power-hungry, the Sapphire RV64 allows for the deactivation of unused logic gates, keeping the power profile tightly coupled to the application’s specific requirements.
Official Responses: Insights from Efinix Leadership
The launch has been met with enthusiasm from Efinix’s executive team, who view the RV64 as a bridge to a more intelligent, autonomous future.
Tony Ngai, Co-Founder, Co-President, and CTO of Efinix, emphasized the importance of seamless migration for existing clients:
"Sapphire RV64 was designed so customers already using our 32-bit Sapphire SoCs can scale up without re-architecting their systems. It carries forward the same configurable, FPGA-native approach, but now with larger addressable memory, caching, and debug capability that 64-bit embedded Linux and AI workloads demand."
Mark Oliver, Vice President of Marketing at Efinix, highlighted the "customer-first" development approach:
"Customers building increasingly sophisticated embedded and edge AI systems have been asking us for a path to 64-bit processing without giving up the low power and configurability that made Sapphire successful. Sapphire RV64 gives designers that headroom, with larger memory spaces, deeper caches, and more capable debug, while staying true to the user-configurable philosophy of the entire Sapphire SoC suite."
Implications: The Future of Edge Computing
The introduction of the Sapphire RV64 has several far-reaching implications for the industry at large.
1. The Democratization of Edge AI
By providing a 64-bit RISC-V platform that integrates easily with FPGA fabrics, Efinix is lowering the barrier to entry for companies wanting to develop proprietary, AI-accelerated hardware. Developers no longer need to choose between a power-efficient but limited microcontroller and a high-performance but power-inefficient ASIC.
2. Streamlined Development Ecosystems
Efinix’s commitment to the Efinity® IP Manager and the Eclipse-based IDE signals a shift toward "Platform-as-a-Service" for hardware designers. By consolidating the toolchain, the company reduces the time-to-market for complex systems. A designer can now move from a basic RTL design to a fully functional, 64-bit Linux-based AI platform using the same software environment.
3. Strengthening the RISC-V Ecosystem
Efinix’s continued investment in RISC-V underscores the architecture’s dominance in the embedded space. As a non-proprietary ISA, RISC-V allows Efinix to innovate without the licensing constraints associated with legacy architectures. This flexibility is critical for automotive and industrial clients who demand long-term supply chain security and the ability to customize their instruction sets for specialized tasks.
4. Expansion into High-Performance Verticals
The RV64 allows Efinix to compete more aggressively in sectors that were previously out of reach, such as high-end autonomous drones, advanced robotics, and intelligent medical imaging. These fields require the deterministic response times of an FPGA combined with the compute power of a 64-bit processor—a combination that the Sapphire RV64 is uniquely positioned to deliver.
Conclusion
As the industry moves toward a future where "intelligence at the edge" is the standard rather than the exception, the demand for hardware that is both performant and flexible will only increase. The Sapphire RV64 SoC serves as a critical milestone in this trajectory, offering a mature, scalable, and highly capable platform for the next generation of embedded designers. By building upon a foundation of RISC-V and FPGA agility, Efinix has successfully provided the tools necessary to turn complex AI concepts into efficient, real-world deployments.
