SUSTAINABLE AI & CRYPTO MINING HARDWARE
DESIGNED IN JAPAN

Lenzo builds ultra-efficient hardware designed for modern parallel workloads — from blockchain to AI inference. Our proprietary CGLA architecture delivers breakthrough performance per watt in a compact, cost-effective form.

INTRODUCING LENZO CORE, BUILT FOR REAL WORKLOADS

Utilizing Proprietary CGLA (Coarse-Grained Linear Array) architecture.
Parallel compute units scale across cascades and memory lanes.
Execution, memory, and data flow are coordinated directly in hardware.

This is visible, predictable performance.

This visualization reveals the internal state of LENZO Core, LENZO’s CGLA (Coarse-Grained Linear Array) compute architecture. Each ring represents a scalable CGLA fabric where parallel processing units operate across cascades and memory lanes. Real-time, color-coded states expose execution flow, cache activity, DMA transfers, and external memory interaction. Hardware-level multi-threading and buffering eliminate pipeline stalls, enabling predictable high throughput. LENZO Core delivers scalable performance, energy efficiency, and simplified programming for real-world workloads. Full length demo video available here.

PRODUCTS & SERVICES

CGLA-based compute systems, platforms, and integration services

Lenzo M-Series: Crypto Mining Rig

Mine smarter and run cooler—without performance tradeoffs.

M-Series is built for next-generation mining infrastructure, optimized for SHA256 efficiency and high hashrate-per-watt delivery.

Runs custom ASICs – enabling long-term reliability and simplified 24/7 operation at scale.

Lenzo A-Series: AI Server Platform

Train smarter and run faster—without GPU lock-in.

A-Series is built for next-generation AI infrastructure, optimized for LLM training and inference with ultra-low power draw.

Runs open-source models – enabling flexible deployment and full ownership of your stack.

Technical Consulting Services

Our team combines deep expertise in high-efficiency computing and custom silicon design.

We assist with architecture, tuning, and rollout across edge, AI, and blockchain systems.

Whether you’re designing hardware or scaling compute, Lenzo supports every step from chip to deployment.

KEY BENEFITS

Advantages That Drive Your Bottom Line

High Hash Rate

Outperforms standard ASIC rigs in energy efficiency (J/TH)

Cost-Efficient Hardware

Designed to reduce total cost of ownership over time

Low Power Consumption

Maximizes performance per watt using a novel dataflow architecture

Broad Algorithm Support

Supports SHA256 out of the box; other algorithms can be added upon request

ABOUT

The team behind the technology

Crafted in Japan, Lenzo is built by the engineers behind some of the world’s fastest chips—from Sony’s PlayStation CPU/GPU teams to the designers of Fujitsu’s supercomputers. Our founding team includes researchers from NAIST and ITRI Taiwan, with deep expertise in large-scale compute infrastructure. We are system architects and builders, committed to engineering high-performance, energy-efficient hardware for a new generation of workloads.

BACKED BY INVESTORS

Supported by a group of venture capital and strategic investors

CAREERS

Join Our Growing Team

System on Chip (SOC) Engineer

  • Responsibilities: Register-transfer level design, SoC integration, verification, and performance optimization.
  • Requirements: Experience in SoC/ASIC/FPGA design, EDA tools (Synopsys/Cadence), and interest in AI or cryptographic accelerators.
  • Location: Yokohama, Kyoto, Remote

C/C++ Programmer (Blockchain / Mining)

  • Responsibilities: C/C++ programming, system integration, open-source miner contributions (e.g., CGMiner, Bitaxe/AxeOS, Tether MOS), and hardware communication.
  • Requirements: Strong C/C++ skills, blockchain protocol knowledge, and familiarity with CGMiner, Bitaxe/AxeOS, Tether MOS or similar open-source projects.
  • Location: Yokohama, Kyoto, Remote

RESOURCES & PRESS

White Papers, Research, Media Coverage, and Press Releases
5 Jun

Thank You for Visiting LENZO at COMPUTEX 2026

COMPUTEX 2026 has come to a close, and we would like to extend our sincere thanks to everyone who visited the LENZO booth during the event.

Over the course of the week, we had the opportunity to meet with industry leaders, technology partners, investors, researchers, and innovators from around the world. We are grateful for the many insightful conversations and the strong interest shown in LENZO’s approach to next-generation compute architecture.

As part of the Garage+ Pavilion at InnoVEX, we were pleased to introduce LENZO and share our vision for more power-efficient, scalable computing through our proprietary TokenProcessor™ architecture.

The feedback, questions, and discussions we received throughout the event were invaluable and will help shape the next stage of our development.

We would also like to thank the teams at Garage+, InnoVEX, and COMPUTEX for creating an excellent environment for innovation, collaboration, and meaningful connections.

For those we met during the exhibition, we look forward to continuing the conversation. If you would like to learn more about LENZO or explore potential collaboration opportunities, please feel free to contact our team.

Thank you once again for your time, support, and interest in LENZO.

We look forward to seeing you again soon.

28 May

LENZO A-Series product materials are now available

Built on LENZO’s proprietary TokenProcessor™ architecture, the A-Series is designed for low-power, reconfigurable AI inference across edge, embedded, and distributed Physical AI deployments.

The newly released flyer provides an overview of:
• Reconfigurable dataflow architecture
• Up to 65 TOPS (INT8) AI performance
• LPDDR4 memory scalability
• Edge-to-server deployment flexibility and use cases
• Linux-based software environment

Materials available upon request.

7 May

LENZO to Exhibit at COMPUTEX Taipei 2026

LENZO is pleased to announce that we will exhibit at COMPUTEX Taipei 2026 / InnoVEX 2026 this June as part of the Garage+ Startup Global Program (S26 Batch).

This marks LENZO’s first international trade show appearance and an important milestone as we expand our engagement with the global semiconductor and AI infrastructure ecosystem.

Hosted in Taipei, COMPUTEX is one of the world’s leading technology exhibitions, bringing together semiconductor companies, AI infrastructure providers, system manufacturers, investors, and media from across the industry. Through the Garage+ program, LENZO will participate in a tailored business development track while exhibiting inside the Garage+ area at InnoVEX.

At the event, LENZO will introduce its CGLA (CPU Grounded Linear Array) architecture — a next-generation compute architecture designed to improve performance-per-watt through reduced memory movement and highly efficient parallel processing.

As demand for AI infrastructure continues to grow, power efficiency and scalable compute architectures are becoming increasingly important across both datacenter and edge environments. LENZO is developing specialized compute technologies aimed at addressing these challenges through a more energy-efficient approach to acceleration.

COMPUTEX 2026 also represents an opportunity for LENZO to further strengthen relationships across the semiconductor ecosystem in Taiwan and internationally, while sharing progress on the company’s roadmap and ongoing development efforts.

We look forward to meeting partners, investors, media, and industry participants during the exhibition.

LENZO will be exhibiting within the Garage+ pavilion at InnoVEX 2026. A venue floor map is included below for reference.

If you are attending COMPUTEX and would like to schedule a meeting with LENZO during the event, please contact us through our dedicated form.

Exhibition Information

Event: COMPUTEX Taipei / InnoVEX 2026
Dates: June 2–5, 2026
Venue: Taipei Nangang Exhibition Center Hall 2, 4F
Exhibitor Area: Garage+ Pavilion

30 Apr

LENZO Selected for Garage+ Startup Global Program (S26 Batch)

LENZO has been selected as part of the S26 Batch of the Startup Global Program, organized by Garage+ for market entry into Taiwan.

The program brings together 31 startups selected from over 330 applications across 51 countries, spanning fields such as AI, semiconductors, advanced materials, robotics, and clean energy. LENZO joins a global cohort of companies building next-generation technologies with strong international ambition.

As part of the program, LENZO will participate in a tailored business development track starting in June, with opportunities to engage directly with investors and strategic partners. The program also includes participation in COMPUTEX 2026 and InnoVEX 2026, where LENZO will showcase its architecture and ongoing developments. LENZO will be exhibiting at S0304 "Garage+".

This selection reflects growing recognition of LENZO’s CPU Grounded Linear Array (CGLA) architecture -designed to deliver high-performance, energy-efficient compute for emerging AI workloads.

We look forward to connecting with partners across the ecosystem and sharing more in the months ahead.

List of Publications

View the full list of publications of the NAIST Computing Architecture Lab, including peer-reviewed, invited-talk & awards

INVESTORS

What Our Investors Say

Ready to power the next era of intelligent infrastructure?

Let’s talk. Whether you’re a miner, builder, or investor—we’re here to deliver.

    Subscribe to Newsletter