Mentium Technologies

Name

Mentium Technologies

Legal name

Mentium Technologies Inc.

Website

https://mentium.tech

Company type

Private

Founded year

2018

Operating status

Operating

Headcount range

1–10 employees

Ownership category

akta.pro rank

Product category

Edge AI Semiconductors

NAICS

Semiconductor and Related Device Manufacturing (334413)

SIC

Electronic Computers (3571)

akta.pro primary industry

Compute & Acceleration ICs (CPU/GPU/AI/FPGA/ASIC) (HDAHABAC)

akta.pro secondary industries

AI/Neural Processing Units (NPUs) & ML Accelerators (HDAHAJAF), On-Device Inference Runtimes & SDKs (mobile/embedded) (HDAAAJAB), AI Compiler, Runtime & Kernel Optimization Software (CUDA/ROCm/XLA, graph compilers) (HDAAAAAI)

Keywords

• Edge AI accelerators
• Radiation-hardened semiconductors
• Neuromorphic computing chips
• Space-grade AI processors
• Compute-in-memory architecture

Headquarters

HQ city

Santa Barbara

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Hardware or Manufacturing

Cost components

Technology or R&D, Personnel, Supply Chain, Operations, Marketing or Sales, Infrastructure

Revenue model

1. AI Co-processor Hardware Sales: Sale of radiation-tolerant AI co-processors (Luna family) to government, defense, and commercial customers for space, IoT, robotics, drones, and security applications.
2. SDK and Software Tools: Mentium SDK sales for model compilation, quantization, profiling, and deployment on Mentium co-processors.
3. NASA SBIR Contracts: Revenue from NASA Small Business Innovation Research (SBIR) Phase II Sequential awards and Civilian Commercialization Readiness Pilot Program (CCRPP) contracts for space AI technology development.

Pricing tiers

Go-to-market motion1 record

Distribution channels1 record

Marketing channels4 records

Core offering

Mentium Technologies designs and sells ultra-low-power, radiation-tolerant AI co-processors based on proprietary neuromorphic compute-in-memory architectures. Its Luna family of co-processors delivers AI inference below 0.5W for size, weight, and power (SWaP)-constrained environments across space, defense, drones, robotics, and commercial IoT applications. The product line is supported by the Mentium SDK toolchain for model compilation, quantization, and deployment.

Product overview

Mentium Technologies offers a unified family of ultra-low-power AI co-processors called Luna, designed for SWaP-constrained environments. The portfolio spans commercial IoT (Luna-1, Luna-2) and radiation-tolerant space-grade variants (Luna-R1, Luna-R2). All co-processors are supported by the Mentium SDK, a comprehensive toolchain enabling model compilation, quantization, profiling, and deployment with a high-fidelity emulator. The family delivers AI inference below 0.5W, with Luna-2 representing the next-generation 4.2 TOPS accelerator. The Luna Development Kit provides the development and evaluation platform for rapid prototyping.

Differentiator

Problem solved

Functional benefit

Products and services

• Luna-1 Extreme-low power AI co-processor for IoT applications delivering 2.4 TOPS at 70mW typical power, featuring xSPI and GPIO interfaces, designed for SWaP-constrained edge devices.
• Luna-R1 Radiation-tolerant extreme-low power AI co-processor for space applications, delivering 2.4 TOPS at 75mW typical power with RHBD features and up to 200kRad TID tolerance.
• Luna-2 Next-generation ultra-low power AI accelerator for IoT and edge applications, delivering 4.2 TOPS at 130mW typical power with xSPI and PCIe 3.0x1 interfaces.
• Luna-R2

Quantifiable outcome

• Energy efficiency improvement: 2-3 orders of magnitude better than CMOS through compute-in-memory architecture
• +2 more outcomes

Named customers3 records

Segments5 records

Ideal customer profiles4 records

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

AI capability7 records

Feature4 records

Partnerships

Six partnerships are on record, tiered core and minor.

• EarthTraqcoreChannel Partner/ Reseller/ Distributor · 3 June 2026EarthTraq selected Luna-R1 AI chips for deployment on ET-01 constellation mission (4 satellites in LEO). Mission sponsored by NASA's Small Spacecraft and Distributed Systems (SSDS) project. Luna-R1 enables distributed AI inference across constellation for autonomous intelligence in LEO satellite networks.
• UFO companycoreChannel Partner/ Reseller/ Distributor · 12 March 2026UFO company selected Luna-1 AI chip for deployment on their 6U Cubesat Odyssey orbital mission launching in 2026. Luna-1 will gather real-time sensor data from orbit and perform onboard AI analytics. The CubeSat is hosted by Xenti's FANTM-RiDE dispenser.
• NASAcoreStrategic or Co-development Partner · 24 February 2026NASA awarded Mentium a Civilian Commercialization Readiness Pilot Program (CCRPP) Small Business Innovation Research (SBIR) contract, the final stage of the NASA SBIR program. CCRPP requires recipients to secure outside investment that NASA matches with SBIR/STTR funds ($500,000 to $2.5 million). NASA previously awarded SBIR Phase II Sequential award 80NSSC22CA230 and funded flight testing under 80NSSC23CA207. NASA JPL provided technical consultation for radiation testing at Berkeley Accelerator Space Effects Facility.
• SynopsyscoreTechnology or Integration · 17 September 2024Mentium Technologies leverages Synopsys Cloud's EDA-as-a-Service platform to tape out designs faster in a cost- and technology-prohibitive semiconductor industry. The platform helps AI chip startups accelerate chip design lifecycles with 40% faster time-to-results on average, reducing weeks of configuration to days with per-minute licensing automation and SOC 2 compliance.
• NASA Flight Opportunities / Worldview / SPEED TeamcoreStrategic or Co-development Partner · 20 September 2023Mentium developing test AI system for suborbital flight on High-Altitude Balloon in collaboration with Worldview and NASA's SPEED team from Ames Research Center. This represents first step toward developing 1U CubeSat AI system for LEO and Lunar orbit testing.
• Silicon CatalystminorStrategic or Co-development Partner · 4 December 2022Mentium presented at the Silicon Catalyst Portfolio Companies' Update event (2022 Fall Meeting), attended by over 150 investors and semiconductor industry veterans. Silicon Catalyst is a semiconductor-focused startup accelerator and investor.

Scale indicators4 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Mythic AI: Mythic develops analog compute-in-memory AI inference accelerators using flash memory transistors, directly analogous to Mentium's floating-gate CIM architecture. Both target ultra-low-power edge AI inference with the same fundamental device-physics approach.
• BrainChip: BrainChip's Akida neuromorphic processor IP targets ultra-low-power edge AI inference for IoT, security, and aerospace applications. Like Mentium, BrainChip focuses on event-driven neural network inference at the edge under severe SWaP constraints.
• Syntiant: Syntiant produces ultra-low-power neural decision processors (NDPs) for always-on edge AI in IoT, audio, and sensor applications. Mentium's Luna-1 and Luna-2 compete head-to-head with Syntiant's NDP120/NDP120-S in the power-constrained edge AI category.
• Hailo: Hailo develops edge AI inference processors (Hailo-8, Hailo-15) optimized for low-power computer vision and multimodal AI at the edge. It is a primary competitor in the broader edge AI accelerator market where Mentium's Luna family also competes.
• GrAI Matter Labs: GrAI Matter Labs produces neuromorphic-inspired edge AI processors (GrAI VIP) for ultra-low-latency, low-power inference in vision and sensor applications. Its biologically-inspired architecture mirrors Mentium's neuromorphic approach for power-constrained edge inference.
• GreenWaves Technologies: GreenWaves' GAP9 processors target ultra-low-power AI inference for IoT, hearables, and sensor-based edge devices. Both companies compete in the sub-100mW edge AI processor category with comparable SWaP-constrained design philosophies.

Emerging players

• Anaflash: Anaflash develops compute-in-memory AI chips based on embedded flash memory technology, sharing Mentium's floating-gate CIM foundation. It is an emerging competitor pursuing similar analog in-memory inference advantages for edge applications.
• PolyN Technology: PolyN Technology builds ultra-low-power analog in-memory compute AI chips for always-on edge AI in hearables, wearables, and IoT sensors. It pursues a similar compute-in-memory value proposition to Mentium's CIM architecture.

Broad incumbents

• Microchip Technology: Microchip supplies radiation-hardened and radiation-tolerant microcontrollers, FPGAs (via Microsemi), and DSPs widely used in space and defense applications. It is the dominant incumbent supplier of space-grade processors that Mentium's Luna-R1/R2 must displace.
• Xilinx (AMD): Xilinx (now part of AMD) supplies space-grade FPGAs and adaptive SoCs (e.g., Kintex UltraScale, Versal AI Edge) that are widely deployed in satellites and defense systems. It represents the incumbent programmable-logic alternative for in-orbit AI workloads.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks6 records

Key highlights6 records

Customer concentration

• LinkedInlinkedin.com/company/mentium-technologies-inc

Compliance2 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles11 records

Funding overview

Funding rounds5 records

Investors8 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Mentium Technologies