Aspinity
Aspinity is a Pittsburgh-based semiconductor company developing fully analog neuromorphic processors (AnalogML) that perform machine-learning inference in the analog domain, enabling near-zero power always-on AI for automotive security, smart home, and IoT applications.
- Company typePrivate
- Founded2015
- HeadquartersPittsburgh, United States
- Headcount11–50
- GTM typeB2B
- OfferingHardware or Manufacturing
What Aspinity does
Aspinity, Inc. is a Pittsburgh-based, venture-backed semiconductor company founded in 2015 to commercialize analog machine learning research originating from West Virginia University. The company designs neuromorphic all-analog integrated circuits that perform machine learning inference entirely in the analog domain — before any analog-to-digital conversion — using its proprietary RAMP (Reconfigurable Analog Modular Processor) platform and patented analog non-volatile memory with greater than 10-bit precision. Its two core products are the AML100, launched in February 2022 as the world's first pure analog AI processor consuming under 20µA in always-sensing mode, and the AML200, announced in June 2024 as a next-generation 22nm processor delivering up to 2 TOPS peak performance in a sub-100µW typical power footprint and available as chip, die, or licensable IP.
Aspinity targets three primary end markets: automotive (parked vehicle monitoring, dashcams, security event detection), smart home and home security (glass break, T3/T4 alarm tones, voice activity detection), and broader IoT/industrial IoT (vibration monitoring, predictive maintenance, keyword spotting). Revenue is generated through direct sales of AML100 and AML200 processors and reference modules, evaluation kit sales (EVK1, EVK2, AML100-REF-1, dashcam kit), and an emerging IP licensing channel. Distribution combines direct enterprise engagement with channel partnerships — notably Unitrontech for Korean and global automotive OEMs, and Renesas, STMicroelectronics, and Infineon partner programs for IoT design-in. Pricing is quote-based and not publicly disclosed; the company emphasizes system-level cost benefits from 95%+ reductions in always-on power and 10x battery-life extension rather than chip-level pricing.
The company is led by CEO Richard Hegberg (appointed January 2024) with co-founder Tom Doyle serving as COO and Dr. Brandon Rumberg as CTO; co-founders also include Dr. David Graham (CSO) and Vinod Kulathumani. Aspinity has raised more than $19 million across seed, Series A, and Series B rounds from investors including Anzu Partners, Birchmere Ventures, Amazon Alexa Fund, Mountain State Capital, Riverfront Ventures, and strategic investor Unitrontech, and operates a small organization of 11–50 employees headquartered in Pittsburgh's Strip District.
Aspinity firmographics
Firmographics- Name
- Aspinity
- Legal name
- Aspinity, Inc.
- Website
- https://aspinity.com
- Company type
- Private
- Founded year
- 2015
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- Aspinity is a Pittsburgh-based semiconductor company developing fully analog neuromorphic processors (AnalogML) that perform machine-learning inference in the analog domain, enabling near-zero power always-on AI for automotive security, smart home, and IoT applications.
- Ownership category
- akta.pro rank
Aspinity industry classification
Industry- Product category
- Low-Power Edge AI Semiconductor Processors
- NAICS
- Semiconductor and Related Device Manufacturing (334413), Semiconductor and Other Electronic Component Manufacturing (33441)
- SIC
- Semiconductors & Related Devices (3674)
- akta.pro primary industry
- Compute & Acceleration ICs (CPU/GPU/AI/FPGA/ASIC) (HDAHABAC)
- akta.pro secondary industries
- Custom & Application-Specific ASIC Design Services (HDAHABAJ), On-Device Inference Runtimes & SDKs (mobile/embedded) (HDAAAJAB)
Keywords
Where Aspinity is headquartered
LocationHeadquarters
- HQ city
- Pittsburgh
- HQ country
- United States
- HQ region
- North America
Offices1 record
Markets served
Aspinity business model
Business model- GTM type
- B2B
- Offering type
- Hardware or Manufacturing
- Cost components
- Technology or R&D, Personnel, Supply Chain, Marketing or Sales, Operations
Revenue model
- Analog ML Processor Sales (AML100, AML200): Aspinity generates revenue primarily from the sale of its proprietary analog ML processor chips — the AML100 and the next-generation AML200. These are sold as standalone chips, dies, or IP (licensable IP blocks) to semiconductor companies, automotive manufacturers, and IoT device makers. Volume production of AML100 was planned for Q4 2022; AML200 was sampling in Q1 2025.
- IP Licensing: The AML200 is available as chip, die, and IP, allowing other semiconductor companies to license and integrate Aspinity's analog neural processing technology into their own products.
- Evaluation and Development Kits: Aspinity sells evaluation kits including EVK1 (glass break detection), EVK2 (voice-first with preroll), AML100-REF-1 wireless reference module, and dashcam evaluation kits. These kits facilitate rapid prototyping and customer evaluation, driving chip adoption.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Other | Multi-year contract | AML100 Processor |
| Other | Multi-year contract | AML200 Processor (chip, die, IP) |
| Unit Pricing | Pay-as-you-go | Evaluation Kits (EVK1, EVK2, AML100-REF-1, Dashcam Kit) |
Go-to-market motion2 records
Distribution channels4 records
Marketing channels7 records
Aspinity product offering
Product offeringCore offering
Aspinity designs and sells near-zero-power analog machine learning (AnalogML) semiconductor processors that perform ML inference entirely in the analog domain before digitization. Its flagship AML100 and next-generation AML200 processors, along with reference modules, evaluation kits, and an IoT enablement platform, are sold as chips, dies, or licensable IP to semiconductor companies, automotive OEMs, smart home device makers, and IoT manufacturers.
Product overview
Aspinity offers a portfolio of near-zero power AI processors based on its proprietary AnalogML (analog machine learning) technology. The core product is the AML100 processor, which was the world's first pure analog AI processor enabling always-on system power under 50µA. The AML200 is the next-generation processor delivering TOPS-level performance (up to 2 TOPS peak) in a µW power footprint. Supporting products include the AML100-REF-1 reference module for rapid prototyping and the AML100 IoT Enablement Platform providing complete hardware/software building blocks including multi-sensor modules, gateway, and mobile app for system deployment.
Differentiator
Problem solved
Functional benefit
Products and services
- AML100 World's first pure analog AI processor for near-zero-power always-on detection, classification, notification, and prevention. Consumes less than 20µA when always-sensing and enables always-on system power under 50µA with greater than 20 TOPS/W equivalent compute efficiency.
- AML200 Next-generation edge AI processor delivering TOPS-level performance (up to 2 TOPS peak) in a µW power footprint. Designed at 22nm, supports up to 125k parameters, and is the only edge AI processor to deliver TOPS-level performance in a battery-operated power footprint.
- AML100-REF-1 Fully integrated reference module for prototyping battery-operated always-on sensing solutions. Incorporates the AML100 processor, STMicroelectronics STM32WB5MMG 64MHz Arm Cortex-M4 MCU with 2.4GHz wireless, and TDK ICS-40310 analog MEMS microphone; system consumes less than 40µA in always-on mode for acoustic event detection.
- AML100 IoT Enablement Platform Complete hardware/software solution for quick deployment, testing, and development of AML100-based always-on monitoring solutions for vehicle, smart home, and security applications. Includes up to 8 multi-sensor modules, a gateway module with BTLE/WiFi/LTE, and a mobile app.
- EVK1 Acoustic Event Detection Kit Evaluation kit featuring the analogML core and Infineon XENSIV IM73A135 analog MEMS microphone, targeting smart home devices for acoustic triggers such as glass break, voice, or other acoustic events.
- EVK2 Voice-First Evaluation Kit Voice-first evaluation kit featuring the RAMP chip for ultra-low-power analog voice activity detection and preroll collection. Demonstrates an end-to-end voice wake-up solution integrated with STMicroelectronics STM32H743ZI MCU.
Quantifiable outcome
- Always-on system power reduced by 95%+ vs. digital approaches, enabling 10x longer battery life
- +5 more outcomes
Companies that use Aspinity
Customer profileSegments3 records
Ideal customer profiles4 records
Aspinity technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- No
- API docs
- API detail
Core technology
AI maturity
App detail
Integration1 record
AI capability6 records
Feature7 records
Aspinity partnerships and signals
Strategic signalPartnerships
Four partnerships are on record, tiered core and major.
- UnitrontechcoreStrategic partnership with Unitrontech, a Korean automotive semiconductor distributor and key partner of leading automotive manufacturers. The partnership was established alongside Aspinity's $5M Series B funding round. Unitrontech serves as a distributor of Aspinity's technology to automotive customers globally, accelerating time-to-market for AI-based automotive impact detection and other AIoT solutions such as always-on dashcams with sensor fusion and AI to detect and record only relevant events.
- Renesas ElectronicscoreAspinity joined the Renesas Preferred Partners Program and released the AB2 AML100 Application Board integrating with Renesas Quick-Connect IoT platform (Arduino-compatible). This enables rapid prototyping of the AML100 with Renesas MCUs. The AB2 board was demonstrated at Embedded World 2023 and allows system designers to implement the AML100 as a front-end analog event detector that keeps Renesas MCUs in their lowest power state unless relevant events are detected.
- STMicroelectronicscoreAspinity joined the ST Partner Program to integrate its ultra-low-power analogML core with STMicroelectronics' high-performance MCUs (STM32 series). The collaboration enables customers to rapidly adopt analogML for always-on sensing applications. The Voice-First Evaluation Kit (EVK2) integrates the AML/RAMP chip with the STM32H743ZI MCU and demonstrates ultra-low-power analog voice activity detection with preroll collection for accurate keyword analysis.
- Infineon TechnologiesmajorPartnership with Infineon Technologies AG to combine Aspinity's RAMP technology with Infineon's XENSIV family of sensors (including MEMS microphones) for battery-operated always-on sensing products. The collaboration enables devices that continuously monitor their environment without compromising battery life. Infineon's high-performance XENSIV sensors combined with Aspinity's RAMP chip promote high accuracy with power efficiency across consumer, IoT, and industrial applications.
Scale indicators9 records
Recent moves6 records
Expansion highlights6 records
Aspinity competitors and assessment
Company assessmentDirect peers
- Syntiant: Syntiant makes ultra-low-power neural decision processors for always-on voice and sensor edge AI. It is the closest direct competitor to Aspinity, serving similar use cases (keyword spotting, audio event detection) with an ultra-low-power digital neural accelerator, often contrasted head-to-head against Aspinity's analog ML approach.
- Ambiq Micro: Ambiq designs sub-threshold power-optimized microcontrollers and SoCs that enable always-on AI in battery-operated endpoints. Like Aspinity, Ambiq targets wearable, hearable, and IoT customers who need near-zero-power operation, making it a primary comparable in edge AI power efficiency benchmarks.
- BrainChip (Akida): BrainChip offers the Akida neuromorphic processor IP and chips for edge AI. It is a direct peer in the neuromorphic/event-driven AI space and competes for the same always-on, low-power edge inference workloads across IoT and industrial applications.
Emerging players
- Eta Compute: Eta Compute develops low-power edge AI processors and SoCs for IoT sensor applications. It competes in the same always-on edge AI niche as Aspinity with a digital architecture focused on ultra-low-power inference.
- Hailo: Hailo produces edge AI processors focused on higher-throughput computer vision inference at the edge. While it targets higher TOPS workloads than Aspinity, it is a relevant emerging peer in the broader edge AI accelerator market.
- Mythic: Mythic developed analog compute-in-memory AI accelerators for edge inference. While focused on higher-power vision workloads, Mythic represents the most analogous analog-in-memory competitor to Aspinity's fully-analog approach.
Broad incumbents
- Synopsys (ARC Edge AI IP): Synopsys offers licensable ARC processor IP including ARC NPX neural network processors for edge AI. As a broad EDA and IP incumbent, Synopsys competes for analog/neuromorphic-style design-in wins with established digital IP, and competes more broadly in the IP-licensing business Aspinity is pursuing.
- Cadence Design Systems (Tensilica): Cadence offers Tensilica DSP and HiFi neural network processor IP for edge AI applications. As a broad EDA/IP incumbent, Cadence is a competitive threat for IP licensing deals and serves overlapping audio and always-on AI markets.
- Himax Technologies: Himax designs mixed-signal and display driver ICs, including ultra-low-power AI imaging sensors with on-chip AI inference for always-on vision. Its WiseEye product line competes for always-on edge AI sensor applications in IoT and automotive markets.
- Qualcomm (Edge AI / Always-On Voice): Qualcomm's low-power audio SoCs and AI accelerators power many always-listening smart speakers and hearables. As a broad semiconductor incumbent with deep voice/AI software stacks, it competes for the same design wins Aspinity targets, especially in Alexa-enabled and voice-first devices.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat6 records
Key risks6 records
Key highlights7 records
Customer concentration
Aspinity social profiles
Digital presenceAspinity financial estimates
Financial estimateRevenue estimate
Valuation estimate
Aspinity leadership team
Management profileNumber of profiles
Profiles6 records
Aspinity funding detail
Funding detailFunding overview
Funding rounds8 records
Investors12 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Aspinity M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Aspinity
What does Aspinity do?
Aspinity designs and sells near-zero-power analog machine learning (AnalogML) semiconductor processors that perform ML inference entirely in the analog domain before digitization. Its flagship AML100 and next-generation AML200 processors, along with reference modules, evaluation kits, and an IoT enablement platform, are sold as chips, dies, or licensable IP to semiconductor companies, automotive OEMs, smart home device makers, and IoT manufacturers.
Is Aspinity a public or private company?
Aspinity is a private company. It is classified as venture growth investor backed and is currently operating.
When was Aspinity founded?
Aspinity was founded in 2015. It employs 11 to 50 people.
Where is Aspinity based?
Aspinity is headquartered in Pittsburgh, United States, in the North America region.
How does Aspinity make money?
Three revenue lines are on record. Analog ML Processor Sales (AML100, AML200) is the primary driver. The others are IP Licensing and evaluation and Development Kits.
Who are Aspinity's main competitors?
Direct peers on record are Syntiant, Ambiq Micro and BrainChip (Akida). Emerging players are Eta Compute, Hailo and Mythic. Broad incumbents are Synopsys (ARC Edge AI IP), Cadence Design Systems (Tensilica), Himax Technologies and Qualcomm (Edge AI / Always-On Voice).
Does Aspinity have an API?
No public API is recorded for Aspinity.
What industry is Aspinity in?
Aspinity's product category is Low-Power Edge AI Semiconductor Processors. Its primary akta.pro industry code is HDAHABAC, Compute & Acceleration ICs (CPU/GPU/AI/FPGA/ASIC), with a secondary code of HDAHABAJ, Custom & Application-Specific ASIC Design Services. Its NAICS code is 334413 and its SIC code is 3674.