ANAFLASH
ANAFLASH develops data-centric compute-in-memory semiconductor IP and processors, including Logic-EFLASH embedded flash memory, Legato-Logic time-domain neural network IP, and the Reflex-CU AI microcontroller, for battery-powered edge AI devices in wearables, automotive, and IoT markets.
- Company typePrivate
- Founded2019
- HeadquartersSunnyvale, United States
- Headcount11–50
- GTM typeB2B
- OfferingHardware or Manufacturing
What ANAFLASH does
ANAFLASH is a semiconductor IP and processor company developing data-centric compute-in-memory (CiM) technology for always-on edge AI. The company's portfolio centers on three integrated offerings: Logic-EFLASH™, a standard-logic-compatible embedded flash memory IP that requires no additional fabrication masks beyond standard CMOS and stores up to 4 bits per cell; Legato-Logic™, a time-domain neural network processor IP that performs multiply-accumulate operations directly inside the memory array to eliminate the processor-memory data bottleneck (acquired via the February 2025 acquisition of Legato Logic); and Reflex-CU™, an always-on near-sensor AI microcontroller that runs inference locally with zero standby power consumption. The technology is targeted at battery-powered edge AI applications including wearables, smart home devices, automotive sensors, and robotics, with reported energy efficiency of 333 TOPS/W and accuracy of 91.8%-98.5% on benchmark workloads.
ANAFLASH monetizes primarily through semiconductor IP licensing to foundries and chip designers (Samsung Foundry, GlobalFoundries, ADTechnology), supplemented by sale of AI microcontroller products (Reflex-CU sampling) and an SDK/professional services layer for deploying pre-trained ML models. The company operates a wholly-owned Korean R&D subsidiary, SEMIBRAIN Inc., which has secured Korean government grants for advanced node development, and partners with KAIST and the University of Minnesota for research. Customers and partners include Samsung Foundry (28nm process), GlobalFoundries (22nm showcase), DRIMAES (automotive infotainment), and ADTechnology (Samsung DSP). The company is venture-backed with more than $22 million in disclosed funding including an $18.2M Series A in November 2024, and has received NSF, NASA, and DoD Air Force SBIR awards alongside funding from Berkeley SkyDeck, Stonebridge Ventures, Lotte Ventures, L&S Venture Capital, Plug and Play Ventures, and Mirae Asset Venture Investment.
ANAFLASH firmographics
Firmographics- Name
- ANAFLASH
- Legal name
- ANAFLASH Inc.
- Website
- https://anaflash.com
- Company type
- Private
- Founded year
- 2019
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- ANAFLASH develops data-centric compute-in-memory semiconductor IP and processors, including Logic-EFLASH embedded flash memory, Legato-Logic time-domain neural network IP, and the Reflex-CU AI microcontroller, for battery-powered edge AI devices in wearables, automotive, and IoT markets.
- Ownership category
- akta.pro rank
ANAFLASH industry classification
Industry- Product category
- Neuromorphic Edge AI Processors
- NAICS
- Computer and Peripheral Equipment Manufacturing (33411), Computer Storage Device Manufacturing (334112), Computer Systems Design Services (541512)
- SIC
- Electronic Computers (3571), Services-Prepackaged Software (7372)
- akta.pro primary industry
- AI Hardware IP & Chiplet/Packaging for Accelerators (EDA/IP, advanced packaging) (HDAAAAAL)
- akta.pro secondary industries
- Model Serving, Inference & Deployment Platforms (APIs, Edge/On-Prem) (HDAEANAC), AI/ML Solution Integration & MLOps Enablement (BPAEAAAJ)
Keywords
Where ANAFLASH is headquartered
LocationHeadquarters
- HQ city
- Sunnyvale
- HQ country
- United States
- HQ region
- North America
Offices2 records
Markets served
ANAFLASH business model
Business model- GTM type
- B2B
- Offering type
- Hardware or Manufacturing
- Cost components
- Technology or R&D, Personnel, Operations, Marketing or Sales, Supply Chain, Infrastructure
Revenue model
- Semiconductor IP Licensing: Licensing of proprietary Logic-EFLASH and Legato-Logic IP to semiconductor companies, foundries, and chip designers. IP can be licensed to collaborative partners for integration into their products.
- AI Microcontroller Products: Sale of AI microcontroller products (Reflex-CU) featuring zero-standby power weight memory. Currently sampled to partners and customers with supporting software development packages.
- SDK and Software Tools: Cross-platform software development toolchain and RCU neural engine converters for pre-trained ML models provided alongside hardware products.
Go-to-market motion1 record
Distribution channels3 records
Marketing channels6 records
ANAFLASH product offering
Product offeringCore offering
ANAFLASH develops and licenses data-centric processor IP and AI microcontrollers for edge AI applications. Its core offerings include Logic-EFLASH™ (standard logic-compatible embedded flash memory IP), Legato-Logic™ (time-domain compute-in-memory neural network processor IP), and Reflex-CU™ (an always-on near-sensor AI computing unit), accompanied by a cross-platform SDK for deploying pre-trained ML models.
Product overview
ANAFLASH is a data-centric processor solution provider for real-time edge intelligence. The company offers a unified portfolio of complementary products: Reflex-CU™️ (always-on near-sensor reflex computing unit for low-power, low-latency AI inference), Logic-EFLASH™️ (standard logic compatible embedded flash memory IP), and Legato-Logic™️ (time-domain neural network processor IP for compute-in-memory). Together, these products enable AI models to be stored and executed directly where data is generated, reducing data movement between processor and memory for significantly lower power consumption. The SDK provides cross-platform development tools for deploying pre-trained ML models onto these hardware platforms.
Differentiator
Problem solved
Functional benefit
Brands
- Reflex-CU: Always-on Intelligent Near-Sensor Reflex Computing Unit for low power and low latency smart sensors featuring nonvolatile AI computing technology coupled with FLASH memory
- Logic-EFLASH
- Legato-Logic
- SEMIBRAIN
Products and services
- Reflex-CU™ (Reflex Computing Unit) An always-on intelligent near-sensor Reflex Computing Unit (RCU) for low-power and low-latency smart sensors featuring non-volatile AI computing technology coupled with FLASH memory. Inspired by the human nervous system, it provides rapid reflex-like responses without engaging the main processor, even during sleep mode. Targeted at manufacturers of battery-powered smart edge devices.
- Logic-EFLASH™ Standard Logic Compatible Embedded Flash Memory IP — a non-volatile memory that can be built with no additional chip fabrication steps beyond standard CMOS logic process. Scalable in advanced logic process technologies, stores multi-level (4-bits/cell) data for AI weight storage, and provides zero standby power. Licensed to semiconductor companies and foundries for integration into their products.
- Legato-Logic™ Time-Domain Neural Network Processor IP for Low Power and Low Latency AI Computation — a standard-logic-compatible time-domain compute-in-memory technology that eliminates excessive data traffic between processor and memory. Uses near-threshold supply operation with proprietary time self-tracking architecture for robust operation under PVT variation without power-hungry ADCs and DACs. Licensed to semiconductor companies for integration.
- Software Development Kit (SDK) Cross-platform software development toolchain and RCU neural engine converters for pre-trained ML models. Supports customized neural network models on smart edge devices. Bundled with hardware products and IP evaluations for enterprise customers.
- NAND Flash Based Neural Network Accelerator A 3D NAND flash compatible compute-in-memory array featuring multi-bit weight storage, 128 NAND layers, 40-bitline multiply-and-accumulate operation, and integrated signed weight bitline scheme. Based on exclusively licensed technology from University of Minnesota.
Quantifiable outcome
- 333 TOPS/W energy efficiency achieved with Logic-EFLASH compute-in-memory macro
- +3 more outcomes
Companies that use ANAFLASH
Customer profileNamed customers4 records
Segments3 records
Ideal customer profiles3 records
ANAFLASH technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- No
- API docs
- API detail
Core technology
AI maturity
App detail
AI capability2 records
Feature4 records
ANAFLASH partnerships and signals
Strategic signalPartnerships
Nine partnerships are on record, tiered core, flagship and minor.
- GlobalFoundriescoreGlobalFoundries fabricated ANAFLASH's Legato-Logic IP using their 22nm technology node. This support enables the company to showcase their Time-Domain Neural Network Processor IP for low-power, low-latency AI computation.
- Legato LogicflagshipANAFLASH acquired Legato Logic's time-based compute-in-memory technologies and industry veterans. Shahrzad Naraghi, CEO of Legato Logic, joined ANAFLASH as CTO. The combined company now offers both Compute-near-Flash and Compute-in-Flash technologies.
- DRIMAESflagshipJoint business agreement to develop AI-powered infotainment platform for next-gen mobility solutions. The partnership combines ANAFLASH's edge AI processor technology with DRIMAES's infotainment platform development capabilities. This is the first official collaboration under South Korea's global technology cooperation initiative.
- ADTechnologycoreCollaboration at Embedded Vision Summit in California to showcase 28nm zero layer embedded non-volatile memory IP (28nm Logic E-FLASH IP). ADTechnology is a Samsung Foundry Design Solution Partner focusing on advanced ASIC SoCs.
- University of MinnesotacoreANAFLASH acquired an exclusive license for single-poly based embedded flash memory technology from Professor Chris Kim's group at the University of Minnesota. The technology was recognized by IEEE with the Low Power Design Contest award.
- KAISTcoreResearch partner that fabricated energy-efficient edge AI test-chip using ANAFLASH's logic-compatible embedded flash IP. The IMPACT Lab at KAIST has prototyped a Compute-in-Flash based AI engine and a hybrid-domain sensing technique AI engine.
- Samsung FoundrycoreANAFLASH develops embedded flash memory IP using Samsung Foundry's 28nm process. The collaboration originated through SEMIBRAIN subsidiary winning Samsung's fabless challenge contest in 2022. Multiple research grants were awarded by Korean government agencies to develop Logic-EFLASH technology using Samsung Foundry.
- KICminorPartnership with KIC Silicon Valley to help engage with global customers and investors. Through KIC support, ANAFLASH collaborates with AI technology startups in Korea.
- SEMIBRAINcoreWholly owned subsidiary of ANAFLASH in Korea that received $4 million grant from Korean government funding agencies to develop embedded flash memory IP in Samsung Foundry advanced process nodes. Collaborates with KAIST, Korea University, and ETRI.
Scale indicators10 records
Recent moves6 records
Expansion highlights7 records
ANAFLASH competitors and assessment
Company assessmentDirect peers
- Sima.ai: Sima.ai develops edge AI inference accelerators for vision and sensor applications. Comparable focus on edge AI silicon, with similar customer types (edge device OEMs) and competing for design wins in low-power on-device AI.
- Mythic: Mythic builds analog compute-in-memory AI inference processors for edge applications. Highly comparable technology approach (analog/near-memory compute) and similar end-market focus (vision AI at the edge). Both companies target the power/performance gap between microcontrollers and cloud inference.
- Hailo: Hailo makes edge AI processors optimized for high-performance inference at low power on smart cameras and edge devices. Competes with ANAFLASH in edge AI accelerators and shares target verticals (smart cameras, automotive, industrial), though Hailo focuses on higher-throughput workloads than ANAFLASH's always-on microcontrollers.
- Syntiant: Syntiant develops ultra-low-power neural network processors for always-on edge AI inference in battery-powered devices. Directly comparable to ANAFLASH's Reflex-CU on target use case (battery-powered edge AI), customer segment (IoT, hearables, wearables), and core value proposition (low-power on-device inference without cloud).
- GreenWaves Technologies: GreenWaves produces GAP-series RISC-V based AI processors for ultra-low-power edge AI in hearables, IoT sensors, and wearables. Direct peer on architecture (RISC-V + dedicated AI engine), use case (battery-powered always-on devices), and target customers.
- BrainChip: BrainChip develops neuromorphic processors (Akida) for ultra-low-power edge AI. Comparable as a peer in event-driven, brain-inspired AI silicon for battery-powered smart edge devices, and as a competitor for design wins in industrial IoT and automotive ADAS.
Emerging players
- Graphcore: Graphcore builds Intelligence Processing Units (IPUs) for AI workloads. A fellow emerging AI accelerator company pursuing novel compute architectures (though Graphcore targets datacenter rather than edge), and a useful comparable for technology-validation trajectory and capital intensity in the AI silicon space.
- Esperanto Technologies: Esperanto Technologies develops RISC-V based AI accelerators with thousands of cores. Comparable as an emerging compute-in-memory / massively parallel AI silicon player targeting inference workloads, with similar focus on energy-efficient AI compute.
Broad incumbents
- Arm (Ethos / Cortex-M): Arm offers Ethos NPU IP and Cortex-M microcontrollers widely used in edge AI devices. Represents the dominant incumbent in embedded processor IP that ANAFLASH must displace or complement, with far greater scale, software ecosystem, and design-in footprint.
- Synopsys (ARC Processors): Synopsys offers ARC processor IP including AI/ML extensions for embedded and edge AI applications. A broad incumbent in processor IP that competes with ANAFLASH's Legato-Logic IP licensing model, with a much larger ecosystem, established toolchain, and incumbent customer relationships.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat5 records
Key risks6 records
Key highlights8 records
Customer concentration
ANAFLASH social profiles
Digital presenceANAFLASH financial estimates
Financial estimateRevenue estimate
Valuation estimate
ANAFLASH leadership team
Management profileNumber of profiles
Profiles9 records
ANAFLASH subsidiaries and ownership
Company hierarchySubsidiaries1 record
ANAFLASH funding detail
Funding detailFunding overview
Funding rounds10 records
Investors12 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
ANAFLASH M&A and investment
M&A and investmentM&A1 record
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about ANAFLASH
What does ANAFLASH do?
ANAFLASH develops and licenses data-centric processor IP and AI microcontrollers for edge AI applications. Its core offerings include Logic-EFLASH™ (standard logic-compatible embedded flash memory IP), Legato-Logic™ (time-domain compute-in-memory neural network processor IP), and Reflex-CU™ (an always-on near-sensor AI computing unit), accompanied by a cross-platform SDK for deploying pre-trained ML models.
Is ANAFLASH a public or private company?
ANAFLASH is a private company. It is classified as venture growth investor backed and is currently operating.
When was ANAFLASH founded?
ANAFLASH was founded in 2019. It employs 11 to 50 people.
Where is ANAFLASH based?
ANAFLASH is headquartered in Sunnyvale, United States, in the North America region.
How does ANAFLASH make money?
Three revenue lines are on record. Semiconductor IP Licensing is the primary driver. The others are AI Microcontroller Products and SDK and Software Tools.
Who are ANAFLASH's main competitors?
Direct peers on record are Sima.ai, Mythic, Hailo, Syntiant, GreenWaves Technologies and BrainChip. Emerging players are Graphcore and Esperanto Technologies. Broad incumbents are Arm (Ethos / Cortex-M) and Synopsys (ARC Processors).
Does ANAFLASH have an API?
No public API is recorded for ANAFLASH.
What industry is ANAFLASH in?
ANAFLASH's product category is Neuromorphic Edge AI Processors. Its primary akta.pro industry code is HDAAAAAL, AI Hardware IP & Chiplet/Packaging for Accelerators (EDA/IP, advanced packaging), with a secondary code of HDAEANAC, Model Serving, Inference & Deployment Platforms (APIs, Edge/On-Prem). Its NAICS code is 33411 and its SIC code is 3571.