EnCharge AI

What EnCharge AI does

EnCharge AI designs and develops analog in-memory computing (IMC) AI accelerator chips and supporting software, with a flagship EN100 product launched in May 2025 that delivers approximately 150-200 TOPS/W (200 TOPS at roughly 8W). Its core technology — branded Virtualized In-Memory Compute (VIMC) — performs matrix multiplications directly inside SRAM memory cells using charge-domain computation through metal capacitors fabricated in standard CMOS, eliminating the data-movement bottleneck that dominates digital AI accelerator energy budgets. The company offers the EN100 in PCIe card, ASIC, and chiplet form factors paired with a full software stack (compilers, runtimes, quantization tools) that orchestrates AI inference from edge devices through data centers, and supports a broad model family including Qwen3, GPT-OSS, Mistral, Phi, and Llama.

Name

EnCharge AI

Legal name

EnCharge AI, Inc.

Website

https://enchargeai.com

Company type

Private

Founded year

2022

Operating status

Operating

Headcount range

11–50 employees

Short description

EnCharge AI designs analog in-memory computing AI accelerator chips, led by the EN100 product delivering roughly 200 TOPS at 8W, and sells to enterprise and strategic customers across consumer electronics, defense, aerospace, and AI PC markets.

Ownership category

akta.pro rank

Headquarters

HQ city

Santa Clara

HQ country

United States

HQ region

North America

Offices2 records

Markets served

GTM type

B2B

Offering type

Hardware or Manufacturing

Cost components

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

Revenue model

1. AI accelerator hardware sales: Sale of proprietary AI accelerator products (e.g., EN100) in standard PCIe form-factor cards, ASICs, and chiplets for edge-to-cloud deployment. The company states its core technology leverages the existing semiconductor supply chain and integrates into diverse form factors for partner and customer solutions. Commercial launch began in 2025 following the $100M Series B.
2. Software platform and full-stack AI solutions: Companion software platform (compilers, runtimes, quantization tools, system software) bundled with hardware for orchestration between on-device and cloud deployments. Sold as part of the full-stack AI solution rather than as a standalone product offering.
3. Strategic / OEM licensing (inferred): Strategic capital from Samsung Ventures and HH-CTBC (Foxconn) suggests forthcoming OEM licensing and consumer electronics integration pathways, supporting an embedded/licensing revenue motion alongside direct hardware sales.

Pricing tiers

Go-to-market motion4 records

Distribution channels5 records

Marketing channels8 records

Core offering

EnCharge AI designs and sells analog in-memory computing AI accelerator chips and a bundled full-stack software platform for AI inference at the edge and in the cloud. The flagship product is the EN100 AI accelerator, which performs matrix multiplications directly inside SRAM cells using charge-domain computation via metal capacitors fabricated in standard CMOS, achieving approximately 150-200 TOPS per watt and targeting use cases from milliwatt edge sensors to multi-watt client and data-center deployments.

Product overview

EnCharge AI offers a unified platform-plus-modules product architecture built on its proprietary Virtualized In-Memory Compute (VIMC) technology, which performs matrix operations directly within memory arrays using capacitor-based analog computation in standard CMOS. The flagship core product is the EN100 AI Accelerator — described as the industry's first AI accelerator built on scalable analog in-memory computing — which delivers approximately 150-200 TOPS/W and is available as the EN100 PCIe Card standard form-factor. These hardware offerings are paired with a full software platform/tools stack that orchestrates workloads from edge devices (laptops, drones, robots, vehicles) to cloud data centers. The platform targets generative AI inference, Vision-Language-Action models, and on-device AI assistants, with deployment configurations spanning chiplets, ASICs, and PCIe cards to fit diverse partner and customer solution requirements.

Differentiator

Problem solved

Functional benefit

Brands

• EN100: EnCharge AI's first-of-its-kind AI accelerator for on-device computing, built on precise and scalable analog in-memory computing architecture.

Segments5 records

Ideal customer profiles4 records

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

Integration1 record

AI capability7 records

Feature9 records

Partnerships

Four partnerships are on record, tiered strategic and flagship.

• TSMC (Taiwan Semiconductor Manufacturing Company)strategicStrategic or Co-development Partner · 24 September 2025EnCharge AI participated as a featured speaker at TSMC's 2025 Open Innovation Platform (OIP) Ecosystem Forum in Santa Clara, CA on September 24, 2025. CEO Naveen Verma presented on the future of semiconductor design and the need for high-performance, energy-efficient designs as AI proliferates from the data center to the edge. Anchors foundry/manufacturing ecosystem access.
• National Semiconductor Technology Center (NSTC)flagshipStrategic or Co-development Partner · 27 February 2025EnCharge AI joined the NSTC as one of the first 100 members since its establishment in early 2024. NSTC is a public-private consortium established by the CHIPS Act to support and extend U.S. leadership in semiconductor research, design, engineering, and advanced manufacturing. CEO Naveen Verma delivered a keynote on American innovation at the NSTC Symposium in Washington, D.C.
• DARPA (Defense Advanced Research Projects Agency)flagshipStrategic or Co-development Partner · 8 March 2024EnCharge AI participates in DARPA's $78 million OPTIMA program in collaboration with Princeton University to develop advanced AI processors that improve efficiency for AI workloads. Technology originates from DARPA and DoD-funded research. The company's EN100 chip was developed through partnership with DARPA and is being deployed with defense contractors for tactical-edge AI.
• Princeton UniversityflagshipStrategic or Co-development Partner · 6 March 2024Co-development partnership with Princeton University on groundbreaking AI chip technology, supported by the U.S. Department of Defense (DoD). The partnership builds on six years of foundational in-memory computing research led by EnCharge co-founder and CEO Naveen Verma, a Princeton ECE professor since 2009. The collaboration is part of DARPA's $78M OPTIMA program developing analog in-memory computing chips for AI deployment beyond data centers to edge devices.

Scale indicators16 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Mythic AI: Closest direct competitor — also develops analog in-memory computing AI accelerators targeting edge inference. Compares to EnCharge on core architecture (analog IMC), target use cases (on-device AI), and form factor (PCIe cards/embedded), making it the most directly comparable peer despite a different underlying memory technology (Mythic uses NOR flash, EnCharge uses capacitors in SRAM).
• Cerebras Systems: Develops wafer-scale AI accelerator chips for training and inference. Comparable as a high-profile AI silicon startup chasing compute efficiency advantages through novel architectures, with overlapping enterprise and government customer targets.
• SambaNova Systems: Builds AI accelerator chips and full-stack platforms (DataScale) for enterprise and government. Comparable as a well-funded AI chip startup pursuing both data-center and edge inference with custom architecture and software stack.
• Groq: Develops LPU inference accelerators focused on low-latency, high-throughput AI inference. Comparable as an AI chip startup with a custom architecture targeting efficiency, with overlapping enterprise/government channel partners (including In-Q-Tel).
• Tenstorrent: Designs RISC-V-based AI accelerators and chiplets for training and inference. Comparable as a custom-architecture AI chip startup pursuing chiplet-based designs and developer ecosystem partnerships.
• Hailo: Develops edge AI accelerator chips (Hailo-8, Hailo-15) for automotive, industrial, and IoT. Comparable as a fabless AI chip startup targeting the same edge inference use cases with a custom dataflow architecture.

Emerging players

• Lightmatter: Builds silicon-photonics-based AI accelerators and interconnects for data centers. Comparable as a novel-architecture AI compute startup pursuing massive efficiency gains, though focused on data-center rather than edge deployments.
• Syntiant: Designs ultra-low-power neural decision processors for always-on edge AI in IoT and hearables. Comparable on edge AI silicon for ultra-low-power deployments, though focused on smaller models and sensing rather than large-model inference.

Broad incumbents

• NVIDIA: Dominant GPU and AI accelerator incumbent offering edge (Jetson) and data-center (H100/B200) AI compute. The primary competitor EnCharge must displace; relevant because NVIDIA's roadmap and CUDA software moat are the biggest adoption barriers for any AI chip startup.
• Intel (Habana Labs): Intel's Habana Labs (Gaudi, Greco) develops AI accelerators for training and inference, alongside Intel's broader Xeon and Core Ultra AI PC silicon. Comparable as a broad incumbent offering edge-to-data-center AI accelerators with much larger engineering and channel resources.

Market position

Strengths4 records

Weaknesses4 records

Competitive moat5 records

Key risks5 records

Key highlights6 records

Customer concentration

• YouTubeyoutu.be/StCPm3LIhEA
• LinkedInlinkedin.com/company/encharge-ai

Compliance3 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles14 records

Funding overview

Funding rounds5 records

Investors26 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 EnCharge AI