EnCharge AI
Startup building an analog in-memory chip for efficient on-device AI inference.
On the record
Checkpoint: EN100 first shipment to a named paying customer or tier-1 OEM design win
The call
EnCharge has real chip IP but no shipped product yet, while better-funded rivals and NPU makers already sell into the same laptop market.
EnCharge AI designs computer chips that run artificial intelligence programs more efficiently than existing options, based on research from Princeton University. The company has real patents and talented engineers, but its main product (called EN100) has not been released to actual customers at full scale yet, even 14 months after the company said it would ship, while competitors like Axelera and d-Matrix have raised hundreds of millions of dollars more and are already selling to customers. The biggest risk is that Nvidia (which controls almost all AI chip sales and just bought a competitor called Groq for $20 billion) and phone makers like Qualcomm and Apple can copy EnCharge's efficiency gains using their own teams and resources, so the race comes down to which company can build and sell products faster and who has the most money to survive.
Why
01
AI pressure: Qualcomm Snapdragon and Apple Neural Engine NPUs already built into laptops. Built-in NPUs already run on-device AI tasks, cutting the reason a laptop maker would add a separate EnCharge chip. Could AI do the core job? Only partly.
02
Bigger players: Nvidia (dominant AI inference share, bought Groq for roughly $20B in Dec 2025). Nvidia now owns Groq's inference tech and could fold similar efficiency gains into its own chips without needing EnCharge. Could a giant copy it? Probably.
03
What protects it. Patented analog in-memory computing IP from a 7-year Princeton program that rivals cannot simply copy quickly.
What would change this call
EN100 ships in volume to a named tier-1 OEM or defense customer by early 2027. Would show EnCharge can convert its IP into real commercial traction, not just announcements.
EnCharge closes a Series C at a step-up valuation within the next year. Would show investors still believe it can compete with better-capitalized rivals like Axelera and d-Matrix.
A major laptop OEM (Dell, HP, Lenovo, Samsung) announces a design win with Qualcomm, Arm, or Nvidia NPUs instead. Would show incumbents are locking the channel before EnCharge ships.
EnCharge closes a Series C at a step-up valuation within the next year. Would show investors still believe it can compete with better-capitalized rivals like Axelera and d-Matrix.
A major laptop OEM (Dell, HP, Lenovo, Samsung) announces a design win with Qualcomm, Arm, or Nvidia NPUs instead. Would show incumbents are locking the channel before EnCharge ships.
Key numbers
- EN100 still unshipped 11 months after May 2025 launch
- Axelera AI raised $250M+, total funding over $450M
- d-Matrix raised $275M in latest round
Named risks
- EN100 chip still unshipped 11+ months after May 2025 announcement, per April 2026 fab status.
- No new funding round since Feb 2025 Series B while rivals Axelera and d-Matrix raise $250M-275M+.
- Nvidia's $20B Groq acquisition signals consolidation appetite that could sideline smaller inference-chip startups.
- No named enterprise or OEM customers publicly disclosed; revenue figures are unverified aggregator estimates.
Rivals, as read in July 2026
| Rival | Direction | Evidence |
|---|---|---|
| Nvidia | losing to | Holds ~92% AI-inference hardware share and bought Groq for ~$20B, Dec 2025 |
| Axelera AI | gaining on | Raised $250M+ Feb 2026, total >$450M, claims 500+ customers |
| d-Matrix | gaining on | Reported $275M raise, Nov 2025, ~250 employees vs EnCharge's smaller team |
| Qualcomm/Arm NPUs | contesting | Already shipping in millions of laptops, EnCharge's exact target segment, as of 2024-2026 |
Questions people ask
- Has EnCharge AI actually shipped its EN100 chip yet?
- As of April 28, 2026, EnCharge's own Chief Scientist Shwetank Kumar described EN100 as 'returning from fab imminently,' nearly 11 months after the chip was announced on May 29, 2025. No named paying customer or shipped-volume figure has been publicly disclosed as of July 2026.
- Is EnCharge AI profitable?
- No audited revenue or profitability figures exist. Third-party aggregator estimates (Datanyze, StartupHub, PrivCo) range wildly and conflict, so EnCharge AI's actual revenue and profitability status remain unverified as of July 2026; the company has raised over $144M in venture funding instead of disclosing earnings.
- Who are EnCharge AI's biggest competitors?
- EnCharge competes with better-funded analog/in-memory chip rivals Axelera AI ($450M+ raised, $250M+ round Feb 2026) and d-Matrix ($275M raised, Nov 2025), plus incumbent Nvidia, which holds roughly 92% of AI-inference hardware share and bought Groq for about $20B in December 2025.
- Could Nvidia or Qualcomm just replicate what EnCharge AI is doing?
- EnCharge holds patented 'charge-domain' analog in-memory compute IP from seven years of Princeton research, but Qualcomm and Arm NPUs already ship in millions of laptops, EnCharge's exact target market, and Nvidia's Dec 2025 $20B Groq acquisition shows it prefers buying inference-chip innovators over being blocked.
- Is EnCharge AI a safe company to join right now?
- EnCharge has raised $144M total and made leadership hires (Chief Scientist, VP Finance, HR Director) in 2025, but its flagship EN100 chip remained unshipped as of April 2026, over a year post-announcement, and no new funding round has closed since the February 2025 Series B.
- When did EnCharge AI last raise funding?
- EnCharge AI's most recent confirmed round was a $100M+ oversubscribed Series B led by Tiger Global, closed February 13, 2025, bringing cumulative funding to more than $144M. As of July 2026, no confirmed Series C had been announced, a roughly 17-month gap.
Timeline
- 2026-04-28Chief Scientist says EN100 'returning from fab imminently,' still unshipped 11 months post-announcement
- 2026-02-24Rival Axelera AI raises $250M+, total funding tops $450M
- 2025-12Nvidia acquires inference-chip rival Groq for approximately $20B
- 2025-11Rival d-Matrix reportedly raises $275M
- 2025-05-29EnCharge AI announces and launches EN100 accelerator chip
- 2025-04-17EnCharge adds Chief Scientist, VP Finance, and HR Director post-Series B
- 2025-02-13EnCharge closes oversubscribed $100M+ Series B led by Tiger Global
- 2023-12-05EnCharge raises $22.6M Series A extension with TSMC VentureTech, RTX Ventures
- 2022-12-14EnCharge emerges from stealth with $21.7M Series A led by Anzu Partners
Revision history
24 July 2026
First assessed. Losing its protection. The text above is fixed as of this date and is not edited afterwards.
In the same band
Sources
- EnCharge AI Announces EN100 (2025-05-29)
- Making AI inference Cheap with Analog Computing w/ Shwetank Kumar @ EnCharge AI (2026-04-28)
- EnCharge AI Closes Oversubscribed $100M Series B (2025-02-13)
- EnCharge AI emerges from stealth with $21.7M (2022-12-14)
- EnCharge raises $22.6M to commercialize its AI-accelerating chips (2023-12-05)
- Edge AI chip startup Axelera AI raises $250M+ funding round (2026-02-24)
- Axelera AI Secures More Than $250 Million Funding (2026-07-01)
- Enterprise AI Inference Hardware (2026-02-28)
- From A University Lab to $144M in VC Funding (2026-03-31)
- EnCharge AI
- Management and Employees List
- EnCharge AI - $244M Raised (2026-07-23)