FemtoAI Unlocks 5X Growth With Full-Stack AI Platform

ByRicardo

femtoAI’s Dual Sparsity Delivers 100X Power & Energy Gains, 10X Less Memory for Customers Including Samsung, Marshall

femtoAI, a number one AI inference platform and creator of the Sparse Processing Unit (SPU), as we speak introduced 5X progress in 1H 2026, pushed by engagements with Samsung and Marshall, new buyer use instances and elevated developer adoption.

The demand for AI inference globally continues to outpace provide, and femtoAI is filling a spot. With 200K+ chips already deployed, and prospects throughout knowledge middle, enterprise and shopper electronics, femtoAI delivers AI inference options that eat 100X much less energy & vitality with 10X much less reminiscence to make AI economically viable on each machine.

“We are proud that we’re profitable and scaling throughout a number of markets due to our distinctive acceleration of sparsity within the AI stack to scale back reminiscence and energy with out compromising efficiency,” stated Sam Fok, co-founder and CEO at femtoAI. “From our preliminary wins with our first-generation SPU with corporations similar to NewSound and now Marshall, to new use instances similar to fault monitoring for AI factories, to sensible glass and robotics, femtoAI is solidifying its place as one of many main options for on-device, low-power, low-memory AI on each machine. Our next-generation SPU shall be much more groundbreaking.”

Highlights for femtoAI in 2026 embrace:

Customer Growth

  • Over 200K+ SPU-001 chips at the moment in market, underscoring the rising market demand for a differentiated AI inference answer
  • Six tier-1 next-generation sensible glass distributors together with Orka, have chosen femtoAI as their on-device AI answer
  • AI-enhanced audio will attain extra customers as femtoAI and Marshall companion throughout premium headphone and earbud households
  • 5-year partnership with ABOV Semiconductor and Samsung continues to evolve throughout a wide range of dwelling home equipment

Developer Demand & Engagement at developer.femto.ai:

  • Increased developer engagement with practically half of all prospects constructing their very own algorithms and AI fashions within the femtoAI developer portal
  • femtoAI companions straight with builders to uncover new inference options, each regionally and within the cloud, similar to a brand new knowledge middle vitality administration system, robotics and imaginative and prescient denoising, amongst others

Product Innovation:

  • Launch of femtoAI ClaraCall 3.0 and ClaraCall 3.0-S fashions for next-generation AI name high quality enhancement, out there on the femtoAI developer platform
  • The next-generation SPU is at the moment underway to satisfy growing buyer demand for low-energy, low-power, and low-memory AI inference
  • Decrease reminiscence utilization through mannequin compression, leading to 10X much less reminiscence required for open-source fashions similar to OpenAI Whisper and customized options like native voice instructions

Recognition & Awards:

  • femtoAI was awarded CES 2026 Best-in-Show award for SPU-001
  • Named a slicing–edge AI inference answer by Bessemer Venture Partners
  • Honored as EDGE AI Foundation “Rookie of the Year” 2026

Dual Sparsity is the Key Ingredient
femtoAI is in a position to scale back reminiscence by as much as 10X and enhance vitality effectivity by as much as 100X with its SPU Platform by making use of a novel idea: twin sparsity. Sparsity is a recognized strategy of decreasing pointless compute and storage, enabling the quickest and best path to activity completion. femtoAI’s twin sparsity strategy delivers these positive aspects by leveraging this idea concurrently at each {hardware} (chip) and software program (tooling, functions) ranges to make sure minimal pointless processing happens when finishing a activity.

Swetha Srinivasan, Author of The Thesis, a expertise and semiconductor e-newsletter, lately said, “Getting to the following degree means wanting previous the usual playbook of simply scaling compute. What’s thrilling about femtoAI is the brain-inspired wager on sparsity, and the truth that they’re fixing it at each layer of the stack: sparsifying the mannequin itself and designing silicon particularly for sparse compute. That full-stack alignment is a holistic strategy that AI truly wants.”

The put up FemtoAI Unlocks 5X Growth With Full-Stack AI Platform first appeared on AI-Tech Park.

Similar Posts