BrainChip Launches Sequora: Edge AI That Doesn't Just Recognize — It Understands and Responds

Customers can integrate the licensable design into their own chips and choose their foundry and manufacturing process node.

Rhea-AI Summary

BrainChip (BCHPY, BRCHF) launched Sequora Core IP, a processor design for running generative AI directly on battery-powered devices. Sequora is available now as licensable semiconductor intellectual property, a design customers can incorporate into their own chips.

BrainChip says its state-space architecture keeps memory use fixed as conversations or data streams grow and supports models from a few million parameters up to 3 billion. Paired with the company's Akida platform for processing sensor data, Sequora handles generative and sequential AI, including large language models. Processing takes place on the device without sending data off it.

Key Figures

Maximum model size: 3 billion parameters

Maximum model size

3 billion parameters

Sequora scales from models of a few million parameters

Key Terms

large language models (llms), state-space design, system-on-chip (soc), register-transfer level (rtl)

4 terms

large language models (llms) technical

"including large language models (LLMs), directly on battery-powered edge devices"

Large language models (LLMs) are advanced computer programs trained on massive amounts of text to generate, summarize, translate and understand human-like language. For investors they matter because LLMs can act like a very fast, experienced research assistant—automating customer service, speeding product development and cutting costs—while also creating new revenue opportunities and regulatory, accuracy and ethical risks that can affect a company’s profits and reputation.

state-space design technical

"a state-space design keeps memory use fixed"

A method from control theory for modeling and designing dynamic systems using a set of state variables and matrix equations that describe how those variables evolve and how inputs produce outputs. The system is represented in “state-space” form (typically x' = Ax + Bu, y = Cx + Du), and design work adjusts feedback laws or observer equations (for example state feedback, pole placement, or optimal control like LQR) to meet goals such as stability, speed of response, or disturbance rejection. The approach requires a mathematical model of the system and either direct measurement of the state variables or an observer (estimator) to infer them; designs for linear models do not automatically apply to nonlinear dynamics without modification.

system-on-chip (soc) technical

"integrate the design into their own system-on-chip (SoC)"

A system-on-chip (SoC) is a single microchip that packs the main electronic components of a computer — such as the processor, memory, graphics and input/output controllers — that would otherwise be separate parts. Like a Swiss Army knife replacing multiple tools, an SoC reduces size, power use and manufacturing cost while shaping device performance and features, so investors watch it as a key driver of product competitiveness, margins and market share.

register-transfer level (rtl) technical

"Delivered as process-portable, fully synthesizable register-transfer level (RTL) code"

A register-transfer level (RTL) is a way engineers describe how a digital chip moves and processes data by showing how information flows between storage points called registers and the logic that transforms it. Think of it like a flowchart for an electronic device that specifies the steps and timing for moving and changing data. RTL matters to investors because it is the core design stage that determines a chip’s performance, power use, and how long and costly it will be to bring the product to market.

AI-generated analysis. How Rhea-AI works. Not financial advice.

See more from StockTitan in Google Search and AI answers. Adds StockTitan as a preferred source · opens Google

Add on Google Not now

Sequora extends BrainChip’s proven ultra-low-power, always-on sensing advantage to generative AI

LAGUNA HILLS, Calif., Oct. 14, 2026 (GLOBE NEWSWIRE) -- BrainChip Holdings Ltd. (ASX: BRN, OTCQX: BRCHF, ADR: BCHPY), a global leader in ultra-low-power, fully digital, event-based neuromorphic AI, today announced BrainChip Sequora^™ Core IP, a processor IP core that runs generative AI, including large language models (LLMs), directly on battery-powered edge devices.

The next step in BrainChip's edge AI roadmap, Sequora builds on the strengths of the company's Akida^™ platform. Its efficiency comes from model structure: a state-space design keeps memory use fixed no matter how long the data stream or conversation runs. As a result, Sequora scales from models of a few million parameters up to 3 billion. Devices that once could only recognize a signal can now understand, explain and respond — at the power levels customers expect from BrainChip.

"Our commercial and aerospace/defense customers are designing AI into many of the next-generation edge systems that provide real-time processing and decision-making in congested and contested environments with minimal footprint in size, weight and power," said Frank Ferrante, chief executive officer at ForwardEdge ASIC. "Sequora puts generative AI in a low-power IP core that fits many of the demands of the custom silicon we build. It gives us a path to add on-device intelligence to mission-critical edge ASICs."

What Sequora Does

Sequora removes the memory and power bottleneck that has kept generative AI out of battery-powered devices such as earbuds and sensor nodes. Paired with Akida, it brings two workloads together in one solution: Akida processes streaming sensor data — vibration, audio, radar, raw radio-frequency (RF) IQ data and biological signals — while Sequora runs generative and sequential AI, including on-device LLMs, voice-to-voice interaction and AI agents.

Markets from consumer voice and wearables to aerospace and defense, automotive and industrial radar, industrial IoT, and medical and biomedical devices need conventional inference and generative inference in the same product.

"The edge is where AI becomes useful in everyday infrastructure, from utility meters to mobility systems," said Jay Kim, CEO of edgeAI Inc. "BrainChip saw early that efficiency would decide who wins at the edge. Now they're applying that same thinking to generative AI, bringing language and reasoning capabilities to devices that run on a small battery."

Key challenges Sequora addresses:

• Generative models stall on memory bandwidth
• Per-word cost that climbs as the conversation (context window) and its cache grow
• Memory traffic caps model size on battery power
• Make run-time branching predictable for automotive, medical and avionics certification
• On-device per-query latency and privacy
  
  

Sensing and Generative AI on One Device, One Battery

“Customers are not asking for just a signal classifier or a language model,” said Sean Hehir, chief executive officer, BrainChip. “They want both, in the same product, on the same battery. Akida and Sequora together deliver exactly that.”

Because Sequora runs generative and sequential AI without sending data off the device, it addresses a growing market priority: keeping data and AI models sovereign while minimizing size, weight and power.

Availability

Sequora is available now as licensable semiconductor IP. Customers integrate the design into their own system-on-chip (SoC) and manufacture it at the foundry and process node of their choice. Delivered as process-portable, fully synthesizable register-transfer level (RTL) code, the Sequora package includes verification IP and testbench, synthesis scripts and timing constraints, and an SDK with a static graph compiler, a C++ runtime library and a model zoo.

About BrainChip Holdings Ltd.

BrainChip is the worldwide leader in edge AI on-chip processing and learning. The company’s first-to-market, fully digital, event-based AI processor, Akida^™, uses neuromorphic principles to analyze only essential sensor inputs at the point of acquisition, processing data with unmatched efficiency, precision and energy economy. BrainChip’s Sequora^™ is an efficient digital AI processor for GenAI and other large models using state-space architectures. BrainChip’s technology is deployed across industries including aerospace, defense, industrial IoT, healthcare, consumer devices, and wearables. Explore more at www.brainchip.com.

Follow BrainChip:

X: https://www.twitter.com/BrainChip_inc

LinkedIn: https://www.linkedin.com/company/7792006

Investor Contact

IR@brainchip.com

Media Contact

Maddie Coe

PRforBrainchip@bospar.com

^{© 2026 BrainChip Holdings Ltd. All rights reserved. Internal use only}

FAQ

AI-generated questions and answers. How Rhea-AI works. Not financial advice.

Is BrainChip Sequora available for licensing?

Sequora is available now as licensable semiconductor intellectual property. Customers integrate the processor design into their own system-on-chip and manufacture it at the foundry and process node of their choice.

What does BrainChip include in the Sequora licensing package?

The Sequora package includes processor design code, verification tools and a software development kit. It is delivered as process-portable, fully synthesizable register-transfer level code, with verification IP and a testbench, synthesis scripts and timing constraints. The software kit includes a static graph compiler, a C++ runtime library and a collection of models.