BrainChip
BrainChip is an Australian technology company developing neuromorphic processors and software for low-power, event-based artificial intelligence at the edge.
Last updated August 31, 2026
AI crawler reads
No AI crawler has read this profile since Oct 28, 2026.
Overview
BrainChip is an Australian technology company focused on neuromorphic computing, spiking neural networks, and edge artificial intelligence. Founded in 2004 by Peter Van Der Made, the company develops hardware and software intended to run machine-learning inference close to where data is generated, including in cameras, sensors, vehicles, industrial equipment, and other embedded systems. Its approach is designed to reduce dependence on cloud processing by using event-driven computation, local memory, and on-device learning. The company's principal technology platform is Akida, a neuromorphic processor architecture based on spiking neural networks. Unlike conventional processors that generally process large streams of data through sequential or highly parallel numerical operations, Akida is designed to respond to meaningful changes or events in incoming data. This architecture is intended to reduce power consumption, latency, and data movement, particularly in applications involving continuous streams from cameras, microphones, and other sensors. BrainChip's hardware portfolio has centered on the Akida 1000 processor and related development products. The Akida 1000 is described as an event-based neural processing device containing approximately 1.2 million artificial neurons and 10 billion artificial synapses. Its architecture uses a mesh of processing nodes, on-chip memory, and neural processing units capable of executing neural-network layers in parallel. These features are intended to support efficient inference and incremental learning on edge devices. The company subsequently announced the design completion of an Akida 1500 reference chip and introduced a second-generation Akida platform with support for additional neural-network formats and processing capabilities. BrainChip's software offering includes the MetaTF development environment. MetaTF is designed to help developers train, convert, test, and deploy spiking neural networks using image, video, and sensor data. It uses Python-based workflows and can support the transformation of conventional neural-network models into spiking models for deployment on Akida hardware. The environment may also be used to deploy spiking neural networks on non-Akida processors, making it part of BrainChip's broader developer and ecosystem strategy rather than only a hardware configuration tool. The company's target applications include security and surveillance, autonomous and assisted vehicles, industrial automation, robotics, consumer devices, and other embedded systems requiring real-time AI with constrained power or connectivity. BrainChip has pursued partnerships and ecosystem relationships to broaden adoption, including work with NVISO, the Rochester Institute of Technology, Edge Impulse, Arm, and Intel Foundry Services. These relationships connect BrainChip's technology with computer-vision software, academic research, embedded development tools, processor ecosystems, and semiconductor manufacturing capabilities. Corporate ownership and listing arrangements changed significantly in the mid-2010s. Australian mining company Aziana acquired BrainChip in March 2015, and BrainChip subsequently entered the Australian market through a reverse merger involving the dormant Aziana entity in September 2015. In 2016, former Exar chief executive Louis Di Nardo became BrainChip's chief executive, while Van Der Made moved into the chief technology officer role. BrainChip has remained a publicly traded Australian technology company, with its commercial strategy focused on licensing, development hardware, software tools, and adoption of its neuromorphic processor architecture. Exact current executive assignments, headquarters information, and financial metrics are not established by the supplied reference material.
History
BrainChip was founded in Australia in 2004 by Peter Van Der Made around the idea of developing specialized hardware for brain-inspired artificial intelligence. The company focused on neuromorphic computing, an area that seeks to reproduce some characteristics of biological neural systems in electronic systems. Its work centered on spiking neural networks, in which information is represented and processed through discrete neural events rather than only through continuous numerical operations. A major corporate transition took place in March 2015, when Australian mining company Aziana acquired BrainChip. Later that year, in September 2015, BrainChip entered the Australian public market through a reverse merger involving the then-dormant Aziana entity. This transaction placed the BrainChip business on the Australian Securities Exchange and provided a listed corporate structure through which Van Der Made's processor concept could be commercialized. In 2016, Louis Di Nardo, formerly associated with semiconductor company Exar, was appointed chief executive officer. Van Der Made subsequently served as chief technology officer, concentrating on the technical development of the company's neuromorphic architecture. BrainChip's commercial and engineering efforts increasingly converged around the Akida processor family and the MetaTF software environment. The Akida 1000 became the central hardware expression of the company's approach. It was designed as an event-based neural processor with approximately 1.2 million artificial neurons and 10 billion artificial synapses. Its architecture uses a mesh of 80 nodes, with each node containing neural processing units and associated memory. The processor is intended to execute neural-network layers in parallel and retain relevant data on chip, supporting low-power inference and incremental learning on edge devices. BrainChip complemented the processor with MetaTF, a Python-oriented development environment for working with image, video, and sensor data. Developers can use the environment to construct spiking neural networks or convert other neural-network models into spiking forms for deployment. MetaTF also supports deployment of spiking models on conventional processors, broadening the potential development path beyond dedicated Akida hardware. The company's stated application areas have included security and surveillance, autonomous vehicles, industrial automation, and other embedded AI uses. From 2021 onward, BrainChip began moving its platform toward broader developer access. In October 2021 it announced that orders were being accepted for Akida AI Processor Development Kits. In January 2022 it announced orders for Akida AI Processor PCIe boards, giving developers additional ways to evaluate and integrate the processor. During 2022, the company announced a collaboration with NVISO and added the Rochester Institute of Technology to its University AI Accelerator Program. It also became part of the Intel Foundry Services technology-partner ecosystem. The platform continued to evolve in 2023. BrainChip announced completion of the Akida 1500 reference-chip design in January. In March it presented the second generation of the Akida platform, adding support for 8-bit weights and activations, a Vision Transformer engine, and hardware support for Temporal Event-Based Neural Networks. The company also announced integration of the Akida processor family with Arm's Cortex-M85 processor. In the same period, Edge Impulse announced support for the Akida platform, connecting BrainChip's hardware with an embedded machine-learning development ecosystem. BrainChip's development trajectory has therefore combined corporate restructuring, public-market access, specialized processor design, developer tools, and ecosystem partnerships. Its technology strategy remains centered on bringing AI inference and selected learning functions to devices that operate under power, latency, bandwidth, or privacy constraints. The supplied reference material does not establish current headquarters, current executive leadership beyond the historical appointments described above, or detailed financial performance.
- 2023
Akida 1500 reference-chip design is completed
[1]BrainChip announced completion of the design of its Akida 1500 reference chip.
- 2023
Second-generation Akida platform is introduced
[1]The new platform added 8-bit weights and activations, a Vision Transformer engine, and hardware support for Temporal Event-Based Neural Networks.
- 2023
Akida integrates with Arm Cortex-M85
[1]BrainChip announced integration of its Akida processor family with Arm's Cortex-M85 processor.
- 2022
Akida PCIe boards become orderable
[1]BrainChip announced order availability for Akida AI Processor PCIe boards.
- 2022
BrainChip expands academic and industry relationships
[1]The company announced collaboration with NVISO, added the Rochester Institute of Technology to its university accelerator program, and joined the Intel Foundry Services ecosystem.
- 2021
Akida development kits become orderable
[1]BrainChip announced that customers could place orders for development kits based on its Akida AI processor.
- 2016
Louis Di Nardo becomes chief executive
[1]Former Exar chief executive Louis Di Nardo was appointed BrainChip's CEO, while founder Peter Van Der Made moved into the CTO role.
- 2015
Aziana acquires BrainChip
[1]Australian mining company Aziana acquired BrainChip in March, initiating a major corporate restructuring.
- 2015
BrainChip enters the ASX through a reverse merger
[1]In September, BrainChip was placed on the Australian Securities Exchange through a reverse merger involving the dormant Aziana entity.
- 2004
BrainChip is founded
[1]Peter Van Der Made founded BrainChip in Australia to develop specialized hardware for brain-inspired artificial intelligence.
Products and positioning
BrainChip positions itself as a specialist in energy-efficient, event-based edge AI. Its differentiation is based on neuromorphic hardware and spiking neural networks intended to perform inference and selected forms of learning locally, with lower latency, reduced data movement, and potentially lower power consumption than conventional cloud-dependent or general-purpose AI approaches.
Neuromorphic processor
Akida 1000
The Akida 1000 is BrainChip's event-based neuromorphic processor for edge AI. It is described as containing approximately 1.2 million artificial neurons and 10 billion artificial synapses. Its 80-node mesh architecture combines neural processing units with local memory and is designed to execute neural-network layers in parallel. The design supports local inference and incremental learning for applications that need low latency and constrained power consumption.
Reference chip · 2023
Akida 1500
The Akida 1500 is a later reference-chip design in BrainChip's neuromorphic processor family. BrainChip announced completion of its design in January 2023. The supplied material does not provide complete specifications, commercial availability details, or a definitive production status.
Neuromorphic AI platform · 2023
Second-generation Akida platform
BrainChip's second-generation Akida platform extends the original architecture with support for 8-bit weights and activations, a Vision Transformer engine, and hardware support for Temporal Event-Based Neural Networks. The additions are intended to broaden the types of modern AI models that can be used in power-efficient edge-computing systems.
AI development environment
MetaTF
MetaTF is BrainChip's software environment for developing and deploying spiking neural networks. It uses Python-based workflows and supports image, video, and sensor data. Developers can create spiking models or convert conventional neural networks into spiking forms for Akida deployment. The software can also support spiking-neural-network deployment on conventional processors.
Development hardware · 2021
Akida AI Processor Development Kits
These kits provide developers with hardware for evaluating BrainChip's Akida processor and building prototype edge-AI applications. They are intended to connect the company's processor architecture and MetaTF software with practical development and testing workflows.
AI accelerator board · 2022
Akida AI Processor PCIe boards
Akida PCIe boards provide a host-computer form factor for evaluating and developing applications with BrainChip's neuromorphic processor technology. BrainChip announced that it was taking orders for the boards in January 2022.
Marketing campaigns
- 2022University AI Accelerator Program
Academic and research markets
BrainChip's university-focused accelerator program sought to encourage academic experimentation and education around its AI and neuromorphic-computing technology. The Rochester Institute of Technology was added to the program in November 2022.
Outcome. The supplied material confirms the addition of Rochester Institute of Technology but does not quantify program participation or results.
Brand decisions
2023
Advance the Akida architecture to its second generationGeneration changeBackground
BrainChip sought to broaden the model types and embedded workloads supported by its neuromorphic platform.
Action taken
The second-generation Akida platform added 8-bit weights and activations, a Vision Transformer engine, and hardware support for Temporal Event-Based Neural Networks.
Aftermath
The changes expanded the platform's stated compatibility with contemporary computer-vision and temporal AI workloads.
2021
Commercialize Akida development hardwareProduct launchBackground
BrainChip needed to give developers practical access to its neuromorphic processor architecture.
Action taken
The company announced that it was taking orders for Akida AI Processor Development Kits.
Aftermath
The announcement expanded the platform beyond research and processor design toward external development and evaluation.
2015
Accept acquisition by AzianaM&ABackground
BrainChip was a private technology business when Australian mining company Aziana acquired it in March 2015.
Action taken
The acquisition placed BrainChip within Aziana's corporate structure and preceded the later public-market transaction.
Aftermath
BrainChip subsequently entered the ASX through a reverse merger involving the dormant Aziana entity in September 2015.
2015
Use a reverse merger to enter the Australian public marketM&ABackground
Following Aziana's acquisition of BrainChip, the company required a listed structure through which its technology could be commercialized and financed.
Action taken
BrainChip was placed on the Australian Securities Exchange through a reverse merger with the dormant Aziana entity.
Aftermath
The transaction established BrainChip as a publicly traded Australian technology company.
Recent events
- 2023
Edge Impulse adds support for BrainChip's Akida platform[1]
Other
Edge Impulse announced support for BrainChip's Akida processor, extending the platform's connection with embedded machine-learning development workflows.
- 2023
BrainChip completes Akida 1500 reference-chip design[1]
Product generation
BrainChip announced completion of the design of its Akida 1500 reference chip, marking a development step for the next stage of its neuromorphic processor family.
- 2023
BrainChip introduces second-generation Akida platform[1]
Product generation · Product launch
The second-generation platform added support for 8-bit weights and activations, a Vision Transformer engine, and hardware support for Temporal Event-Based Neural Networks.
- 2023
Akida processor family integrates with Arm Cortex-M85[1]
Other
BrainChip announced integration between the Akida processor family and Arm's Cortex-M85 processor, linking its neuromorphic accelerator technology with a widely used embedded CPU architecture.
- 2022
BrainChip opens orders for Akida AI Processor PCIe boards[1]
Product launch
The company announced availability for orders of PCIe boards intended to provide developers with a computer-hosted platform for evaluating and developing applications with Akida processing.
- 2022
BrainChip partners with NVISO[1]
Other
BrainChip announced a collaboration with NVISO involving applications and technologies connected with AI and neuromorphic processing.
- 2022
Rochester Institute of Technology joins BrainChip's University AI Accelerator Program[1]
Other
BrainChip added the Rochester Institute of Technology to a university-focused program intended to encourage research, education, and experimentation using its AI technology.
- 2022
BrainChip joins Intel Foundry Services technology ecosystem[1]
Other
BrainChip became associated with Intel Foundry Services as part of a technology-partner ecosystem involving semiconductor manufacturing and advanced AI hardware.
- 2021
BrainChip begins taking orders for Akida AI Processor Development Kits[1]
Product launch
BrainChip announced that orders had opened for development kits based on its Akida AI processor, supporting evaluation and application development for its neuromorphic edge-AI technology.
Sources5 sources
- 1BrainChipen.wikipedia.org
- 2BrainChip Appoints Geoffrey Carrick as Non-Executive Directoribtimes.com.au
- 3BrainChip Inc Appoints Vice President of Worldwide Salesibtimes.com.au
- 4BrainChip Launches AKD1500 Raspberry Pi 5 Kit: Giving Developers a Direct Path from Prototype to Productiondesign-reuse.com
- 5BrainChip Unveils AKD1500 SPI Chip, Bringing Neuromorphic AI to Everyday Edge Devicesdesign-reuse.com
Cite this profile: Cite the canonical profile. /brand-wiki/brainchip · Editorial policy · How profiles are compiled