Valve founder Gabe Newell has thrown his support behind a new startup, Starfish Technologies, which is focused on the development of brain-computer interfaces (BCI) for treating neurological disorders. This was reported by TechSpot.
Starfish Technologies is not connected to the gaming industry; rather, it is dedicated to medical applications. The company aims to introduce its first chips by the end of 2025. Developers are working on miniature, energy-efficient, and wireless implants capable of reading and stimulating activity in multiple areas of the brain simultaneously.
The technology targets disorders associated with disrupted neural connectivity, such as Parkinson's disease. Starfish is seeking partnerships with experts in wireless power transfer, neurointerfaces, and communication systems.
Previously, Valve explored brain implants in a gaming context, and interest in such technology is also demonstrated by Sony, Tencent, and Apple. However, Starfish intends to focus initially on medical use cases.
Starfish has already released the technical specifications of its chip. The device measures 2×4 mm and consumes only 1.1 mW of power during standard recording. It can record brain activity (spikes and local field potentials) from 16 channels simultaneously at a frequency of 18.75 kHz, using 32 electrodes. The chip also supports electrical stimulation, has built-in impedance monitoring, voltage measurement during stimulation, and digital signal processing, enabling it to operate through low-speed wireless interfaces. The chip is manufactured using TSMC's 55 nm process technology.
Interestingly, Starfish's main competitor is Neuralink, founded by Elon Musk. Neuralink is already testing its BCI systems on humans – the first patient received an implant in January 2024, and as of now, its functionality has been maintained, despite some disconnection of thread parts.
Neuralink uses a larger implant with 1024 electrodes, consuming about 6 mW of power and requiring periodic wireless charging. The company targets both medical and experimental applications of brain interfaces.