Start up company Untetherer launches low-energy AI chip based on RISC-V technology
The startup Untetherer has launched an artificial intelligence (AI) chip aimed at supporting AI applications in automobiles, agricultural equipment, and other extreme situations.
The flagship AI chips produced by NVIDIA and AMD are designed to run in data centers connected to thousands or tens of thousands of other chips. This type of installation requires a significant amount of energy and additional equipment, typically used to build or train large language models that support applications such as ChatGPT.
Untetherer, headquartered in Toronto, Canada, stated that its chips are designed to run AI applications after model construction and can operate with higher efficiency.
Bob Beachler, Vice President of Product at Untetherer, stated that by 2027, the market for chips that only run AI applications (also known as "inference") will far surpass the training market. The company expects the size of the inference market to grow to $102 billion by 2027.
The application scope of reasoning technology is very broad, "said Bob Beachler.
Untether's chip, named 240 Slim, has excellent performance but much lower energy consumption. It is suitable for use in autonomous vehicle, automatic agricultural equipment and some data centers.
Mercedes Benz said that it is cooperating with Untether to develop its next generation autonomous vehicle.
The chips produced by Untech are developed around open-source RISC-V technology. RISC-V technology competes with Arm's design.
In a recent peer review benchmark test conducted by the non-profit organization MLCommons, Untetherer's chip performed well in testing applications designed for hardware.
Untetherer, headquartered in Toronto, Canada, stated that its chips are designed to run AI applications after model construction and can operate with higher efficiency.
Bob Beachler, Vice President of Product at Untetherer, stated that by 2027, the market for chips that only run AI applications (also known as "inference") will far surpass the training market. The company expects the size of the inference market to grow to $102 billion by 2027.
The application scope of reasoning technology is very broad, "said Bob Beachler.
Untether's chip, named 240 Slim, has excellent performance but much lower energy consumption. It is suitable for use in autonomous vehicle, automatic agricultural equipment and some data centers.
Mercedes Benz said that it is cooperating with Untether to develop its next generation autonomous vehicle.
The chips produced by Untech are developed around open-source RISC-V technology. RISC-V technology competes with Arm's design.
In a recent peer review benchmark test conducted by the non-profit organization MLCommons, Untetherer's chip performed well in testing applications designed for hardware.