DEEPX and Hyundai to Develop Generative AI Robots Using Low-Power Chips
South Korean AI chip startup DEEPX is expanding its collaboration with Hyundai Motor Group to develop a computing platform for generative AI robots. The platform will use DEEPX’s second-generation DX-M2 chips, designed for low power consumption and optimized for on-device AI processing.
The DX-M2 chips are scheduled for mass production next year using Samsung Electronics’ 2-nanometer manufacturing process. DEEPX stated that the chips are 20 times more power-efficient and significantly less expensive than comparable products, such as Nvidia’s Jetson Orin. The new chips are intended to help prevent overheating in energy-intensive humanoid robots.
Hyundai’s upcoming robotics platform will integrate the DX-M2 chips, following the company’s earlier use of DEEPX’s processors in its four-wheeled delivery robots. Hyundai unveiled its Atlas humanoid robot in January and plans to build a factory capable of producing 30,000 robot units annually by 2028.
DEEPX is also in discussions with investors and the government to raise over 600 billion won (around $408 million) as part of an ongoing funding round ahead of a potential South Korean IPO. The company reported that it aims to generate $40 million in revenue this year and may consider a future U.S. listing through American Depositary Receipts.
We hope you enjoyed this article
Consider subscribing to one of our newsletters like AI Funding Brief, Robotics Brief or Daily AI Brief.
Also, consider following us on social media:
More from Funding
Oct 16 Goodweek Raises $12 Million for AI Management Platform Oct 16 Intuit Agrees to Acquire AI Billing Startup Sequence for Reported $200 Million Oct 16 Matters Raises $5 Million to Build AI Consumer Bank Oct 16 KP Labs raises €4.42 million for AI data processing in orbit Oct 15 Origin Raises $20 Million for Autonomous Construction RobotsMore from Robotics
Oct 16 Rareview and Tuttle Launch Humanoid Robotics ETF HOID Oct 16 Element to Manage Waymo's Toronto Fleet for 2028 Launch Oct 16 RoboParty launches open source RP1 humanoid robot Oct 16 Jodell Robotics Unveils Dexterous Hand and Tactile Sensors Oct 16 Saab Completes Sea Trials of Large Uncrewed Underwater VehicleSilicon Brief
Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.
Whitepaper
Tensordyne Napier: What If One Rack Could Do the Work of Nine?
Tensordyne
This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.
Read moreYou may also like
DexForce Uses Synthetic Data to Train Humanoid Robots
D-Robotics Raises $400 Million for Robotics Chips and Software
XPENG Opens AI and Vehicle Technology to Global Partners
SEMIFIVE and Mobilint Sign Robotics AI Chip Contract
POSCO DX Invests 2 Billion Won in Enhans to Build Task Executing AI Agents
Daily AI Brief: the AI news that matters, in your inbox.