Latest news with #HussamAmrouch
Business Mayor
21-05-2025
- Business
- Business Mayor
AI chip developed for decentralized use without the cloud
A new AI chip developed at the Technical University of Munich (TUM) works without the cloud server or internet connections needed by existing chips. The AI Pro, designed by Prof Hussam Amrouch, is modelled on the human brain. Its innovative neuromorphic architecture enables it to perform calculations on the spot, ensuring full cyber security. It is also up to ten times more energy efficient. The professor of AI processor design at TUM has already had the first prototypes produced by semiconductor manufacturer Global Foundries in Dresden. Unlike conventional chips, the computing and memory units of the AI Pro are located together. This is possible because the chip applies the principle of 'hyperdimensional computing': This means that it recognizes similarities and patterns, but does not require millions of data records to learn. Instead of being shown countless images of cars, as with the deep learning method used in conventional AI chips, this chip combines various pieces of information, such as the fact that a car has four wheels, usually drives on the road, and can have different shapes. Like the new chip, explains Prof. Amrouch, 'humans also draw inferences and learn through similarities.' An important advantage of brain-like thinking: it saves energy. For the training of a sample task, the new chip consumed 24 microjoules, while comparable chips required ten to a hundred times more energy — 'a record value,' notes Prof. Amrouch. 'This mix of modern processor architecture, algorithm specialization and innovative data processing makes the AI chip something special.' This also sets it apart from all-rounders like the chips from industry giant NVIDIA. 'While NVIDIA has built a platform that relies on cloud data and promises to solve every problem, we have developed an AI chip that enables customized solutions. There is a huge market there.' Read More Democrats and Republicans condemn espionage-driven Chinese hack Neuromorphic chips: Modelled on the human brain The one square millimeter chip currently costs 30,000 euros. With around 10 million transistors it is not quite as densely packed or as powerful as NVIDIA chips with 200 billion transistors. But that is not Prof. Amrouch's primary concern. His team specializes in AI chips that perform the processing directly on site instead of having to send the data to the cloud to be processed along with millions of other data sets before being sent back again. This saves time and server computing capacity and reduces the carbon footprint of AI. The chips are also customized for specific applications. 'That makes them very efficient,' says chip expert Amrouch. For example, they focus on processing heartrate and other vital data collected via smartwatch or navigation data of a drone. Because this personal and sometimes sensitive data remains on board the device, issues with stable internet connections or cybersecurity do not even arise. The chip expert is convinced: 'The future belongs to the people who own the hardware.'
Forbes
20-05-2025
- Science
- Forbes
Brain-Inspired AI Chip Enables Energy-Efficient Off-Grid Processing
AI Processor In an advancement for efficient and secure AI, researchers at the Technical University of Munich (TUM) have unveiled the AI Pro chip, a neuromorphic processor that functions independently of cloud servers and without internet connections. Designed by Professor Hussam Amrouch, this brain-inspired chip performs on-device computations, enhancing cybersecurity and energy efficiency. The AI Pro's architecture emulates the human brain, integrating computing and memory units to process data locally. This design eliminates the need for data transmission to external servers, reducing latency and potential security vulnerabilities. By employing hyperdimensional computing, the chip recognizes patterns with minimal data, streamlining the learning process. Neuromorphic processors are computer chips designed to mimic the structure and function of the human brain. Unlike traditional processors, which separate memory and processing units, neuromorphic chips integrate them, similar to neurons and synapses in the brain. This allows them to process information more efficiently, especially for tasks like pattern recognition, learning from small data sets, and operating with very low power. Professor Amrouch explains, "humans draw inferences and learn through similarities" in much the same way as the AI Pro chip. This approach allows the AI Pro to function effectively with fewer training examples, making it suitable for applications where data availability is limited. The AI Pro demonstrates remarkable energy efficiency, consuming just 24 microjoules for specific tasks, up to ten times less than comparable chips. This efficiency is crucial for enabling powerful AI capabilities on battery-powered devices and scenarios where power resources are constrained. By processing data on-device, the AI Pro enhances cybersecurity. Sensitive information remains within the device, mitigating risks associated with data transmission and storage in external servers. This feature is particularly beneficial for applications in healthcare, environmental monitoring, and autonomous systems. 'While Nvidia has built a platform that relies on cloud data and promises to solve every problem, we have developed an AI chip that enables customized solutions. There is a huge market there,' explains Prof. Amrouch. The AI Pro's capabilities align with the growing demand for edge computing solutions. Its ability to operate without internet connectivity makes it ideal for remote or mobile applications, such as drones, wearable health monitors, and IoT devices. Compared to general-purpose GPU chips like those from Nvidia, which rely heavily on cloud-based processing, the AI Pro offers a specialized solution focused on efficiency and security. While it contains around 10 million transistors, significantly fewer than the over 200 billion on Nvidia's latest Blackwell B200 GPU, the AI Pro's design prioritizes targeted performance over broad applicability. However this power and efficiency comes at a price, with the one square millimeter chip currently priced at 30,000 euros. The AI Pro has progressed beyond the conceptual stage, with prototypes manufactured by Global Foundries in Dresden. This development indicates the chip's readiness for integration into commercial products. As industries seek to enhance data security and reduce energy consumption, the AI Pro represents a promising advancement in AI hardware. Its design philosophy underscores a shift towards localized, efficient processing, potentially setting a new standard for future AI applications. With the increasing capability and power of AI-specific processors, Prof. Amrouch is convinced that 'the future belongs to the people who own the hardware.'
