Artificial intelligence (AI) applications are spreading to more industries every day. However, the amount of energy used by these AI systems has become a significant issue. Modern deep neural networks require a considerable amount of computing power. This article examines five key hardware strategies for building energy-efficient AI acceleration: dedicated accelerator architectures, analog in-memory computing, […]
Artificial intelligence/ML
RAID storage system supports PCIe Gen 5 and Gen 4 connectivity
HighPoint Technologies introduces its next-generation NVMe RAID storage solutions, engineered to meet the demanding data ingestion and processing requirements of AI, autonomous systems, and professional media applications. Combining industry-leading PCIe switching architecture with advanced RAID technologies, these solutions deliver unparalleled speed, reliability, and scalability for high-volume data capture, real-time analytics, and long-term storage. HighPoint’s proprietary […]
Embedded compute modules enable Windows with ARM architecture transition for industrial automation
Tria has introduced a family of embedded compute modules featuring Qualcomm Dragonwing processors with multi-operating system support. The modules support Android, Windows 11 IoT Enterprise, and Yocto Linux, enabling deployment across industrial, medical, agriculture, and construction sectors where edge computing, machine learning, and AI capabilities provide operational advantages. The modules utilize Qualcomm Dragonwing processors designed […]
What is ‘compute-in-memory’ and why is it important for AI?
If you are running AI workloads, here is something that might surprise you. Your processors are wasting more energy shuffling data around than actually doing the calculations you care about. This inefficiency is becoming a serious limit for the next generation of artificial intelligence systems. As neural networks grow to billions of parameters, traditional von […]
AI-accelerated MCU combines voice and vision processing capabilities
Renesas Electronics announced the RA8P1, an AI-accelerated microcontroller designed for AIoT (Artificial Intelligence of Things) applications. The device combines ARM Cortex-M85 and Cortex-M33 cores with an Ethos-U55 Neural Processing Unit (NPU), integrating AI processing capabilities directly into a microcontroller form factor. The RA8P1 offers up to 1 GHz of CPU performance and 256 GOPS of […]
SRAM recovers 15% silicon area on 2 nm process with 6 gigabit memory capacity
Marvell Technology, Inc. expanded its custom technology platform with the launch of the industry’s first 2nm custom Static Random Access Memory (SRAM), designed to boost the performance of custom XPUs and devices powering cloud data centers and AI clusters. Combining advanced custom circuitry and software from Marvell with core SRAM and cutting-edge 2nm process technology, Marvell […]
How do AI agents and model context protocol work together?
An AI agent, combined with a model context protocol (MCP), can create more contextual AI systems that can perform a wide range of tasks, from automating processes to providing insights and interacting with users. MCP helps AI agents connect to data sources, enabling them to access information, collaborate with other systems, and make decisions based […]
Linear redrivers support PCIe 6 with PAM4 modulation
Diodes Incorporated introduces the industry’s first ReDriver to reach PCI Express (PCIe) 6.0 protocol speeds (up to 64 GT/s) while being backward-compatible with PCIe 5.0/4.0/3.0 protocols. The PI3EQX64904 is a low-power, high-performance, 64GT/s, four-differential-channel, PAM4, linear ReDriver. Target applications include AI data centers (storage and servers), workstations, 5G networking, CPU-to-network (PCIe NIC card) and CPU-to-storage (NVME) […]
MCUs integrate CAN FD controller and 3-phase PWM with AI accelerator
ROHM Semiconductor has announced the development of AI-equipped microcontrollers designated ML63Q253x-NNNxx and ML63Q255x-NNNxx that perform fault prediction and degradation forecasting using sensor data in industrial equipment including motors. These MCUs execute both learning and inference operations without requiring network connectivity, operating as standalone devices for real-time monitoring applications. The MCUs implement ROHM’s Solist-AI technology using […]
Multi-gigahertz clocks boast LVDS, CML, LVPECL output options
Mixed Signal Devices has announced its MS11X0 oscillator and MS15X0 jitter attenuator families for computing and communications systems. The timing devices are built on 28 nm CMOS technology and incorporate the company’s Virtual Crystal architecture to address requirements in AI data centers, 5G base stations, and radar applications. The MS1130 and MS1150 oscillators operate at […]
