Samtec now second-sources its Si-Fly HD co-packaged and near-chip systems through a licensing agreement with Molex. It’s not the first time the two companies have been second sources. Having a second source for critical components can save the day. That’s even more important today given supply-chain disruptions of raw materials. For basic devices such as […]
AI/ML
What type of interconnects and connectors link accelerator cards in AI data centers?
Many data centers are packed with racks of high-performance graphics processing units (GPUs) and tensor processing units (TPUs). These accelerators process massive artificial intelligence (AI) and machine learning (ML) datasets, executing complex operations in parallel and exchanging data at high speed. This article explores the interconnects and connectors that link AI accelerator clusters together. Scaling […]
What is the role of liquid cooling connectors in AI data centers?
Artificial intelligence (AI) and machine learning (ML) applications consume significant power and generate considerable heat in data centers. High-performance AI accelerators — such as graphics processing units (GPUs), tensor processing units (TPUs), and application-specific integrated circuits (ASICs) — increasingly require more efficient cooling methods to maintain safe and optimal thermal operating levels. This article discusses […]
How do the OCP immersion cooling requirements effect connectors?
The Open Compute Project (OCP) is focused on developing scalable computing for cloud data centers, artificial intelligence, and other large-scale computing applications. Immersion cooling can improve the energy efficiency and sustainability of those facilities. So, the interest of the OCP in immersion cooling is not surprising. The OCP immersion cooling requirements define materials compatibility needs […]
Intelligent interconnect optimizes AI data transmission with DSP-free design
TeraSignal introduced TSLink, the world’s first intelligent chip-to-module (C2M) interconnect designed to revolutionize data transmission between large ASICs and linear optical modules. Leveraging existing microcontroller resources in optical modules, the TSLink solution automates link training and performance monitoring without the need for additional digital signal processors (DSPs). TSLink dramatically reduces power consumption and latency while simplifying […]
How twin axial cable assemblies support high-performance computing for AI/ML systems
The high-performance computing platforms used for artificial intelligence (AI) and machine learning (ML) in hyperscale data centers need high-speed interconnects like 112 Gbps PAM 4 and faster inside the servers. High-speed interconnects are also required between the servers and storage devices. Twin axial (Twinax) cable assemblies are one way to address those needs. This article […]
What interconnects are used with memory for HPC and AI?
High-performance computing (HPC) for artificial intelligence (AI) applications places extreme demands on memory connectors. They must deliver high-speed connectivity, high levels of signal integrity (SI), and support high-density solutions. This article reviews connector needs for Peripheral Component Interconnect Express 6.0 (PCIe 6.0), Serial Attached SCSI (SAS) including SAS 3.0, 4.0 and SAS/PCIe 5.0, and Data […]
