Direct-To-Chip Liquid-Cooled Data Centers (Binghamton, Nvidia)


A new technical paper titled "Parameters of performance: A deep dive into liquid-to-air CDU assessment" was published by researchers at Binghamton University-SUNY and NVIDIA. Abstract: "The rapid growth in data center workloads and the increasing complexity of modern applications have led to significant contradictions between computational performance and thermal management. Traditional air... » read more

Chip Industry Week In Review


Europe's top court ruled in Intel's favor, voiding a $1.1 billion fine imposed by the European Union and dismissing charges of anti-competitive behavior. IBM released yield benchmarks for high-NA EUV, which serve as proof points that the newest advanced litho equipment will enable scaling beyond the 2nm process node. Also on the lithography front, Nikon is developing a maskless digital litho... » read more

Chip Industry Week In Review


The U.S. Department of Commerce and Texas Instruments (TI) signed a non-binding preliminary memorandum of terms to provide up to $1.6 billion in CHIPS Act funding towards TI’s investment of over $18 billion for three 300mm semiconductor wafer fabs under construction in Texas and Utah. TI also expects to get about $6 billion to $8 billion from the U.S. Department of Treasury’s Investmen... » read more

Power Delivery Challenged By Data Center Architectures


Processor and data center architectures are changing in response to the higher voltage needs of servers running AI and large language models (LLMs). At one time, servers drew a few hundred watts for operation. But over the past few decades that has changed drastically due to a massive increase in the amount of data that needs to be processed and user demands to do it more quickly. NVIDIA's G... » read more

Chip Industry Week In Review


The University of Texas at Austin’s Texas Institute for Electronics (TIE) was awarded $840 million to establish a Department of Defense microelectronics manufacturing center. This center will focus on developing advanced semiconductor microsystems to enhance U.S. defense systems. The project is part of DARPA's NGMM Program. The U.S. Dept. of Commerce announced preliminary terms with Global... » read more

How Does Reclaiming Data Center Lost Capacity Result in Return on Investment?


As the importance of data center performance, efficiency, and sustainability grows, owners and operators must move beyond “quick fixes” when managing their data center and IT deployments. These temporary solutions are neither effective nor sustainable. Long-term solutions and effective management over time are required to achieve meaningful efficiency gains. Capacity planning is one such di... » read more

Data Center Evolution: The Leap to 64 GT/s Signaling with PCI Express 6.1


The PCI Express (PCIe) interface is the critical backbone that moves data at high bandwidth and low latency between various compute nodes such as CPUs, GPUs, FPGAs, and workload-specific accelerators. With the rapid rise in bandwidth demands of advanced workloads such as AI/ML training, PCIe 6.1 jumps signaling to 64 GT/s with some of the biggest changes yet in the standard. Find more inform... » read more

Sea Of Processors Use Case


Core counts have been increasing steadily since IBM's debut of the Power 4 in 2001, eclipsing 100 CPU cores and over 1,000 for AI accelerators. While sea of processor architectures feature a stamp and repeat design, per-core workloads aren't always going to be symmetrically balanced. For example, a cloud provider (AI or compute) will rent out individual core clusters to customers for specialize... » read more

Application-Specific Power Performance Optimizer Based On Chip Telemetry


As datacenter power consumption continues to pose cooling and cost challenges, and battery driven devices are expected to last longer between charges, the search for advanced power management mechanisms continues. A modern design must balance between maximizing performance, consuming the least amount of power, and guaranteeing no failures in field. The latter requires safety margins which tr... » read more

Building Scalable And Efficient Data Centers With CXL


The AI boom is giving rise to profound changes in the data center; demanding AI workloads are driving an unprecedented need for low latency, high-bandwidth connectivity and flexible access to more memory and compute power when needed. The Compute Express Link (CXL) interconnect offers new ways for data centers to enhance performance and efficiency between CPUs, accelerators and storage and move... » read more

← Older posts