The Best DRAMs For Artificial Intelligence


Artificial intelligence (AI) involves intense computing and tons of data. The computing may be performed by CPUs, GPUs, or dedicated accelerators, and while the data travels through DRAM on its way to the processor, the best DRAM type for this purpose depends on the type of system that is performing the training or inference. The memory challenge facing engineering teams today is how to keep... » read more

Chiplet-to-Chiplet Gateway Architecture, A C2C Interface Bridging Two Chiplet Protocols (Peter Grünberg, Jülich Supercomputing Centre)


A new technical paper titled "Modeling Chiplet-to-Chiplet (C2C) Communication for Chiplet-based Co-Design" was published by researchers at Peter Grünberg Institute and Jülich Supercomputing Centre. Abstract "Chiplet-based processor design, which combines small dies called chiplets to form a larger chip, enables scalable designs at economical costs. This trend has received high attention s... » read more

High-Speed Test IO: Addressing High-Performance Data Transmission And Testing Needs For HPC & AI


By Lakshmi Jain and Wei-Yu Ma The AI and HPC industries are rapidly shifting toward chiplet-based designs to achieve unprecedented levels of performance, as traditional monolithic system-on-chip (SoC) architectures face scaling limitations. This transition is fueled by the rise of heterogeneous integration, which is driving innovation across the semiconductor sector. However, this advancemen... » read more

Effects Of Hardware Prefetchers For Scientific Application Kernels Running on High-End Processors


A new technical paper titled "Memory Prefetching Evaluation of Scientific Applications on A Modern HPC Arm-based Processor" was published by researchers at Jülich Supercomputing Centre and KTH Royal Institute of Technology. Abstract "Memory prefetching is a well-known technique for mitigating the negative impact of memory access latencies on memory bandwidth. This problem has become more p... » read more

Big Changes Ahead For Interposers And Substrates


Interposers and substrates are undergoing a profound transformation from intermediaries to engineered platforms responsible for power distribution, thermal management, high-density interconnects, and signal integrity in the most advanced computing systems. This shift is being driven by AI, high-performance computing (HPC), and next-generation communications, where the need for heterogeneous ... » read more

Lines Blurring Between Supercomputing And HPC


Supercomputers and high-performance computers are becoming increasingly difficult to differentiate due to the proliferation of AI, which is driving huge performance increases in commercial and scientific applications and raising similar challenges for both. While the goals of supercomputing and high-performance computing (HPC) have always been similar — blazing fast processing — the mark... » read more

Bold Prediction: 50% Of New HPC Chip Designs Will Be Multi-Die In 2025


Monolithic chips have been the workhorses behind decades of technological advancement. But just as the industrial revolution saw workhorses replaced with more efficient and powerful machinery, the semiconductor industry is on the cusp of a similar revolution. Multi-die and chiplet-based designs — which integrate multiple specialized dies in a single package or stack integrated circuits ver... » read more

Signal Integrity Plays Increasingly Critical Role In Chiplet Design


Maintaining the quality and reliability of electrical signals as they travel through interconnects is proving to be much more challenging with chiplets and advanced packaging than in monolithic SoCs and PCBs. Signal integrity is a fundamental requirement for all chips and systems, but it becomes more difficult with chiplets due to reflections, loss, crosstalk, process variation, and various ... » read more

Uncore Frequency Scaling For Energy Optimization In Heterogeneous Systems (UIC, Argonne)


A new technical paper titled "Exploring Uncore Frequency Scaling for Heterogeneous Computing" was published by researchers at University of Illinois Chicago and Argonne National Laboratory. Abstract "High-performance computing (HPC) systems are essential for scientific discovery and engineering innovation. However, their growing power demands pose significant challenges, particularly as sys... » read more

Chiplets: Where Are We Today?


The 3rd annual Chiplet Summit was held in Santa Clara from January 21st to 23rd at the Convention Center. The conference continues to grow from its 1st year when it was held at the San Jose Doubletree Hotel (almost exactly 2 years ago). During his Chairman’s Welcome presentation, Chuck Sobey mentioned that there were 41 exhibitors at this year’s conference. Chuck was also the moderator f... » read more

← Older posts