Efficiently Process Large RM Datasets In Underlying Memory Pool, Disaggregated Over CXL (KAIST)


A technical paper titled "Failure Tolerant Training with Persistent Memory Disaggregation over CXL" was published (preprint) by researchers at KAIST and Panmnesia. "TRAININGCXL can efficiently process large-scale recommendation datasets in the pool of disaggregated memory while making training fault tolerant with low overhead," states the paper. Find the technical paper here. or here (IEE... » read more

CXL Picks Up Steam In Data Centers


CXL is gaining traction inside large data centers as a way of boosting utilization of different compute elements, such as memories and accelerators, while minimizing the need for additional racks of servers. But the standard is being extended and modified so quickly that it is difficult to keep up with all the changes, each of which needs to be verified and validated across a growing swath of h... » read more

What Does 2023 Have In Store For Chip Design?


Predictions seem to be easier to make during times of stability, but they are no more correct than at any other period. During more turbulent times, fewer people are courageous enough to allow their opinions to be heard. And yet it is often those views that are more well thought through, and even if they turn out not to be true, they often contain some very enlightening ideas. 2022 saw some ... » read more

Adapting To Broad Shifts Essential In 2022


Change creates opportunity, but not every company is able to respond quickly enough to take advantage of those opportunities. Others may respond too quickly, before they properly understand the implications. At the start of a typical year, optimism is in plentiful supply. Any positive trend is seen as continuing, and any negative is seen as turning around. Normally the later in the year that... » read more

Is UCIe Really Universal?


Chiplets are rapidly becoming the means to overcome the slowing of Moore's Law, but whether one interface is capable of joining them all together isn't clear yet. The Universal Chiplet Interconnect Express (UCIe) believes it will work, but some in the industry remain unconvinced. At least part of the problem is that interconnect standards are never truly finished. Even today, the protocols tha... » read more

CXL 3.0: From Expansion To Scaling


At the Flash Memory Summit in August, the CXL Consortium released the latest, and highly anticipated, version 3.0 of the Compute Express Link (CXL) specification. This new version of the specification builds on previous generations and introduces several compelling new features that promise to increase data center performance and scalability, while reducing the total cost of ownership (TCO). ... » read more

How Memory Design Optimizes System Performance


Exponential increases in data and demand for improved performance to process that data has spawned a variety of new approaches to processor design and packaging, but it also is driving big changes on the memory side. While the underlying technology still looks very familiar, the real shift is in the way those memories are connected to processing elements and various components within a syste... » read more

Week In Review: Auto, Security, Pervasive Computing


The great EV ramp EV-related developments are everywhere. California’s move to ban sales of new internal-combustion vehicles by 2035, and the U.S. government’s sweeping embrace of clean-energy, are in lockstep with recent moves by the auto industry and related supply chains, as well as cutting-edge research. One of the big breakthroughs is the ability to charge an EV in 10 minutes witho... » read more

Week In Review: Design, Low Power


Quantum computing Baidu introduced a 10-qubit quantum computer called Qianshi and what it described as “the world's first all-platform quantum hardware-software integration solution that provides access to various quantum chips via mobile app, PC, and cloud.” The company said it has also completed the design of a 36-qubit quantum chip. Scientists said “levitating” nanoparticles co... » read more

Finding the Scope of CXL-Enabled Tiered Memory System in Production


This new technical paper titled "TPP: Transparent Page Placement for CXL-Enabled Tiered Memory" is presented by researchers at University of Michigan and Meta Inc. Abstract (partial) "We propose a novel OS-level application-transparent page placement mechanism (TPP) for efficient memory management. TPP employs a lightweight mechanism to identify and place hot and cold pages to appropriate... » read more

← Older posts Newer posts →