10-Year Roadmap for AI + Hardware (UIUC, UCLA, Stanford et al.)


Researchers from University of Illinois Urbana-Champaign, UCLA, Stanford University, Nvidia, Google, et al. have released “AI+HW 2035: Shaping the Next Decade”. Abstract “Artificial intelligence (AI) and hardware (HW) are advancing at unprecedented rates, yet their trajectories have become inseparably intertwined. The global research community lacks a cohesive, long-term vision t... » read more

LtRAM And StRAM: Specialized Memory Architectures Leveraging Workload-Specific Access Characteristics (Stanford, Microsoft)


A new technical paper titled "Towards Memory Specialization: A Case for Long-Term and Short-Term RAM" was published by researchers at Stanford University and Microsoft, and an independent researcher. Abstract "Both SRAM and DRAM have stopped scaling: there is no technical roadmap to reduce their cost (per byte/GB). As a result, memory now dominates system cost. This paper argues for a parad... » read more

Report: The Future of AI Processing


AI is now emerging in everyday use cases thanks to advances in foundational models, powerful chip technology, and abundant data. As a result, new approaches in AI compute are required to deploy these advanced use cases with minimal effort. This report features insights on how to move AI forward from industry leaders at AWS, Meta, Samsung and Arm, and includes research from MIT Technology Rev... » 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

Enhancing Compute Security Architecture For New-Age Applications


New-age AI-powered applications are becoming increasingly essential in our daily lives. Continuing to do so requires that these applications and services meet three primary challenges: Achieving high performance for complex compute tasks. Ensuring cost-effectiveness and seamless integration with existing infrastructure. Maintaining robust security and privacy measures. Historicall... » read more

An Energy Efficient, Linux-Capable RISC-V Host Platform Designed For The Seamless Plug-In And Control Of Domain-Specific Accelerators


A technical paper titled “Cheshire: A Lightweight, Linux-Capable RISC-V Host Platform for Domain-Specific Accelerator Plug-In” was published by researchers at ETH Zurich and University of Bologna. Abstract: "Power and cost constraints in the internet-of-things (IoT) extreme-edge and TinyML domains, coupled with increasing performance requirements, motivate a trend toward heterogeneous arc... » read more

Heterogenous Computing & Cache Attacks


Researchers at imec-COSIC, KU Leuven presented this paper titled "Double Trouble: Combined Heterogeneous Attacks on Non-Inclusive Cache Hierarchies" at the USENIX Security Symposium in Boston in August 2022. Note, this is a prepublication paper. Abstract: "As the performance of general-purpose processors faces diminishing improvements, computing systems are increasingly equipped with domai... » read more

Audio, Visual Advances Intensify IC Design Tradeoffs


A spike in the number of audio and visual sensors is greatly increasing design complexity in chips and systems, forcing engineers to make tradeoffs that can affect performance, power, and cost. Collectively, these sensors generate so much data that designers must consider where to process different data, how to prioritize it, and how to optimize it for specific applications. The tradeoffs in... » read more

Dynamically Reconfiguring Logic


Dynamic reconfiguration of semiconductor logic has been possible for years, but it never caught on commercially. Cheng Wang, co-founder and senior vice president of software and engineering at Flex Logix, explains why this capability has been so difficult to utilize, what’s changed, how a soft logic layer can be used to control when to read, compute, steer, and write data back to memory, and ... » read more

ACAP At The Edge With The Versal AI Edge Series


This white paper introduces the AI Edge series to the Versal ACAP portfolio, a domain-specific architecture (DSA) that meets the strenuous demands of systems implemented in the 7nm silicon process. This series is optimized to meet the performance-per-watt requirements of edge nodes at or near the analog-digital boundary. Here, immediate response to the physical world is highly valued, and in ma... » read more

← Older posts