BlitzCoin: Fully Decentralized Hardware Power Management for Accelerator-Rich SoCs
Martin Cochet, Karthik Swaminathan, Erik Jens Loscalzo, Joseph Zuckerman, Maico Cassel dos Santos, Davide Giri, Alper Buyuktosunoglu, Tianyu Jia, David Brooks, Gu-Yeon Wei, Kenneth L. Shepard, Luca P. Carloni, Pradip Bose
摘要
On-chip power-management techniques have evolved over several processor generations. However, response time and scalability constraints have made it difficult to translate existing power-management strategies to current or next-generation System-on-Chip (SoC) architectures, which are expected to comprise tens to hundreds of cores and accelerators. In this work we present BlitzCoin, a fully decentralized hardware power-management strategy for large, accelerator-rich SoCs, coupled with optimized unified voltage and frequency regulation. We evaluated BlitzCoin through RTL simulations of multiple SoCs targeted toward different application domains. The results are further validated through silicon measurements of a fabricated 12 nm many-accelerator SoC that includes BlitzCoin. Our evaluations show that BlitzCoin is markedly faster, with 8× to 12× lower response times, which provides 25%-34% throughput improvement and allows for scaling to 7 × to 13 × larger SoCs compared to state-of-the-art centralized power-management strategies, all with an area overhead of <1%.
问问这篇 Paper
智能体会读完全文。
Lune 把这篇 Paper 索引到了每一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。
它引用的顶会 Paper4
- SysScale: Exploiting Multi-domain Dynamic Voltage and Frequency Scaling for Energy Efficient Mobile ProcessorsJawad Haj-Yahya, Mohammed Alser, Jeremie S. Kim, Abdullah Giray Yaglikçi 等ISCA 2020 · 被引用 48 次
- Cohmeleon: Learning-Based Orchestration of Accelerator Coherence in Heterogeneous SoCsJoseph Zuckerman, Davide Giri, Jihye Kwon, Paolo Mantovani 等MICRO 2021 · 被引用 24 次
- AgileWatts: An Energy-Efficient CPU Core Idle-State Architecture for Latency-Sensitive Server ApplicationsJawad Haj-Yahya, Haris Volos, Davide B. Bartolini, Georgia Antoniou 等MICRO 2022 · 被引用 22 次
- AgilePkgC: An Agile System Idle State Architecture for Energy Proportional Datacenter ServersGeorgia Antoniou, Haris Volos, Davide B. Bartolini, Tom Rollet 等MICRO 2022 · 被引用 6 次
相关 Paper
- Area Bloating and the Future of SpecializationQixuan Yu, David WentzlaffHPCA 2026
- Centralized Training and Decentralized Control through the Actor-Critic Paradigm for Highly Optimized MulticoresBenedikt Dietrich, Heba Khdr, Jörg HenkelDAC 2025 · 被引用 2 次
- PROPHET: Predictive On-Chip Power Meter in Hardware Accelerator for DNNJian Peng, Tingyuan Liang, Zhiyao Xie, Wei ZhangDAC 2023 · 被引用 1 次
- ScalaGraph: A Scalable Accelerator for Massively Parallel Graph ProcessingPengcheng Yao, Long Zheng, Yu Huang, Qinggang Wang 等HPCA 2022 · 被引用 30 次
- NCPU: An Embedded Neural CPU Architecture on Resource-Constrained Low Power Devices for Real-time End-to-End PerformanceTianyu Jia, Yuhao Ju, Russ Joseph, Jie GuMICRO 2020 · 被引用 21 次
