Invited: The Magnificent Seven Challenges and Opportunities in Domain-Specific Accelerator Design for Autonomous Systems
Article No.: 367, Pages 1 - 4
Abstract
The end of Moore's Law and Dennard Scaling has combined with advances in agile hardware design to foster a golden age of domain-specific acceleration. However, this new frontier of computing opportunities is not without pitfalls. As computer architects approach unfamiliar domains, we have seen common themes emerge in the challenges that can hinder progress in the development of useful acceleration. In this work, we present the Magnificent Seven Challenges in domain-specific accelerator design that can guide adventurous architects to contribute meaningfully to novel application domains. Although these challenges appear across domains ranging from ML to genomics, we examine them through the lens of autonomous systems as a motivating example in this work. To that end, we identify opportunities for the path forward in a successful domain-specific accelerator design from these challenges.
References
[1]
K. Asanovic, et al., "The landscape of parallel computing research: A view from berkeley," University of California Technical Report, 2006.
[2]
H. Esmaeilzadeh, et al., "Dark silicon and the end of multicore scaling," in ISCA, 2011.
[3]
M. B. Taylor, "Is dark silicon useful? harnessing the four horsemen of the coming dark silicon apocalypse," in DAC, 2012.
[4]
J. L. Hennessy and D. A. Patterson, "A new golden age for computer architecture," CACM, 2019.
[5]
W. J. Dally, et al., "Domain-specific hardware accelerators," CACM, 2020.
[6]
S. K. Esser, et al., "Backpropagation for energy-efficient neuromorphic computing," NeurIPS, 2015.
[7]
Y.-H. Chen, et al., "Eyeriss: An energy-efficient reconfigurable accelerator for deep convolutional neural networks," JSSC, 2016.
[8]
Y. Chen, et al., "Diannao family: energy-efficient hardware accelerators for machine learning," CACM, 2016.
[9]
A. Parashar, et al., "Scnn: An accelerator for compressed-sparse convolutional neural networks," SIGARCH News, 2017.
[10]
N. P. Jouppi, et al., "In-datacenter performance analysis of a tensor processing unit," in ISCA, 2017.
[11]
A. Ankit, et al., "Puma: A programmable ultra-efficient memristor-based accelerator for machine learning inference," in ASPLOS, 2019.
[12]
A. Gondimalla, et al., "Sparten: A sparse tensor accelerator for convolutional neural networks," in MICRO, 2019.
[13]
E. Qin, et al., "Sigma: A sparse and irregular gemm accelerator with flexible interconnects for dnn training," in HPCA, 2020.
[14]
S. Murray, et al., "The microarchitecture of a real-time robot motion planning accelerator," in MICRO, 2016.
[15]
J. Sacks, et al., "Robox: an end-to-end solution to accelerate autonomous control in robotics," in ISCA, 2018.
[16]
A. Suleiman, et al., "Navion: A 2-mw fully integrated real-time visual-inertial odometry accelerator for autonomous navigation of nano drones," JSSC, 2019.
[17]
B. Asgari, et al., "Pisces: power-aware implementation of slam by customizing efficient sparse algebra," in DAC, 2020.
[18]
Y. Liu, et al., "Hardware acceleration of monte-carlo sampling for energy efficient robust robot manipulation," in FPL, 2020.
[19]
Y. Han, et al., "Dadu series-fast and efficient robot accelerators," in ICCAD, 2020.
[20]
W. Liu, et al., "Archytas: A framework for synthesizing and dynamically optimizing accelerators for robotic localization," in MICRO, 2021.
[21]
S. M. Neuman, et al., "Robomorphic computing: a design methodology for domain-specific accelerators parameterized by robot morphology," in ASPLOS, 2021.
[22]
M. Bakhshalipour, et al., "Racod: Algorithm/hardware co-design for mobile robot path planning," in ISCA, 2022.
[23]
S. M. Neuman, et al., "Roboshape: Using topology patterns to scalably and flexibly deploy accelerators across robots," in ISCA, 2023.
[24]
Y. Turakhia, et al., "Darwin: A genomics co-processor provides up to 15,000 x acceleration on long read assembly," ACM SIGPLAN Notices, 2018.
[25]
M. Doblas, et al., "Gmx: Instruction set extensions for fast, scalable, and efficient genome sequence alignment," in MICRO, 2023.
[26]
D. S. Cali, et al., "Segram: A universal hardware accelerator for genomic sequence-to-graph and sequence-to-sequence mapping," in ISCA, 2022.
[27]
J. A. Placed, et al., "A survey on active simultaneous localization and mapping: State of the art and new frontiers," TRO, 2023.
[28]
Z. Zhang, et al., "Visual-inertial odometry on chip: An algorithm-and-hardware co-design approach," in RSS, 2017.
[29]
S. Murray, et al., "Robot motion planning on a chip." in RSS, 2016.
[30]
"Realtime Robotics." [Online]. Available: https://rtr.ai/
[31]
B. Sundaralingam, et al., "Curobo: Parallelized collision-free robot motion generation," in ICRA, 2023.
[32]
M. Quigley, et al., "Ros: an open-source robot operating system," in ICRA OSS Workshop, 2009.
[33]
W. Thomason, et al., "Motions in microseconds via vectorized sampling-based planning," arXiv, 2023.
[34]
I. A. Sucan, et al., "The open motion planning library," RAM, 2012.
[35]
V. Sze, et al., "How to evaluate deep neural network processors: Tops/w (alone) considered harmful," SSC-M, 2020.
[36]
M. Moll, et al., "Hyperplan: A framework for motion planning algorithm selection and parameter optimization," in IROS, 2021.
[37]
S. Dave, et al., "Hardware acceleration of sparse and irregular tensor computations of ml models: A survey and insights," Proceedings of the IEEE, 2021.
[38]
S. Krishnan, et al., "Automatic domain-specific soc design for autonomous unmanned aerial vehicles," in MICRO, 2022.
[39]
B. Plancher, et al., "Grid: Gpu-accelerated rigid body dynamics with analytical gradients," in ICRA, 2022.
[40]
R. Hadidi, et al., "Quantifying the design-space tradeoffs in autonomous drones," in ASPLOS, 2021.
[41]
S. M. Neuman, et al., "Tiny robot learning: Challenges and directions for machine learning in resource-constrained robots," in AICAS, 2022.
[42]
B. Boroujerdian, et al., "Mavbench: Micro aerial vehicle benchmarking," in MICRO, 2018.
[43]
D. Nikiforov, et al., "Rosé: A hardware-software co-simulation infrastructure enabling pre-silicon full-stack robotics soc evaluation," in ISCA, 2023.
[44]
M. Huzaifa, et al., "Illixr: Enabling end-to-end extended reality research," in IISWC, 2021.
[45]
S. Sudhakar, et al., "Data centers on wheels: emissions from computing onboard autonomous vehicles," IEEE Micro, 2022.
[46]
D. Patterson, et al., "Energy and emissions of machine learning on smartphones vs. the cloud," CACM, 2024.
[47]
S. Eldridge, et al., "Mlir as hardware compiler infrastructure," in Workshop on Open-Source EDA Technology (WOSET), 2021.
[48]
R. Nigam, et al., "A compiler infrastructure for accelerator generators," in ASPLOS, 2021.
[49]
F. Winterstein, et al., "High-level synthesis of dynamic data structures: A case study using vivado hls," in FPT, 2013.
[50]
R. Nikhil, "Bluespec system verilog: efficient, correct rtl from high level specifications," in MEMOCODE, 2004.
[51]
V. J. Reddi, et al., "Mlperf inference benchmark," in ISCA, 2020, pp. 446--459.
[52]
U. Gupta, et al., "Architectural co2 footprint tool: Designing sustainable computer systems with an architectural carbon modeling tool," IEEE Micro, 2023.
[53]
C. C. Sudarshan, et al., "Greenfpga: Evaluating fpgas as environmentally sustainable computing solutions," HPCA, 2024.
[54]
P. Vlaskovits, "Henry ford, innovation, and that "faster horse" quote," Harvard Business Review, 2011.
Index Terms
- Invited: The Magnificent Seven Challenges and Opportunities in Domain-Specific Accelerator Design for Autonomous Systems
Recommendations
Software architecture challenges and opportunities for the 21st century: dynamism, mobility, autonomy, services, grids, and clouds
ISEC '12: Proceedings of the 5th India Software Engineering ConferenceDevelopments in the area of software architecture over the past two decades have pushed architecture to the forefront of a number of critical software engineering activities: modeling, design, analysis, simulation, implementation, deployment, ...
Reconfigurable Coprocessor for Multimedia Application Domain
A new reconfigurable architectural template is presented. Such a template is composed of coarse-grained and fine-grained reconfigurable datapath and control to obtain performances at custom designed chip level. To show the adaptability/performance of ...
Domain-Specific Language for HW/SW Co-design for FPGAs
DSL '09: Proceedings of the IFIP TC 2 Working Conference on Domain-Specific LanguagesThis article describes FSMLanguage, a domain-specific language for HW/SW co-design targeting platform FPGAs. Modern platform FPGAs provide a wealth of configurable logic in addition to embedded processors, distributed RAM blocks, and DSP slices in order ...
Comments
Please enable JavaScript to view thecomments powered by Disqus.Information & Contributors
Information
Published In
Copyright © 2024 Copyright is held by the owner/author(s). Publication rights licensed to ACM.
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].
Sponsors
In-Cooperation
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 07 November 2024
Check for updates
Author Tags
Qualifiers
- Research-article
Conference
DAC '24
Sponsor:
Acceptance Rates
Overall Acceptance Rate 1,770 of 5,499 submissions, 32%
Upcoming Conference
DAC '25
- Sponsor:
- sigda
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 41Total Downloads
- Downloads (Last 12 months)41
- Downloads (Last 6 weeks)41
Reflects downloads up to 12 Nov 2024
Other Metrics
Citations
View Options
Get Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in