Abstract
Innovative processor architectures aim to play a critical role in future sustainment of performance improvements under severe limitations imposed by the end of Moore’s Law. The Reconfigurable Optical Computer (ROC) is one such innovative, Post-Moore’s Law processor. ROC is designed to solve partial differential equations in one shot as opposed to existing solutions, which are based on costly iterative computations. This is achieved by leveraging physical properties of a mesh of optical components that behave analogously to lumped electrical components. However, virtualization is required to combat shortfalls of the accelerator hardware. Namely, (1) the infeasibility of building large photonic arrays to accommodate arbitrarily large problems and (2) underutilization brought about by mismatches in problem and accelerator mesh sizes due to future advances in manufacturing technology. In this work, we introduce an architecture and methodology for lightweight virtualization of ROC that exploits advantages borne from optical computing technology. Specifically, we apply temporal and spatial virtualization to ROC and then extend the accelerator scheduling tradespace with the introduction of spectral virtualization. Additionally, we investigate multiple resource scheduling strategies for a system-on-chip (SoC)-based PDE acceleration architecture and show that virtual configuration management offers a speedup of approximately 2×. Finally, we show that overhead from virtualization is minimal, and our experimental results show two orders of magnitude increased speed as compared to microprocessor execution while keeping errors due to virtualization under 10%.
- [1] . 2007. Course materials for 6.002 Circuits and Electronics. Retrieved from http://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-002-circuits-and-electronics-spring-2007/video-lectures/6002_l1.pdf.Google Scholar
- [2] . 2018. Energy-quality scalable integrated circuits and systems: Continuing energy scaling in the twilight of moore’s law. IEEE J. Emerg. Select. Topics Circ. Syst. 8, 3 (2018).Google Scholar
- [3] . 2020. LTSpice Simulator. Retrieved 2020 from https://www.analog.com/en/design-center/design-tools-and-calculators/ltspice-simulator.html.Google Scholar
- [4] . 2020. Virtualizing analog mesh computers: The case of a photonic PDE solving accelerator. In Proceedings of the International Conference on Rebooting Computing (ICRC). 133–142.
DOI: Google ScholarCross Ref
- [5] . 2020. ROC: A reconfigurable optical computer for simulating physical processes. ACM Trans. Parallel Comput. 7, 1 (2020).Google Scholar
- [6] . 2018. Recent advances in tunable and reconfigurable metamaterials. Micromachines 9, 11 (2018).Google Scholar
- [7] . 2002. Nonlinear PDE’s in Condensed Matter and Reactive Flows. Springer.Google Scholar
Cross Ref
- [8] . 1984. Adaptive mesh refinement for hyperbolic partial differential equations. J. Comput. Phys. 53, 3 (1984), 484–512.Google Scholar
Cross Ref
- [9] . 2002. Configuration relocation and defragmentation for run-time reconfigurable computing. IEEE Trans. Very Large Scale Integr. Syst. 10, 3 (
June 2002), 209–220.DOI: Google ScholarDigital Library
- [10] . 2020. HWMonitor: volate, temperatures and fan speed monitoring. Retrieved 2020 from https://www.cpuid.com/softwares/hwmonitor.html.Google Scholar
- [11] . 2017. Current Trends in High Performance Computing and Challenges for the Future. Retrieved from https://www.acm.org/binaries/content/assets/education/lc-monthly-bulletins/january2017.html.Google Scholar
- [12] . 2010. Space and time sharing of reconfigurable hardware for accelerated parallel processing. In Reconfigurable Computing: Architectures, Tools and Applications. ARC 2010. Lecture Notes in Computer Science 5992 (2010). Springer.
DOI: Google ScholarDigital Library
- [13] . 2009. Virtual configuration management for efficient use of reconfigurable hardware.
Patent No. US20090187733A1, Filed March 3rd, 2007, Issued July. 23rd, 2009. Google Scholar - [14] . 1997. Wavelet-based image registration on parallel computers. In Proceedings of the ACM/IEEE Conference on Supercomputing (SC’97).Google Scholar
Digital Library
- [15] . 2019. Reconfigurable optical computer.
Patent No. US10318680B2, Filed December 5th, 2017, Issued June. 8th, 2019. Google Scholar - [16] . 2007. Circuits with light at nanoscales: Optical nanocircuits inspired by metamaterials. Science 317, 5845 (2007), 1698–1702.
DOI: Google ScholarCross Ref
- [17] . 2005. Circuit elements at optical frequencies: Nanoinductors, nanocapacitors, and nanoresistors. Phys. Rev. Lett. 95, 9 (
Aug. 2005), 095504.DOI: Google ScholarCross Ref
- [18] . 2003. Virtualizing hardware with multi-context reconfigurable arrays. In Proceedings of the International Conference on Field Programmable Logic and Applications (FPL’03).Google Scholar
Cross Ref
- [19] . 2015. Virtualized FPGA accelerators for efficient cloud computing. In Proceedings of the IEEE 7th International Conference on Cloud Computing Technology and Science (ICCTS’15), Vol. 1. IEEE.Google Scholar
Digital Library
- [20] . 2020. Lumerical 2020a Release Speeds Photonic Design through High Performance Computing. Retrieved 2020 from https://www.prweb.com/releases/prweb16705671.htm.Google Scholar
- [21] . 2019. Towards integrated metatronics: A holistic approach on precise optical and electrical properties of indium tin oxide. Sci. Rep. 11279, 9 (
Aug. 2019).Google Scholar - [22] . 2008. Partial Differential Equations in Fluid Dynamics (1st ed.). Cambridge University Press.Google Scholar
Cross Ref
- [23] . 2016. Container and microservice driven design for cloud infrastructure DevOps. In Proceedings of the IEEE International Conference on Cloud Engineering (IC2E’16). IEEE.Google Scholar
Cross Ref
- [24] . 2020. Software stack for an analog mesh computer: The case of a nanophotonic PDE accelerator. In Proceedings of the International Conference on Computing Frontiers (CF’20). ACM.Google Scholar
Digital Library
- [25] . 2021. NIR Product Page. Retrieved from https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=4737.Google Scholar
- [26] . 2003. Ordinary and Partial Differential Equation Routines in C, C++, Fortran, Java, Maple and MATLAB. CRC Press, Boca Raton, FL.Google Scholar
Cross Ref
- [27] . 1950. Solution of partial differential equations with a resistance network analogue. Brit J. Appl. Phys. 1, 4 (1950).Google Scholar
Cross Ref
- [28] . 2019. A survey of coarse-grained reconfigurable architecture and design: Taxonomy, challenges, and applications. Comput. Surv. 52, 6 (
Oct. 2019).Google ScholarDigital Library
- [29] . 2015. Indium-tin-oxide for high-performance electro-optic modulation. Nanophotonics4 (
Mar. 2015).Google Scholar - [30] . 2018. Energy-performance considerations for data offloading to FPGA-based accelerators over PCIe. ACM Trans. Archit. Code Optim. 15, 1 (
Mar. 2018).DOI: Google ScholarDigital Library
- [31] . 2021. Approximate analog computing with metatronic circuits. Commun. Phys. 4, 196 (
Aug. 2021). Retrieved from https://www.nature.com/articles/s42005-021-00683-4.Google Scholar - [32] . 2020. EA-HRT: An energy-aware scheduler for heterogeneous real-time systems. In Proceedings of the 25th Asia and South Pacific Design Automation Conference (ASP-DAC). 500–505.
DOI: Google ScholarDigital Library
- [33] (Ed.). 2011. Latency Hiding. Springer US, Boston, MA, 1006–1006.
DOI: Google ScholarCross Ref
- [34] . 1959. Investigations into the Use of an Electrical Resistance Analogue for the Solution of Certain Oscillatory-flow Problems.
Reports and Memoranda 312. Aeronautical Research Council.Google Scholar - [35] . 2021. Extending Course-grained Reconfigurable Arrays with Multi-kernel Dataflow. Retrieved 2021 from https://people.ece.uw.edu/hauck/publications/CARLMosaic2.pdf.Google Scholar
- [36] . 2020. Coherent Beam Combining. Retrieved 2020 from https://www.rp-photonics.com/coherent_beam_combining.html.Google Scholar
- [37] . 1998. A special-purpose parallel computer for solving partial differential equations. In Proceedings of the 6th Euromicro Workshop on Parallel and Distributed Processing (PDP’98). IEEE.Google Scholar
Cross Ref
- [38] . 1974. Formal requirements for virtualizable third generation architectures. Commun. ACM 17, 7 (
July 1974).DOI: Google ScholarDigital Library
- [39] . 2017. A Brief Survey of Partial Differential Equations. Springer.Google Scholar
Cross Ref
- [40] . 2020. The Spice Page. Retrieved 2020 from http://bwrcs.eecs.berkeley.edu/Classes/IcBook/SPICE/f.Google Scholar
- [41] . 2021. Dynamic scheduling of task graphs in multi-FPGA systems using critical path. J. Supercomput. 77 (2021), 597–618.
DOI: Google ScholarDigital Library
- [42] . 2000. Digitally-configurable analog VLSI chip and method for real-time solution of partial differential equations.
Patent No. US6141676, Filed July 22, 1998, Issued October 31, 2000. Google Scholar - [43] . 2017. A comparative performance analysis of remote GPU virtualization over three generations of GPUs. In Proceedings of the 46th International Conference on Parallel Processing Workshops (ICPPW’17). IEEE.Google Scholar
Cross Ref
- [44] . 2015. Memristive accelerator for extreme scale linear solvers. In Proceedings of the Government Microcircuit Applications and Critical Technology Conference (GOMACTECH’15).Google Scholar
- [45] . 2010. Using Reconfigurable Functional Units in Conventional Microprocessors.
Sandia Report SAND2010-8063. Sandia National Laboratories.Google ScholarCross Ref
- [46] . 2019. Summary of Rules for Error Propagation. Retrieved from https://sites.fas.harvard.edu/scphys/nsta/error_propagation.pdf.Google Scholar
- [47] . 2001. Context Switching Reconfigurable Computing.
Final Technical Report AFRL-IF-RS-TR-2001-161. Sanders.Google Scholar - [48] . 1973. Cardinal Spline Interpolation. Society for Industrial and Applied Mathematics.Google Scholar
Cross Ref
- [49] . 2000. Partial Differential Equations in Mechanics 1: Fundamentals, Laplace’s Equation, Diffusion Equation, Wave Equation. Springer.Google Scholar
- [50] . 2000. Reconfigurable computing between classifications and metrics—The approach of space/time-scheduling. In Proceedings of the 10th International Workshop on Field-Programmable Logic and Applications: The Roadmap to Reconfigurable Computing (FPL’00), Vol. 1. ACM.Google Scholar
Cross Ref
- [51] . 2012. Ultra-compact silicon nanophotonic modulator with broadband response. Nanophotonics1 (
May 2012).Google ScholarCross Ref
- [52] . 2006. Adaptive Mesh Refinement for a Finite Difference Scheme Using a Quadtree Decomposition Approach.
Masters of Science Thesis . Texas Agricultural and Mechanical University.Google Scholar - [53] . 2018. Scheduling parallel tasks under multiple resources: List scheduling vs. pack scheduling. In Proceedings of the IEEE International Parallel and Distributed Processing Symposium (IPDPS). 194–203.
DOI: Google ScholarCross Ref
- [54] . 2021. Induced homomorphism: Kirchhoff’s law in photonics. Nanophotonics 10, 6 (2021).Google Scholar
Cross Ref
- [55] . 2009. Virtual configuration management: A technique for partial runtime reconfiguration. IEEE Trans. Comput. 58, 10 (
Oct. 2009).Google ScholarDigital Library
- [56] . 2015. Demonstration of WDM weighted addition for principal component analysis. Opt. Exp. 23, 10 (2015), 12758–12765.
DOI: Google ScholarCross Ref
- [57] . 2013. PDE(ODE)-based image processing methods for optical interferometry fringe. Proc. SPIE - Int. Societ. Optic. Eng. 8769, 87692D.
DOI: Google ScholarCross Ref
- [58] . 2014. Reconfigurable and tunable metamaterials: A review of the theory and applications. Int. J. Anten. Propag. (2014).
DOI: Google ScholarCross Ref
- [59] . 2020. COMSOL Blog: Heat Transfer in Deformed Solids. Retrieved 2020 from https://www.comsol.com/blogs/heat-transfer-deformed-solids/.Google Scholar
- [60] . 2011. A novel multicontext coarse-grained reconfigurable architecture (CGRA) for accelerating column-oriented databases. ACM Trans. Reconfig. Technol. Syst. 4, 2 (
May 2011).DOI: Google ScholarDigital Library
- [61] . 2018. A survey on FPGA virtualization. In Proceedings of the 28th International Conference on Field Programmable Logic and Applications (FPL’18). IEEE.Google Scholar
Cross Ref
- [62] . 2019. SciPy 1.0–Fundamental algorithms for scientific computing in Python. arXiv e-prints, arXiv:1907.10121 (
Jul 2019).Google Scholar - [63] . 2020. Resource partitioning and application scheduling with module merging on dynamically and partially reconfigurable FPGAs. MDPI Electron. 9, 1461 (2020).
DOI: Google ScholarCross Ref
- [64] . 2011. Application of Resistivity-Tool-Response Modeling For Formation Evaluation: AAPG Archie Series. Vol. 2. American Association of Petroleum Geologists.Google Scholar
Cross Ref
- [65] . 2008. Lattice Boltzmann based PDE solver on the GPU. Vis. Comput. 24, 5 (2008).Google Scholar
- [66] . 2017. To prune, or not to prune: Exploring the efficacy of pruning for model compression. arXiv:1710.01878 (
11 2017).Google Scholar
Index Terms
Virtualizing a Post-Moore’s Law Analog Mesh Processor: The Case of a Photonic PDE Accelerator
Recommendations
A Case for Virtualizing Persistent Memory
SoCC '16: Proceedings of the Seventh ACM Symposium on Cloud ComputingWith the proliferation of software and hardware support for persistent memory (PM) like PCM and NV-DIMM, we envision that PM will soon become a standard component of commodity cloud, especially for those applications demanding high performance and low ...
Transparently bridging semantic gap in CPU management for virtualized environments
Consolidated environments are progressively accommodating diverse and unpredictable workloads in conjunction with virtual desktop infrastructure and cloud computing. Unpredictable workloads, however, aggravate the semantic gap between the virtual ...
Application Performance Isolation in Virtualization
CLOUD '09: Proceedings of the 2009 IEEE International Conference on Cloud ComputingModern data centers use virtual machine based implementation for numerous advantages like resource isolation, hardware utilization, security and easy management. Applications are generally hosted on different virtual machines on a same physical machine. ...






Comments