Architecting Optically Controlled Phase Change Memory

Phase Change Memory (PCM) is an attractive candidate for main memory as it offers non-volatility and zero leakage power, while providing higher cell densities, longer data retention time, and higher capacity scaling compared to DRAM. In PCM, data is stored in the crystalline or amorphous state of the phase change material. The typical electrically-controlled PCM (EPCM), however, suffers from longer write latency and higher write energy compared to DRAM and limited multi-level cell (MLC) capacities. These challenges limit the performance of data-intensive applications running on computing systems with EPCMs. Recently, researchers demonstrated optically-controlled PCM (OPCM) cells, with support for 5 bits/cell in contrast to 2 bits/cell in EPCM. These OPCM cells can be accessed directly with optical signals that are multiplexed in high-bandwidthdensity silicon-photonic links. The higher MLC capacity in OPCM and the direct cell access using optical signals enable an increased read/write throughput and lower energy per access than EPCM. However, due to the direct cell access using optical signals, OPCM systems cannot be designed using conventional memory architecture. We need a complete redesign of the memory architecture that is tailored to the properties of OPCM technology. This paper presents the design of a unified network and main memory system called COSMOS that combines OPCM and silicon-photonic links to achieve high memory throughput. COSMOS is composed of a hierarchical multi-banked OPCM array with novel read and write access protocols. COSMOS uses an Electrical-Optical-Electrical (E-O-E) control unit to map standard DRAM read/write commands (sent in electrical domain) from the memory controller on to optical signals that access the OPCM cells. Our evaluation of a 2.5D-integrated system containing a processor and COSMOS demonstrates 2.14× average speedup across graph and HPC workloads compared to an EPCM system. COSMOS consumes 3.8× lower read energy-per-bit and 5.97× lower write energy-per-bit compared to EPCM. COSMOS is the first non-volatile memory that provides comparable performance and energy consumption as DDR4 in addition to increased bit density, higher area efficiency and improved scalability.

[1]  Ferdinando Bedeschi,et al.  A Multi-Level-Cell Bipolar-Selected Phase-Change Memory , 2008, 2008 IEEE International Solid-State Circuits Conference - Digest of Technical Papers.

[2]  M. Wuttig,et al.  Phase-change materials for rewriteable data storage. , 2007, Nature materials.

[3]  Jure Leskovec,et al.  {SNAP Datasets}: {Stanford} Large Network Dataset Collection , 2014 .

[4]  Kuanping Shang,et al.  Low-loss compact multilayer silicon nitride platform for 3D photonic integrated circuits. , 2015, Optics express.

[5]  A. Cabrini,et al.  Voltage-Driven Multilevel Programming in Phase Change Memories , 2009, 2009 IEEE International Workshop on Memory Technology, Design, and Testing.

[6]  Vijayalakshmi Srinivasan,et al.  Scalable high performance main memory system using phase-change memory technology , 2009, ISCA '09.

[7]  Xiaowei Li,et al.  Wear rate leveling: Lifetime enhancement of PRAM with endurance variation , 2011, 2011 48th ACM/EDAC/IEEE Design Automation Conference (DAC).

[8]  Rami G. Melhem,et al.  Bit mapping for balanced PCM cell programming , 2013, ISCA.

[9]  Sudeep Pasricha,et al.  DyPhase: A Dynamic Phase Change Memory Architecture with Symmetric Write Latency , 2017, 2017 30th International Conference on VLSI Design and 2017 16th International Conference on Embedded Systems (VLSID).

[10]  Moinuddin K. Qureshi,et al.  DEUCE: Write-Efficient Encryption for Non-Volatile Memories , 2015, ASPLOS.

[11]  Shiming Gao,et al.  On-chip reconfigurable optical add-drop multiplexer for hybrid wavelength/mode-division-multiplexing systems. , 2017, Optics letters.

[12]  Y.C. Chen,et al.  Write Strategies for 2 and 4-bit Multi-Level Phase-Change Memory , 2007, 2007 IEEE International Electron Devices Meeting.

[13]  Michal Lipson,et al.  WDM-compatible mode-division multiplexing on a silicon chip , 2014, Nature Communications.

[14]  Thomas Taubner,et al.  Reversible Optical Switching of Infrared Antenna Resonances with Ultrathin Phase-Change Layers Using Femtosecond Laser Pulses , 2014 .

[15]  Joonyoung Kim,et al.  HBM: Memory solution for bandwidth-hungry processors , 2014, 2014 IEEE Hot Chips 26 Symposium (HCS).

[16]  Jung Hee Cheon,et al.  Homomorphic Encryption for Arithmetic of Approximate Numbers , 2017, ASIACRYPT.

[17]  Bruce Jacob,et al.  DRAMSim2: A Cycle Accurate Memory System Simulator , 2011, IEEE Computer Architecture Letters.

[18]  Zhen Liu,et al.  An Adaptive Ensemble Machine Learning Model for Intrusion Detection , 2019, IEEE Access.

[19]  Jennifer Widom,et al.  GPS: a graph processing system , 2013, SSDBM.

[20]  Hei Wong,et al.  A comparative study of charge pumping circuits for flash memory applications , 2012, Microelectron. Reliab..

[21]  Vijayalakshmi Srinivasan,et al.  Enhancing lifetime and security of PCM-based Main Memory with Start-Gap Wear Leveling , 2009, 2009 42nd Annual IEEE/ACM International Symposium on Microarchitecture (MICRO).

[22]  Hyokyung Bahn,et al.  Characterizing Memory Write References for Efficient Management of Hybrid PCM and DRAM Memory , 2011, 2011 IEEE 19th Annual International Symposium on Modelling, Analysis, and Simulation of Computer and Telecommunication Systems.

[23]  Marco Fiorentino,et al.  A ring-resonator-based silicon photonics transceiver with bias-based wavelength stabilization and adaptive-power-sensitivity receiver , 2013, 2013 IEEE International Solid-State Circuits Conference Digest of Technical Papers.

[24]  Aart J. C. Bik,et al.  Pregel: a system for large-scale graph processing , 2010, SIGMOD Conference.

[25]  Hongliang Yu,et al.  Increasing Endurance and Security of Phase-Change Memory with Multi-Way Wear-Leveling , 2014, IEEE Transactions on Computers.

[26]  Harish Bhaskaran,et al.  Integrated all-photonic non-volatile multi-level memory , 2015, Nature Photonics.

[27]  K. Bergman,et al.  Energy-bandwidth design exploration of silicon photonic interconnects in 65nm CMOS , 2016, 2016 IEEE Optical Interconnects Conference (OI).

[28]  Jun Yang,et al.  A low power and reliable charge pump design for Phase Change Memories , 2014, 2014 ACM/IEEE 41st International Symposium on Computer Architecture (ISCA).

[29]  Nathan Youngblood,et al.  Integrated 256 Cell Photonic Phase-Change Memory With 512-Bit Capacity , 2020, IEEE Journal of Selected Topics in Quantum Electronics.

[30]  Vladimir Stojanovic,et al.  Silicon photonics for compact, energy-efficient interconnects [Invited] , 2007, Journal of Optical Networking.

[31]  Christian Bernard,et al.  POPSTAR: a Robust Modular Optical NoC Architecture for Chiplet-based 3D Integrated Systems , 2020, 2020 Design, Automation & Test in Europe Conference & Exhibition (DATE).

[32]  Yvain Thonnart,et al.  WAVES: Wavelength Selection for Power-Efficient 2.5D-Integrated Photonic NoCs , 2019, 2019 Design, Automation & Test in Europe Conference & Exhibition (DATE).

[33]  Rami G. Melhem,et al.  Increasing PCM main memory lifetime , 2010, 2010 Design, Automation & Test in Europe Conference & Exhibition (DATE 2010).

[34]  Mauro Conti,et al.  A Survey on Homomorphic Encryption Schemes: Theory and Implementation , 2017 .

[35]  Nathan Youngblood,et al.  Experimental investigation of silicon and silicon nitride platforms for phase-change photonic in-memory computing , 2020, Optica.

[36]  David A. Patterson,et al.  The GAP Benchmark Suite , 2015, ArXiv.

[37]  Karthick Rajamani,et al.  Energy Management for Commercial Servers , 2003, Computer.

[38]  Sandro Bartolini,et al.  Design Options for Optical Ring Interconnect in Future Client Devices , 2014, JETC.

[39]  David H. Bailey,et al.  The Nas Parallel Benchmarks , 1991, Int. J. High Perform. Comput. Appl..

[40]  Masaya Notomi,et al.  Toward fJ/bit optical communication in a chip , 2014 .

[41]  D. Pappalardo,et al.  Charge Pump Circuits: An Overview on Design Strategies and Topologies , 2010, IEEE Circuits and Systems Magazine.

[42]  Jun Yang,et al.  FPB: Fine-grained Power Budgeting to Improve Write Throughput of Multi-level Cell Phase Change Memory , 2012, 2012 45th Annual IEEE/ACM International Symposium on Microarchitecture.

[43]  Christopher Batten,et al.  Building Many-Core Processor-to-DRAM Networks with Monolithic CMOS Silicon Photonics , 2009, IEEE Micro.

[44]  Onur Mutlu,et al.  Architecting phase change memory as a scalable dram alternative , 2009, ISCA '09.

[45]  Linjie Zhou,et al.  All-optical non-volatile tuning of an AMZI-coupled ring resonator with GST phase-change material. , 2018, Optics letters.

[46]  Shih-Hung Chen,et al.  Phase-change random access memory: A scalable technology , 2008, IBM J. Res. Dev..

[47]  Guangjie Han,et al.  Dynamic Adaptive Replacement Policy in Shared Last-Level Cache of DRAM/PCM Hybrid Memory for Big Data Storage , 2017, IEEE Transactions on Industrial Informatics.

[48]  Chun Jason Xue,et al.  SLC-enabled wear leveling for MLC PCM considering process variation , 2014, 2014 51st ACM/EDAC/IEEE Design Automation Conference (DAC).

[49]  Andrew B. Kahng,et al.  Adaptive Tuning of Photonic Devices in a Photonic NoC Through Dynamic Workload Allocation , 2017, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[50]  R. Ho,et al.  Energy-Efficient Photonics in Future High-Connectivity Computing Systems , 2015, Journal of Lightwave Technology.

[51]  Rajeev J. Ram,et al.  Single-chip microprocessor that communicates directly using light , 2015, Nature.

[52]  Luis A. Lastras,et al.  PreSET: Improving performance of phase change memories by exploiting asymmetry in write times , 2012, 2012 39th Annual International Symposium on Computer Architecture (ISCA).

[53]  C. David Wright,et al.  An optoelectronic framework enabled by low-dimensional phase-change films , 2014, Nature.

[54]  Chen Sun,et al.  Re-architecting DRAM with Monolithically Integrated Silicon Photonics , 2009 .

[55]  Hongzhong Zheng,et al.  Co-Architecting Controllers and DRAM to Enhance DRAM Process Scaling , 2014 .

[56]  Tao Zhang,et al.  Overcoming the challenges of crossbar resistive memory architectures , 2015, 2015 IEEE 21st International Symposium on High Performance Computer Architecture (HPCA).

[57]  C. David Wright,et al.  In-memory computing on a photonic platform , 2018, Science Advances.

[58]  Ricardo Bianchini,et al.  Page placement in hybrid memory systems , 2011, ICS '11.

[59]  Carlos Guestrin,et al.  Distributed GraphLab : A Framework for Machine Learning and Data Mining in the Cloud , 2012 .

[60]  William J. Dally,et al.  Analog/Mixed-Signal Hardware Error Modeling for Deep Learning Inference , 2019, 2019 56th ACM/IEEE Design Automation Conference (DAC).

[61]  Ke Xu,et al.  Mode-Division Multiplexing for Silicon Photonic Network-on-Chip , 2017, Journal of Lightwave Technology.

[62]  Nagarajan Kandasamy,et al.  Enabling and Exploiting Partition-Level Parallelism (PALP) in Phase Change Memories , 2019, ACM Trans. Embed. Comput. Syst..

[63]  Jianmin Wang,et al.  Overview of Phase-Change Materials Based Photonic Devices , 2020, IEEE Access.

[64]  S. Ovshinsky.  Reversible Electrical Switching Phenomena in Disordered Structures , 1968 .

[65]  Joseph Gonzalez,et al.  PowerGraph: Distributed Graph-Parallel Computation on Natural Graphs , 2012, OSDI.

[66]  Ajay Joshi,et al.  Cross-Layer Co-Optimization of Network Design and Chiplet Placement in 2.5-D Systems , 2020, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[67]  Seong Keun Kim,et al.  Future of dynamic random-access memory as main memory , 2018 .

[68]  C. David Wright,et al.  On‐Chip Photonic Memory Elements Employing Phase‐Change Materials , 2014, Advanced materials.

[69]  Tao Zhang,et al.  NVMain 2.0: A User-Friendly Memory Simulator to Model (Non-)Volatile Memory Systems , 2015, IEEE Computer Architecture Letters.

[70]  K. Bergman,et al.  Resolving the thermal challenges for silicon microring resonator devices , 2014 .

[71]  Thomas Taubner,et al.  Phase-change materials for non-volatile photonic applications , 2017, Nature Photonics.

[72]  Jongman Kim,et al.  An energy- and performance-aware DRAM cache architecture for hybrid DRAM/PCM main memory systems , 2011, 2011 IEEE 29th International Conference on Computer Design (ICCD).

[73]  C. David Wright,et al.  Fast and reliable storage using a 5  bit, nonvolatile photonic memory cell , 2018, Optica.

[74]  Weisong Shi,et al.  Edge Computing: Vision and Challenges , 2016, IEEE Internet of Things Journal.

[75]  Mark Chen,et al.  Language Models are Few-Shot Learners , 2020, NeurIPS.

[76]  A. Biberman,et al.  Ultrahigh-Bandwidth Silicon Photonic Nanowire Waveguides for On-Chip Networks , 2008, IEEE Photonics Technology Letters.

[77]  Jian Huang,et al.  LL-PCM: Low-Latency Phase Change Memory Architecture , 2019, 2019 56th ACM/IEEE Design Automation Conference (DAC).

[78]  K. Gopalakrishnan,et al.  Phase change memory technology , 2010, 1001.1164.

[79]  Hsien-Hsin S. Lee,et al.  Security refresh: prevent malicious wear-out and increase durability for phase-change memory with dynamically randomized address mapping , 2010, ISCA.

[80]  Qi Wang,et al.  A 20nm 1.8V 8Gb PRAM with 40MB/s program bandwidth , 2012, 2012 IEEE International Solid-State Circuits Conference.

[81]  Onur Mutlu,et al.  Efficient Data Mapping and Buffering Techniques for Multilevel Cell Phase-Change Memories , 2014, ACM Trans. Archit. Code Optim..

[82]  Four Channel 48Gbps Multicasting in a Coupled Si Ring Resonator with Tunable Channel Spacing , 2018, 2018 Conference on Lasers and Electro-Optics Pacific Rim (CLEO-PR).

[83]  Yong-Zhen Huang,et al.  Silicon nitride three-mode division multiplexing and wavelength-division multiplexing using asymmetrical directional couplers and microring resonators. , 2014, Optics express.

[84]  Yvain Thonnart,et al.  PROWAVES: Proactive Runtime Wavelength Selection for Energy-Efficient Photonic NoCs , 2021, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[85]  Andrew Alduino,et al.  Demonstration of a high speed 4-channel integrated silicon photonics WDM link with hybrid silicon lasers , 2010, 2010 IEEE Hot Chips 22 Symposium (HCS).

[86]  Onur Mutlu,et al.  Memory scaling: A systems architecture perspective , 2013, 2013 5th IEEE International Memory Workshop.

[87]  Nikolaos Hardavellas,et al.  Galaxy: a high-performance energy-efficient multi-chip architecture using photonic interconnects , 2014, ICS '14.

[88]  Mohammad Arjomand,et al.  Boosting Access Parallelism to PCM-Based Main Memory , 2016, 2016 ACM/IEEE 43rd Annual International Symposium on Computer Architecture (ISCA).

[89]  C. Wright,et al.  Behavioral modeling of integrated phase-change photonic devices for neuromorphic computing applications , 2019, APL Materials.

[90]  Christopher Batten,et al.  Designing Chip-Level Nanophotonic Interconnection Networks , 2012, IEEE Journal on Emerging and Selected Topics in Circuits and Systems.

[91]  Jin Xiong,et al.  DWC: dynamic write consolidation for phase change memory systems , 2014, ICS '14.

[92]  Cong Xu,et al.  NVSim: A Circuit-Level Performance, Energy, and Area Model for Emerging Nonvolatile Memory , 2012, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[93]  Jun Yang,et al.  Improving write operations in MLC phase change memory , 2012, IEEE International Symposium on High-Performance Comp Architecture.

[94]  Sang Woon Lee,et al.  Capacitors with an Equivalent Oxide Thickness of <0.5 nm for Nanoscale Electronic Semiconductor Memory , 2010 .

[95]  J Feldmann,et al.  Calculating with light using a chip-scale all-optical abacus , 2017, Nature Communications.

[96]  Mark Wade,et al.  TeraPHY: A Chiplet Technology for Low-Power, High-Bandwidth In-Package Optical I/O , 2020, IEEE Micro.