The Engine. SRAM & DRAM Endurance and Speed with STT MRAM. Les Crudele / Andrew J. Walker PhD. Santa Clara, CA August

Similar documents
Unleashing MRAM as Persistent Memory

Architectural Aspects in Design and Analysis of SOTbased

Recent Advancements in Spin-Torque Switching for High-Density MRAM

Phase Change Memory An Architecture and Systems Perspective

Phase Change Memory An Architecture and Systems Perspective

Can MRAM be a factor for HPC?

Spin-Hall Effect MRAM Based Cache Memory: A Feasibility Study

An Architecture-level Cache Simulation Framework Supporting Advanced PMA STT-MRAM

Couture: Tailoring STT-MRAM for Persistent Main Memory. Mustafa M Shihab Jie Zhang Shuwen Gao Joseph Callenes-Sloan Myoungsoo Jung

Area, Power, and Latency Considerations of STT-MRAM to Substitute for Main Memory

AC-DIMM: Associative Computing with STT-MRAM

MTJ-Based Nonvolatile Logic-in-Memory Architecture

Novel Nonvolatile Memory Hierarchies to Realize "Normally-Off Mobile Processors" ASP-DAC 2014

emram: From Technology to Applications David Eggleston VP Embedded Memory

Mohsen Imani. University of California San Diego. System Energy Efficiency Lab seelab.ucsd.edu

CMP annual meeting, January 23 rd, 2014

Evaluating STT-RAM as an Energy-Efficient Main Memory Alternative

Will Phase Change Memory (PCM) Replace DRAM or NAND Flash?

Loadsa 1 : A Yield-Driven Top-Down Design Method for STT-RAM Array

Evaluation of Hybrid Memory Technologies using SOT-MRAM for On-Chip Cache Hierarchy

MRAM Developer Day 2018 MRAM Update

SOLVING MANUFACTURING CHALLENGES AND BRINGING SPIN TORQUE MRAM TO THE MAINSTREAM

Annual Update on Flash Memory for Non-Technologists

Emerging NV Storage and Memory Technologies --Development, Manufacturing and

Toward a Memory-centric Architecture

Improving Energy Efficiency of Write-asymmetric Memories by Log Style Write

SOLVING THE DRAM SCALING CHALLENGE: RETHINKING THE INTERFACE BETWEEN CIRCUITS, ARCHITECTURE, AND SYSTEMS

Technology, Manufacturing and Markets of Magnetoresistive Random Access Memory (MRAM) Brad Engel, VP- Product Development & Quality

This material is based upon work supported in part by Intel Corporation /DATE13/ c 2013 EDAA

Memory technology and optimizations ( 2.3) Main Memory

OAP: An Obstruction-Aware Cache Management Policy for STT-RAM Last-Level Caches

Embedded 28-nm Charge-Trap NVM Technology

Don t Forget the Memory. Dean Klein, VP Advanced Memory Solutions Micron Technology, Inc.

MRAM - present state-of and future challenges

ODESY: A novel 3T-3MTJ cell design with Optimized area DEnsity, Scalability and latency

Test and Reliability of Emerging Non-Volatile Memories

STT-RAM has recently attracted much attention, and it is

Cascaded Channel Model, Analysis, and Hybrid Decoding for Spin-Torque Transfer Magnetic Random Access Memory (STT-MRAM)

Steven Geiger Jackson Lamp

Design Method of Stacked Type MRAM. with NAND Structured Cell

NAND Flash Memory: Basics, Key Scaling Challenges and Future Outlook. Pranav Kalavade Intel Corporation

Energy-Efficient Spin-Transfer Torque RAM Cache Exploiting Additional All-Zero-Data Flags

NAND Flash Memory. Jinkyu Jeong Computer Systems Laboratory Sungkyunkwan University

Logic and Computer Design Fundamentals. Chapter 8 Memory Basics

Don t Forget the Memory: Automatic Block RAM Modelling, Optimization, and Architecture Exploration

! Memory Overview. ! ROM Memories. ! RAM Memory " SRAM " DRAM. ! This is done because we can build. " large, slow memories OR

Embedded System Application

Emerging NVM Enabled Storage Architecture:

Don t Forget the Memory. Dean Klein, VP Memory System Development Micron Technology, Inc.

From Silicon to Solutions: Getting the Right Memory Mix for the Application

SLC vs MLC: Considering the Most Optimal Storage Capacity

Concept of Memory. The memory of computer is broadly categories into two categories:

CS429: Computer Organization and Architecture

Magnetoresistive RAM (MRAM) Jacob Lauzon, Ryan McLaughlin

Computing with Spintronics: Circuits and architectures

Chapter 8 Memory Basics

Driving the MRAM Revolution. Kevin Conley CEO, Everspin Technologies

Technical Notes. Considerations for Choosing SLC versus MLC Flash P/N REV A01. January 27, 2012

CS 201 The Memory Hierarchy. Gerson Robboy Portland State University

Intel s s Memory Strategy for the Wireless Phone

ECE 152 Introduction to Computer Architecture

Architecture for Carbon Nanotube Based Memory (NRAM)

Cache Memory Configurations and Their Respective Energy Consumption

A Better Storage Solution

Lecture 14. Advanced Technologies on SRAM. Fundamentals of SRAM State-of-the-Art SRAM Performance FinFET-based SRAM Issues SRAM Alternatives

Fault Injection Attacks on Emerging Non-Volatile Memories

Scalable Many-Core Memory Systems Lecture 3, Topic 2: Emerging Technologies and Hybrid Memories

Respin: Rethinking Near- Threshold Multiprocessor Design with Non-Volatile Memory

CS429: Computer Organization and Architecture

Designing Giga-scale Memory Systems with STT-RAM

Middleware and Flash Translation Layer Co-Design for the Performance Boost of Solid-State Drives

ABSTRACT. Mu-Tien Chang Doctor of Philosophy, 2013

CIRCUIT AND ARCHITECTURE CO-DESIGN OF STT-RAM FOR HIGH PERFORMANCE AND LOW ENERGY. by Xiuyuan Bi Master of Science, New York University, 2010

Microprocessor Trends and Implications for the Future

A Low Power 1Mbit MRAM based based on 1T1MTJ Bit Cell Integrated with Copper Interconnects

Memory Overview. Overview - Memory Types 2/17/16. Curtis Nelson Walla Walla University

Semiconductor Memory Types Microprocessor Design & Organisation HCA2102

The Memory Hierarchy. Computer Organization 2/12/2015. CSC252 - Spring Memory. Conventional DRAM Organization. Reading DRAM Supercell (2,1)

DESIGN AND PERFORMANCE ANALYSIS OF A NONVOLATILE MEMORY CELL

Evolving Solid State Storage. Bob Beauchamp EMC Distinguished Engineer

Computing-in-Memory with Spintronics

The Evolving NAND Flash Business Model for SSD. Steffen Hellmold VP BD, SandForce

YEILD, AREA AND ENERGY OPTIMIZATION IN STT-MRAMS USING FAILURE- AWARE ECC

CS 33. Memory Hierarchy I. CS33 Intro to Computer Systems XVI 1 Copyright 2016 Thomas W. Doeppner. All rights reserved.

Semiconductor Memory II Future Memory Trend

Maximize energy efficiency in a normally-off system using NVRAM. Stéphane Gros Yeter Akgul

Semiconductor Memory Classification. Today. ESE 570: Digital Integrated Circuits and VLSI Fundamentals. CPU Memory Hierarchy.

Versatile RRAM Technology and Applications

VM Futures Panel: Are We Near the Beginning of a New Solid State Epoch, or Is That Just a Big Firebal in the Sky?

Could We Make SSDs Self-Healing?

The Memory Hierarchy 1

Programming Characteristics on Three-Dimensional NAND Flash Structure Using Edge Fringing Field Effect

Developing a Prototyping Board for Emerging Memory

Reconfigurable RRAM-based computing: A Case study for reliability enhancement

CS24: INTRODUCTION TO COMPUTING SYSTEMS. Spring 2017 Lecture 13

IEEE TRANSACTIONS ON COMPUTERS, VOL. XX, NO. X, AUGUST TA-LRW: A Replacement Policy for Error Rate Reduction in STT-MRAM Caches

Flash Trends: Challenges and Future

Adapting Controllers for STT-MRAM

Computer Systems Research in the Post-Dennard Scaling Era. Emilio G. Cota Candidacy Exam April 30, 2013

UNIT 2 Data Center Environment

Transcription:

The Engine & DRAM Endurance and Speed with STT MRAM Les Crudele / Andrew J. Walker PhD August 2018 1

Contents The Leaking Creaking Pyramid STT-MRAM: A Compelling Replacement STT-MRAM: A Unique Endurance Conundrum The ENGINE for & DRAM Endurance and Speed The Physics of the ENGINE ENGINE Emulation Data from Emulated ENGINE ENGINE Benefits Takes MRAM Mainstream! The Performance, Energy and Cost Advantage Conclusions August 2018 2

Nonvolatile Volatile The Leaking Creaking Pyramid (1) Leakage Current (ma) CPU Gap DRAM Mind the Gap NAND DISK Cell Area in F 2 FinFET Inefficient Scaling* Planar CMOS Technology Node F (nm) Fundamental Leakage** 1000 Mb 200 Mb 100 Mb 50 Mb 20 Mb 10 Mb * R. de Werdt et al., IEDM87; R.D.J. Verhaar et al., IEDM90; S.H. Kang et al., IEDM17 Temperature ( o C) August 2018 ** Simulations/estimates from 7nm transistor data in IEDM 2017 3

The Leaking Creaking Pyramid (2) Energy per Operation (pj) GigaOperations per Joule CPU On-chip data access Keep Data Local! On-chip data access Operation Off-chip data access * What We Get for a Joule * Gap DRAM Off-chip data access Operation Memory access energy grows as the length of the wire * > 60% of system energy used is in data movements to and from DRAM ** * A. Pedram et al., Dark Memory and Accelerator-Rich System Optimization in the Dark Silicon Era, IEEE Design & Test, vol.34, April 2017 ** A. Boroumand et al., Google Workloads for Consumer Devices: Mitigating Data Movement Bottlenecks, ASPLOS 18, March 2018 August 2018 4

STT-MRAM : A Compelling Replacement Size : Dramatic bitcell reduction Leakage : No array leakage Persistence : Data retained On-chip : Reduces DRAM accesses Rad-Hard Cell Area in F 2 MRAM FinFET Planar But: Limiting Endurance Conundrum Non-symmetric R/W CMOS Technology Node F (nm) Disturbs with fast read August 2018 5

Write Error Rate STT-MRAM: A Unique Endurance Conundrum Magnetic Tunnel Junction (MTJ) Free layer Tunnel barrier Reference layer CoFeB MgO CoPt The Stochastic Top Hat Pulse width Low Write Error Rate needs large tunnel current Limits endurance due to oxide wear out mechanism High endurance with low Write Error Rate needs reduced tunnel current Make Free Layer magnetically less stiff Reduce MTJ area Use special design techniques - The ENGINE MTJ Voltage (V) August 2018 6

Write Error Rate Log(Write Error Rate) The Physics of the ENGINE The Stochastic Top Hat Pulse width Increased WER Reduced electrical stress MTJ Voltage (V) MTJ Voltage (V) Proprietary Circuit Design allows reduced electrical stress Results in large endurance increase (~ 6 orders of magnitude) Circuit Deals with resultant Write Error Rate increase Managed transparently to the user No change in latency Allows for faster pulses at high endurance August 2018 7

ENGINE Emulation Emulation Board (FPGA emulation of Engine) Front Panel (Control timing, address, patterns, W/R ops etc.) Daughter Board (Contains SPIN s own 4Mb STT-MRAMs) Logic Analyzer (Verify addresses, timing, patterns, W/R etc.) August 2018 8

Write Error Rate Cycles to Breakdown Data from Emulated ENGINE 20ns Engine OFF Engine ON 380 mv Voltage (V) Voltage (V) ~ 6 orders of magnitude endurance boost August 2018 9

ENGINE Benefits Takes MRAM Mainstream! /DRAM-like Endurance ~ 6 orders of magnitude improvement with circuitry alone Symmetric R/W 10ns capable Random R/W Looks and feels like an Corrects Read Disturbs allowing Fast Reads No user visible errors August 2018 10

Macro Area Impact versus (%) The Performance, Energy and Cost Advantage 7nm CMOS node Macro Capacity MRAM Cell Cell 140F 2 560F 2 Dramatic area impact Maximize large on-chip cache and memory capacity Allows persistence on-chip Reduces DRAM accesses ~ Zero array leakage Symmetric Read/Write August 2018 11

Conclusions The ENGINE takes MRAM into the mainstream Provides a path for MRAM as /DRAM replacement Uses evolving ecosystem for embedded NV MRAM as a foundation Combines low Write Error Rates with /DRAM-like endurance resulting in zero user-visible errors Achieves fast and symmetric read and write Makes MRAM the most likely candidate for replacement and, eventually, DRAM August 2018 12