Progress in Second-Generation Holographic Data Storage

Transcription

1 Progress in Second-Generation Holographic Data Storage Mark Ayres*, Ken Anderson, Fred Askham, Brad Sissom Akonia Holographics, LLC Optical Data Storage 2014 [ ] Page 1

2 Introduction Introduction 1st Generation Holographic Data Storage InPhase Tapestry Monocular Architecture 2 nd Generation Holographic Data Storage DRED Recording Medium Dynamic Aperture Multiplexing Quadrature Homodyne Detection Phase Quadrature Multiplexing 2 nd Generation Progress Demonstrator Design Density Experiments Technology Outlook Page 2

3 Holographic Data Storage True volumetric data storage record data up to resolution of an optical spot in all three dimensions * Optical resolution in 3D compares favorably to 2D technologies (magnetic, semiconductor, even atomic ) 0 th Generation HDS 1960s 1980s: Research and laboratory demonstrations 1 st Generation HDS 1990s 2000s: Commercial efforts, advanced prototype from InPhase, Inc. Many Moore s-law doublings still available * van Heerden, Theory of optical information storage in solids, Appl. Opt.1963 Page 3

4 1 st Generation HDS: InPhase Tapestry Advanced pre-production prototype developed : 300 GB/disk, 20 MB/s Breakthrough two-chemistry photopolymer media Off-axis angle multiplexing architecture,1.4m pixels/hologram, 320 angle multiplexed holograms/ stack (512 Gbit/in 2 peak) 50+ built, routine automated write and read of whole disks Most Tapestry technology is applicable to 2 nd generation Page 4

5 Tapestry Assembly Akonia demonstrator reuses all of these components Electronics Optical-Mechanical Assembly (OMA) Loader Laser FRU Page 5

6 1 st Generation HDS: Monocular Architecture InPhase invention co-developed with Hitachi Ltd Fast 0.85 NA objective lens >2x Tapestry density Not advanced on Engineered product scale One test bed (optical table), small density (grid) demonstrations Galvo Detector Laser SLM Objective lens Signal Beam Reference Beam Medium K. Anderson, et. al., U.S. Patent 7,742,209, Monocular holographic data storage system architecture, June 22, Page 6

7 1 st Generation Technology Two-chemistry holographic recording medium Zerowave TM disk production Angle-Polytopic multiplexing Phase-conjugate architecture Moving analog phase mask Sub-Nyquist oversampled data channel Tunable field replaceable laser ( FRU ) Multi-axis wobble alignment algorithms Isoplanatic objective lens design Self-terminating medium formulation Monocular architecture plus many more K. Curtis, L. Dhar, W. L. Wilson, A. Hill, M. R. Ayres, Holographic Data Storage: From Theory to Practical Systems, John Wiley & Sons, Ltd. (2010). Page 7

8 What Happened to 1 st Generation? InPhase committed to the off-axis Tapestry architecture before monocular option was available LTO-4 tape came out in 2007 at 800 GB, 120 MB/s; a 300 GB product was just not that impressive anymore Development was sharply curtailed 2008 when InPhase s main investor was hit hard by the financial crisis The investor dropped out altogether in 2009 while new funding was being sought Operations ceased 2010 The dog that didn t bark: InPhase was NOT shut down because of insurmountable technical obstacles Page 8

9 2 nd Generation Holographic Data Storage Introduction 1 st Generation Holographic Data Storage InPhase Tapestry Monocular Architecture 2 nd Generation Holographic Data Storage DRED Recording Medium Dynamic Aperture Multiplexing Quadrature Homodyne Detection Phase Quadrature Multiplexing 2 nd Generation Progress Demonstrator Design Density Experiments Technology Outlook Page 9

10 2 nd Generation: Akonia Holographics Launched August 2012 with an $11M investment from Acadia Woods Partners Acquired all IP, drive and media prototypes, drive and media pilot lines, and much of the equipment from InPhase 18 Employees and contractors, 12,400 ft 2 facility Founded by key technical personnel from InPhase: Dr. Ken Anderson, CEO Dr. Fred Askham, VP Materials Development Dr. Mark Ayres, CTO Page 10

11 The AP1 Akonia Demonstrator Demonstrate 4+ Moore s law doublings to catch up and surpass competing technologies Leverage proven Tapestry technology Add Monocular architecture for ~1 Tbit/in 2 density but 2+ Tbit/in 2 required by market in a few years The 2 nd Generation Akonia Innovations: DRED TM medium formulation Dynamic aperture multiplexing Quadrature homodyne detection Phase quadrature multiplexing Page 11

12 DRED TM Holographic Recording Medium 1 Hologram Exposure 6 Diffusion Control 2 Polymerization 7 3 Polymer Growth 8 Hologram Formation 4 5 Monomer Diffusion Important properties: High photosensitivity and dynamic range (M/#) Low shrinkage High optical quality (Flatness, absorption, scatter ) High archival stability ( Like DNA fossilized in amber ) DRED formula provides 6x M/# Laser Irradiation - Photoinitiator - Writing Monomer - Matrix - Refractive Index Pattern Page 12

13 Dynamic Aperture Multiplexing Dynamically change page to maintain reference/signal separation: Ordinary Monocular 4000 Density and Transfer Rate vs. Multiplexed Holograms Areal Density [Gbit/in 2 ] Transfer Rate [MB/s] Read Write Holograms Multiplexed Density/Transfer rate trade off as smaller pages are included Example: 455 holograms multiplexed 2.0 Tbit/in 2! (4x Tapestry) M. R. Ayres, K. Anderson, F. Askham, B. Sissom, Dynamic Aperture Holographic Multiplexing, Patent Pending Page 13

14 Dynamic Aperture Architecture Dynamic aperture recording implemented by scanning galvo while composing shrinking pages on the SLM Laser z x y ASE Detector SLM y z x Aperture sharing element folds reference beam into signal path without blocking it (To SLM) Polytopic Aperture ASE Galvanometer Reference Beam Objective Lens Signal Beam Medium Beams near end of scan range Page 14

15 Quadrature Homodyne Detection Traditional direct detection: I Low signal level Non-linear response Phase is lost Homodyne detection: I det ( x, y) = E ( x y) 2 det D, Amplified, linearized interference term 2 2 ( x, y) = E + E ( x, y) + 2 E E ( x, y) cosφ ( x y) LO D LO D LO D, Optical amplification Linearized signal, noise +6 db SNR PSK, QPSK modulation possible +3 db SNR 2 2 Δn r E r + E r + E r E r + E ( ) ( ) ( ) ( ) ( ) ( r) E ( r) R S PSK ambiguity terms add incoherently, no phase mask R S R S >2x Capacity M.R. Ayres, U.S. Patent 7,623,279, Method for holographic data retrieval by quadrature homodyne detection, 2009 Page 15

16 Phase Quadrature Multiplexing Phase + amplitude = 2 information channels Add Q channel, 90 o out of phase from I channel with no BER penalty Implementation: Record 2 pages at each reference angle with 90 o phase difference Another 2x Capacity ASK BPSK QPSK Please see [ ] Homodyne detection of holographic memory systems by Adam Urness for experimental results using Akonia s coherent channel M. R. Ayres, Coherent techniques for terabyte holographic data storage, Optical Data Storage Topical Meeting (ODS 2010), May (Invited paper) Page 16

17 2 nd Generation Progress Introduction 1st Generation Holographic Data Storage InPhase Tapestry Monocular Architecture 2 nd Generation Holographic Data Storage DRED Recording Medium Dynamic Aperture Multiplexing Quadrature Homodyne Detection Phase Quadrature Multiplexing 2 nd Generation Progress Demonstrator Design Density Experiments Technology Outlook Page 17

18 AP1 Demonstrator Layout PBS Assembly Reference Scanner Relay Objective Lens Laser FRU Optical-Mechanical Assembly (OMA) Loader Disk Demonstrator inherits high level of functionality from Tapestry Page 18

19 Optical Engineering for Gen 2 HDS Gen 2 HDS lab prototype developed using newly integrated: lens design, signal processing, and holographic simulation tools Many optical solutions, unique to HDS, were developed with the new toolset Gen 2 HDS Requirements Optics Solutions High density 0.85 NA FT Lens Large angle multiplexing 120 degree f-θ scanner Monocular architecture Joint design of FT and scanner lenses Transfer rate Angular magnification of galvo scans Phase conjugation λ/10 PV WFE (29 mλ RMS)* Shift tolerance Extreme isoplanatism (USP 7,532,374) Pitch correction f-θ also scans in spherical coordinates Obliquity correction Very low pupil aberration f-θ scanner Polytopic bandlimiting Object space telecentricity, filtering optics Thermal compensation Achromatic design for tunable laser +/-3nm Plan for passive alignment and element count reduction using adaptive techniques Future cost reduction and special calibrations *Lab prototype incorporated extra precision and adjustment required for toolset validation. Active alignment steps performed by Opto-Alignment Technologies and Akonia. Page 19

20 Density Grid Configuration Reference Beam Signal Beam Grid of 6 x 9 Book of 360 Books Holograms Inner 2 x 3 Fully Overlapped Density Books Polytopic Footprint ,000 bits/hologram [avg] 360 holograms/book / (0.012 ) 2 = 1.52 Tbits/in 2 Page 20

21 1.5 Tbit/in 2 Density Hologram Detected Image 4 db Dynamic aperture page recovered from density book Camera A Image 200 SNR quality metric determined from embedded reserved blocks SNR µ 1 µ 0 20 log 10 σ1 + σ 0 Page LDPC code will correct all bit errors if 2 SNR 2.0 x pixels 1 pixels Regional SNR Pixel Count Pixel Value Page 21-10

22 1.5 Tbit/in 2 Density Book SNR and power for 360 page 1.5 Tbit/in 2 density book Average 2.6 db, book ECC will correct if 90% > 2.0 db 4 Book is recoverable at 1.5 Tbit/in 2 x Hologram SNR [db] Diffracted Power [W] Reference Angle [deg] 0 Page 22

23 Areal Density Trends 1 Tb/in 2 HDD 1.5 Tb/in 2 Akonia Demo 2 Tb/in 2 Dynamic Aperture 4 Tb/in 2 QPSK Homodyne 16 Tb/in 2 Akonia 3 rd Generation LTO-6 Tape BDXL Marchon, et al., The Head-Disk Interface Roadmap to an Areal Density of 4Tbit/in 2, Advances in Tribology, vol Page 23

24 Volumetric Storage Densities HDD TB drive x x 26.1 mm 251 GB/in 3 BDXL 1 TB 12 disk cartridge 132 x 130 x 26.5 mm 36 GB/in 3 (Future 3 TB Generation 108 GB/in 3 ) SSD TB drive x x 9.5 mm 246 GB/in 3 Dynamic Aperture HDS at 2 Tbit/in 2 HDS Cartridge 25 TB 10 Holographic cards 135 x132.5.x 30 mm 772 GB/in 3 Room To Grow LTO TB cartridge x x 21.5 mm 177 GB/in 3 (Future ~4 TB Generation 284 GB/in 3 ) Homodyne QPSK (100 TB HDS Cartridge) 3.1 TB/in 3 3 rd Generation HDS (400 TB HDS Cartridge) 12.4 TB/in 3 Page 24

25 Bits in Frequency Space Other technologies record bits as spots; HDS records them as distributed gratings 3D Spatial Spectrum Planar grating spot-like spatial spectrum Planar gratings orthogonal Fourier basis; focusing beams for layered spots are not Band volume 1D Temporal Spectrum Band width mag -f C +f C freq Page 25

26 Holographic Shannon Limit Shannon capacity Usable band volume For slab-like media, C = B log 2 1+ B = f G ( S N ) 4π 2n0 3 λ f G < Spatial Frequency Claude Shannon Areal density in 1.5 mm thick media with λ=405 nm ( 1.5 mm) B = in ~ 165 Tbit/in Shannon capacity of HDS cartridge (S/N = 1.0) 4.9 PB Page 26

27 Conclusions Akonia has demonstrated 1.5 Tbit/in 2 and will soon reach 2.0 Tbit/in 2 using DRED media and dynamic aperture multiplexing Akonia is on-track to reach 4 Tbit/in 2 (and quadruple user density) using quadrature homodyne detection and phase quadrature multiplexing Akonia has a mature platform ready to serve as a reference for a product design cycle Akonia is ready to seek partners, suppliers, and investors who wish to join the inception of a major new technology Holographic Data Storage is back, Baby! Page 27

28 End Thank you for your attention! Questions? Page 28