High Density, High Performance, Holographic Data Storage: Viable at last?

William L. Wilson Photonics Materials Research Department Bell Laboratories, Lucent Technologies Murray Hill NJ 07974 Phone:+1-908-582-7919, Fax:+1-908-582-3958 E-mail: [email protected]

Presented at the THIC Meeting at the Naval Surface Warfare Center Carderock, 9500 MacArthur Blvd West Bethesda MD 20817-5700 October 3, 2000
(1)

High Density, High Performance, Holographic Data Storage: Viable at last?

William L. Wilson Photonics Materials Research Department Bell Laboratories, Lucent Technologies Murray Hill, NJ 07974

Introduction to holography Optical storage via volume holography Technology review and current status Conclusion and prognosis
(2)

Partial List of Contributors

Kevin Curtis William Wilson Mike Tackitt Adrian Hill Tom Richardson Peter Littlewood Partha Mitra Scott Campbell

System

Lisa Dhar Alex Harris Marcia Schilling Howard Katz Melinda Schnoes Arturo Hale Carol Boyd Adam Olsen

Material

(3)

Holography Primmer
*Total recording of the optical field

(4)

Holographic Recording Fundamentals

Parallel readout of data - fast rates Multiplexing - much greater density

(5)

The Holographic Advantage

Storage of many holograms in the same volume with selective addressing of the individual holograms generates density!!
Issue: How do we access each hologram independently? (Selectivity, allows us to exceed the diffraction limit)

Angle Multiplexing* Wavelength Multiplexing* Shift Multiplexing* Peristropic Multiplexing Phase-code Multiplexing*
*Most techniques are based on or driven by the Bragg effect or strongly enhanced by Bragg processes

The Bragg Effect

Constructive interference achieved at specific for a given

The higher the number of scattering planes, The finer the angular or wavelength resolution!!

Bragg Selectivity of Volume Holograms

change angle change wavelength
Angular Selectivity of 38 um thick photopolymer film

1.0

0.8

Diff. Int (AU)

0.6

0.4

0.2

0.0 -2 -1 0 1 2

Deviation from Bragg (Degrees)

(8)

Promise of Holographic Storage The Dream? Maximum density >1bit/3

**The Library of Congress on a sugar cube!!

(9)

We are not the First!

The Essential Components of a Block-organized Holographic Memory

4A laser Source 4A laser beam deflector* 4A flys eye lens 4A page composer 4A detector array 4A storage matrix**
Bernhard Hill, Philips Labs Advances in Holography, 1976

(10)

Past Problems with Holography

Material - No acceptable material and media requirements were ill defined. Methods - Complex and difficult, limited density Lasers - Extremely costly and unreliable Detector - Cost and performance SLM - Performance at issue, (frame rates, through put, etc.)
(11)

But we cant ignore the technology!

10000 Library of Congress

/ 4 " Disk)

1000

100 Movies

Capacity, GBytes/(5

Holographic Storage Holographic Storage

100

10

1 Movie (DVD)

Industry Trend Industry Trend

1 1996 CD-ROM

1980

1985

1990

1995

2000

Characteristics: High Transfer Rates - > 1GB/sec High Density - initially 500 Gbytes on a 5-1/4 format Low Cost/MB - comparable to tape Write Once, Removable storage - WORM Parallel Read/Write - page access instead of bit access
(12)

Past Problems with Holography

Material - No acceptable material and media requirements were ill defined. Methods - Complex and difficult, limited density Lasers - Extremely costly and unreliable Detector - Cost and performance SLM - Performance at issue, (frame rates, through put, etc.)
(13)

Holographic Recording Fundamentals :Translational/Shift Multiplexing

Shift direction

Object

Material Reference

Hologram forms in Overlap Region

Parallel readout of data - fast rates leverages current technologies

(14)

New Method: Correlation Multiplexing (CM)

Signal Beam
60

Reference Beam

Multiplex by moving the medium

Diffracted Intens.

40

20 10 5 0 -10 -5
X

Simple to implement Selectivity independent of material thickness ~5x density over other holographic approaches Written 16,000 ~300Kbit holograms at 350 bits/um2

0 (m)

5 10 -10

Selectivity Map

(15)

-5

( m )

Array of Stored Holograms

30m
x

30m

Hologram Dia. ~3mm

(16)

Hologram Fidelity
75

Y Microns

50 25 0 -25 -50 -75 -75 -50 -25 0 25 50 75

X Microns

0.5

0.6

0.7

0.8

0.9

1.0

Intensity (arb. units)

(17)

First hologram

Data Stored and Recovered from Polymeric Media

3.5 3.0 2.5

< > ~ 10

-5

amplitude SNR

Last hologram

2.0 1.5 1.0 0.5 0.0 500 1000 1500 2000 2500 3000
~50% Overhead for Channel Modulation and ECC Raw BER O~10
-5

Hologram number

Density: ~48 channel bits/m2

(18)

Methods, a Prognosis

New innovative simple methods developed! High density recording demonstrated High fidelity holograms recorded

Physics not our problem!!

Past Problems with Holography

Material - No acceptable material and media requirements were ill defined. Methods - Complex and difficult, limited density Lasers - Extremely costly and unreliable Detector - Cost and performance SLM - Performance at issue, (frame rates, through put, etc.)
(20)

Requirements for holographic storage media

Dynamic Range - High storage densities & rapid read rates Photosensitivity - Rapid write rates Millimeter Thickness - High storage densities Dimensional Stability - High fidelity data recovery

High Storage Capacity Rapid Write/Read Rates

Optical Flatness - High fidelity imaging of data pages Low Scatter - Low levels of noise in data recovery Processing - Heat/Solvent Free Non-volatile readout Long shelf-life of media Long archival life of stored data Environmental/thermal stability

(21)

Past Candidates for Materials for Holographic Data Storage

LiNbO3 Volatile Readout Low photosensitivity Low dynamic range Photorefractive polymers Volatile Readout Requirement of poling voltages (large) Low photosensitivity Photopolymers . . .

Whats the problem?

(22)

Material Dynamic Range is King!

**But an overall system view is Queen!!
You need to decide what to build!
System capacity determines the density needed a fixed media size Density at fixed media size provides M Power budget effects minimal Signal detection floor effect limiting

Materials

Systems

(23)

M/# as Standard Metric

Can be used as a empirical parameter for any material system!

= (M/#/M)2

In general, for all systems the diffracted intensity of an ensemble of holograms written under identical conditions scales as 1/M2

The M-number allow us to quantify a materials utility as a storage media.

*a much better metric for storage!

Our View: Go WORM/ROM

Punt read/write, and try to solve other problems with the technology
Concentrate

on a WORM or ROM based system to shake out technology

Materials for our assessment: Photopolymers

(25)

Grating Formation in Conventional Photopolymer Systems

Mechanism
System consists of monomers dissolved in a matrix.

Advantages
Holographic exposure produces a spatial pattern of photoinitiated polymerization. High photosensitivity Permanent holograms Low cost Concentration gradient in unreacted monomers induces diffusion of species.

Diffusion produces a compositional gradient, establishing a refractive index grating.

(26)

Dynamic Range in Bell Labs Low-Tg Photopolymer Media:

10 9 8

M# @ 200 m

7 6 5 4 3 2 1 0

Goal: M# = 6 @ 200 m ( M# = 30 @ 1 mm)

Moores Law

Measurements performed in materials exhibiting 0.3% shrinkage during recording

(27)

6/95

9/95

7/96

8/97

3/98 1/98

Date

Optical Quality in Low Tg Photopolymer Materials

/cm flatness achieved in 1 mm thick media 3x3 10
Interferogram of a Photopolymer-Based Disk
100

Data page is imaged through sample with a raw BER ~ 1x10-6.

(28)

Optical Quality: Pixel Matching with Low Bit Error Rates

Expanded view of corner pixels

Transmitted 800x600 data page

100 Number of Pixels

150

10

Number of Pixels

Histogram of pixel intensities, a measure of fidelity of data recovery. Raw BER ~ 10-6

100

1 -0.5

0.0 0.5 1.0 1.5 Normalized Pixel Intensities

2.0

50

0 -0.5 0.0 0.5 1.0 1.5 2.0

Normalized Pixel Intensities

(29)

Demonstrated Density in Photopolymer Media

100 Channel Density User Density - Material 1 User Density - Material 2 100

Density (bits/ m2)

10
DVD user density

10

2 (4.4 bits/mm )

CD-ROM user density

0.1

(0.4 bits/mm 2)

0.1

0.01

New Material and System Development

0.01

Jan '96 Jul '96

Jan '97 Jun '98

Jul '98 Aug '98

(30)

GBytes/ (5 1/4" Disk)

Conclusions
Design, fabrication, & demonstration of high-performance photopolymer media for high density holographic data storage ! High dynamic range, high photosensitivity with controlled recording-induced dimensional change ! Optically flat, millimeter-thick, low scatter formats ! Demonstrated high density digital data storage capabilities Currently in program with Imation Corporation to further jointly develop the photopolymer media.

WORM Media Capabilities - 5.25 disk

Current Current Material Material (532 nm, (650 nm, 50 mW laser) 50 mW laser) User Capacity per disk (channel density) 150 GB (106 Gbits/in2) 250 GB (175 Gbits/in2) Current Material (425 nm, 15 mW laser) 400 GB (300 Gbits/in2)

Write Rate Read Rate

20 MB/sec 70 MB/sec

23 MB/sec 64 MB/sec

20 MB/sec 40 MB/sec (limited by laser power)

(31)

Materials, a Prognosis
New strategy developed! All issues on the table! High M# media in hand! Requirements out in the open High density digital recording demonstrated Digital data recovered at low BER, (10-5)

*Still the key problem, but on the run!!

Past Problems with Holography

Material - No acceptable material and media requirements were ill defined. Methods - Complex and difficult, limited density Lasers - Extremely costly and unreliable Detector - Cost and performance SLM - Performance at issue, (frame rates, throughput, etc.)
(33)

Component Summary
Laser - uchip/small cavity Tested Uniphase and Micracor models Cost, ~$2K going to $20 High volume is possible Power, 200-600mW

Detectors - CMOS Cameras

Custom detector functioning Cheaper to produce than CCDs =>$300 going to $30 Lower power, Less heat, Lower noise

SLM - DMD Status

(34)

Fixed Problems with windows 2,000 frames/sec 80% light throughput Contrast, ~1000:1 800x600 pixels

Components, a Prognosis
All driven by other technologies! SLM a commercial product Camera soon to be a commercial high volume item

Only laser at issue, problem: price!!

DEMO: System Architecture

100mW laser

SLM CMOS Camera HOE for Reference 1 x 2 foot area Stages & Sample

(36)

(37)

System Risks
Engineering and Drive Design Long term life issues for Media and Components Market
If you build it will they buy???

(38)

Past Problems with Holography (Current Status)

Material - Two chemistry materials class has been developed,
study of issues relevant to manufacture underway

Methods - An array of simple, high density method developed.

A compact demonstration device built, (1ft x 2ft).

Laser - Moderate to low cost options coming on line. Detector - Optimal device type in hand. SLM - High throughput, fast frame rate devices available.
(39)