DSNet-UK

1
UK Optical Storage Roadmap

• DSNet-UK General Meeting
• Manchester
• 15th Dec 2004

Dr Andrew Pauza
2
Optical Data Storage

• Aim to roadmap optical storage RD within the
  UK.
• Will ignore potentially disruptive technologies
  in other areas e.g.
• increasingly higher capacity 1GB USB memory
  sticks
• 2 and 4GB microdrives (personal data storage)
• MAID Massive Array of Inactive Discs (archive
  arena)
• It is artificial to consider UK optical data
  storage
• any successful research will have links with
  other EU countries and probably also Japanese and
  US companies/universities
• In optical data storage, there is a very strong
  manufacturing equipment base in Europe
• M2, Unaxis, ODC Nimbus, Dr Schenk, Ciba, ETA
  Optic, Singulus, AudiDev, Axxicon, AWM, Netstal,
  Dr Schwab etc

3
Optical State of the Art

• Blue-ray Disc Dual layer (BD-DL)
• 50GB/side, 4.5Mb/sec (1x) speed ,120mm disc
• could be bought in the shops in Japan in July
  2004
• Higher speeds are on the drawing board (2x, 4x)
• Quad layer (100GB/side) - will almost certainly
  only be WO.
• Holographic Media (Holographic Versatile Disc,
  HVD)
• HVD Alliance formed as well as ECMA Technical
  committee (TC44). Draft HVD specification
  underway
• 200GB/side, 30MB/sec, 120mm disc, WO format
• Due for sale in 2005

4
Optical Disc Capacity

• Optical disc (CD, DVD, HD-DVD, BD) capacities
  compared to HDD and development with time
  (1GB/in21.55bit/um2)

5
Optical Disc Data Rate

• Separate out ROM, WORM (-R, -WO) and ReWritable
  (-RW) as historically, the factors limiting the
  data rate (and write or read) have been different

6
Technical Challenges/Holy Grails

• The strengths that optical disc has and which
  need to be core to any new optical media are
• Cheap to manufacture - low cost/GB relative to
  other storage media.
• Removable - low energy costs for large amounts
  of near-line data.
• Interchangeable - require strong standards.
  Often break down DVD/-RW format war, HD-DVD vs
  BD format war
• Backwards compatible (CDlt-gtDVDlt-gt HD-DVD lt-gt BD
  ). A new technology -such as holography, 2-photon
  or near field will suffer from a lack of a user
  base unless first applied in niche areas, or
  giving an extremely strong competitive advantage.
• These strengths also present technical challenges

7
Influence of Standards

• the requirement for strong standards makes
  discussion of UK-only optical storage roadmap is
  artificial
• implies you have either invented the technology
  and/or
• have a strong technological leadership
• These situations do not exist in the UK for
  optical data storage.
• What influence can the UK have on optical DS?

8
Importance of Standards
Gartner/Dataquest "Standards efforts improve
the market opportunity for all vendors especially
when such efforts are timely and supported by a
majority of vendors in an industry. The
availability of products written to published
specifications will undoubtedly help to increase
adoption rates and ease the application
development and integration process" - Daya
Nadamuni, market analyst with Gartner/Dataquest

• Philips
• Many technologies of tomorrow will be too
  expensive and too risky for a single company to
  develop. lt will require two or more companies to
  combine RD and agree ahead of time on standards
  so that millions of dollars won't be lost just
  because a company backs the wrong format, such as
  the Beta VCR.
• - Gerard Kleisterlee, Philips CEO (USA Today -
  November 19, 2002)

Nokia industry must start focusing on adopting
common technology standards rather than
one-upmanship. If compelling applications are
created with a common technological base,
companies developing applications will be able to
benefit from a competitive, innovative
environment. We have to differentiate on the
services and applications side, not the
technology. - Paul Chapple, marketing director
Nokia US mobile software group (Ottawa Business
Journal Feb 18, 2003)
9
UK Optical DS Influence

• RD in the UK can realistically only hope to
  influence the direction of optical data storage
  by
• Working on improvements to current optical
  technologies which show clear technological and
  manufacturing advantages and which are backwardly
  compatible to current optical discs.
• Working with technologies which are likely to be
  adopted by, the US or Japan e.g. holographic
  materials compatible with InPhase Technologies
  (US) or OPTWARE (Japan)
• E.g. Holographic join the (recently formed) HVD
  alliance
• Having long term research goals which produce a
  technology which is such a clear technological
  winner (cheap to manufacture, high capacity, high
  data rate, compatible with DVD, or BD) that it
  becomes a disruptive technology
• E.g. MODS?

10
UK Optical RD Directions

• Avenues which the UK could followed in order to
  improve the prospects for any future return on
  the research
• New materials
• development of materials which can be applied to
  make current optical formats cheaper to produce,
  or fit within standards definitions of future
  media e.g. holographic, shorter wavelength
  (250nm).
• income could be obtained from licensing of the
  material and/or the process.
• New formats
• we need to be aware of, get involved with or even
  help initiate standards for holography, MODS
  (Multiplexed Optical Data Storage) etc.
• applies to all storage formats such as probe
  storage and is especially the case where media is
  potentially backwards compatible to current
  formats.

11
UK Optical RD Directions

• Format Neutral
• Be aware of the requirements for markets where
  the dependence on (consumer) optical disc
  standards can be bypassed such as in archival
  storage where the discs and drives are generally
  used within a jukebox - the standardized
  interface is then moved to the connectivity
  channel (which is based on SCSI or Fiber Channel
  etc). New formats could be developed for use in
  this environment for proving before pushing them
  as consumer formats, which require significantly
  more agreement on standards
• New support technologies
• work on aspects of an optical disc product which
  are applicable across all types, such as disc
  dynamics, channel coding and signal processing,
  optical modelling, manufacturing techniques etc.
  Longer term returns on this would have to be from
  licensing through careful patent protection

12
Over-riding factors for success

• The winning formula for next generation storage
  will be primarily determined by economic factors
  (such as production costs) and by customer needs.
• Using patent protection (and licensing) is always
  subject to the problems of companies attempting
  to break the patent
• high costs of enforcement, especially in less
  well regulated foreign jurisdictions.

13
Single Layer Formats

• Super RENS (most of Asia!) much work over a few
  years now but cant see this going anywhere.
• Near field (e.g. ex Terastor) now Philips, Sony
  etc. issues with dust and removability. Philips
  seem to think dust isnt an issue. Cant go
  multilayer?
• Phase change organic phase change dyes can
  they be used for multilayer or multilevel?.
• MO/MSR (e.g. MAMMOS) - cant see this going
  anywhere inorganic phase change dominates now
  and is possible to use in multilayer.
• MODS (Multiplexed Optical Data Storage) -Imperial
  college, Peter Török. Essentially a new coding
  format using polarization to give an extra
  dimension. The technique has been evaluated by
  static tester and rigorous modelling has been
  performed. Can be backwards compatible DVD, BD.
  Might be possible to use in multilayer.
• Phase change - carbon

14
Volumetric-multilayer addressing

• Multilayer BD coding can give 35GB/layer,
  theory plus some prototypes suggest up to 8
  layers can be made giving 280GB/side (Sony ODS
  2004). Probably there is no non competitive work
  which can be done (on format, materials) in this
  area, though there is significant work still to
  be done on manufacturing issues.
• Low absorbtion multilayer (Thomson) pick up
  idea in UK to study? Issues of manufacturing
  costs of tens of layers? Could be modelled by
  Imperial.
• VMD (is this C3Ds Fluorecent Multilayer Disc
  rebranded?) Costs of manufacturing 10s of
  layers?
• Electrochromic discs contacts from drive to
  selected layer on disc makes it opaque and
  sensitive to be written/read.

15
Volumetric- true 3D addressing

• Holographic -some work in companies (InPhase
  Technologies Tapestry format) and universities
  in UK (e.g P3 holographics, Manchester,
  Nottingham Trent).
• HVD alliance formed recently to create
  standards
• Two photon -volumetric bit writing, usually
  requires high laser powers.
• New materials developed which substantially lower
  this threshold power - e.g. Landauer AlOx.
• Call/Recall FROST program multilayer and
  multitrack for high data rate demo of 22Gb on
  90mm dia x 3mm 1MB/s recording rate. 64 parallel
  chanel, 64MB/s demo.
• Interferometry-requires very good laser
  stability.
• Spectral Hole burning expensive narrowband
  lasers, restriction of FIFO data (until
  addressing schemes are developed)
• See Coufal and Burr in International Trends in
  Optics, 2002. Also, more info below
  (Alternative Optical Storage Technologies)

16
Supporting and cross disciplinary technologies

• Mastering (e.g. Phase Transition Mastering for BD
  no longer pre- competitive) but adoption for
  other technologies could be a fruitful research
  area.
• Channel Improve error margins, increase data
  rate and capacity with perhaps new modulation
  and ECC codes (e.g. error correcting modulation
  code?), reduced complexity schemes, better
  inter-symbol interference (ISI) and cross talk
  (XT) tolerance, neural net based schemes. One
  solution is for example the optical format
  TwoDOS which gives a 2x capacity increase but a
  10x data rate increase.
• Servos work on feed forward, tilt compensation,
  tolerance to larger accelerations, shorter seek
  times etc.

17
Supporting and cross disciplinary technologies

• Optics design improvements, lower cost, better
  margin optics or optics for future media (e.g.
  transparent at 250nm)
• Sputtering or deposition technologies
  sputtering, spreading etc new coating or active
  layer materials more cheaply, using better
  materials etc.
• Moulding technologies e.g materials which can be
  moulded and are transparent at 250nm
• Spin coating technologies To enable BD
  manufacture, make multilayer cheaper, can apply
  across several proposed formats. EU framework
  program with M2 being formulated.

18
Roadmaps- Consumer

• Low cost, streaming data, moderate data rate
• A potential view of the near future is
• 1-4 years - HD-DVD up to 2/3 layers, 15GB/layer
  
• 1-3 years - BD up to 1 layers, 35Gb/layer
  lowish data rate
• 3-5 years - BD up to 2 layers, 70Gb/layer
  lowish data rate (or 3 beam XTC, 55GB one layer,
  100Mbps)
• 5-7 years - BD up to 4 layers, 140Gb/layer
  lowish data rate next 5-7 years
• 7-9 years - up to 8 layers, 280Gb/layer, lowish
  data rate
• 5-15 years - Holographic/MODS/2-photon next.

19
Roadmaps-Professional

• data rate, data authenticity and longevity are
  emphasised
• A potential view of the near future is
• 1-2 years - UDO/PDD (channel improvements) up
  to 1 layers- 30/40Gb layer, lowish data rate
• 2-4 years - UDO/PDD up to 2 layers- 60/80Gb
  layer, lowish data rate
• 4-6 years - UDO/PDD up to 4 layers- 120/160Gb
  layer, lowish data rate
• 6-10 years (1) - UDO/PDD up to 8 layers-
  240/320Gb layer, lowish data rate
  OR
• 6-10 years (2) - use WO MODS/volumetric bit
  writing/2-photon after 2 BD layers?

20
UK Case Studies

• See White Book for more information
• MODS Imperial College
• Spectral Hole burning Exeter
• Holographic dyes Nottingham Trent