The Future of Tape. Dr. Mark Lantz Manager Advanced Tape Technologies Principal Research Staff Member IBM Research - Zurich.

Transcription

1 August, 2017 The Future of Tape Dr. Mark Lantz Manager Advanced Tape Technologies Principal Research Staff Member IBM Research - Zurich 2013 IBM Corporation

2 Outline Introduction: The role of tape in the era of big data The Future of Tape Tape areal density trends and future scaling potential World record particulate BaFe tape areal density demo of 123 Gb/in 2 World record sputtered metal tape areal density demo of 201 Gb/in 2 Demo technologies INSIC tape technology roadmap Conclusions 2

3 Tape s Renaissance Data continues to grow exponentially while HDD scaling has stagnated à Driving demand for cost effective storage IDC Projection for Data Growth HDD Areal Density Scaling 42% CAGR $/GB scaling of HDD (raw disk cost) 3 80% of all files created are inactive no access in at least 3 months! Ref:

4 Tape advantages for long-term storage Very energy efficient: no power needed once data is recorded Very secure: Data is inaccessible when cartridge is not mounted à air gap Drive level encryption Portable à offsite vaulting Very long expected media lifetime (30+ years) Very reliable: Read while write verification Typically no data loss in case of drive failure Main net advantage of tape for archival storage is cost Recent studies from the Clipper Group: à Investigate 9 year TCO of a 1PB archive that grows to 52 PB (55% CAGR) Main Finding: 6.7x TCO advantage of LTO Tape over Disk 1) Continuing the Search for the Right Mix of Long-Term Storage Infrastructure A TCO Analysis of Disk and Tape Solutions (15 July 2015) Report # TCG ) The Impact of LTO-7 on The TCO of Long-Term Storage (15 Sept. 2015) Report #TCG

5 Magnetic tape (r)evolution Product / Year: IBM 726 /1952 LTO7 / 2015 TS1155 /2017 Demo 2017 Capacity: 2.3 MBytes 6 TBytes 15 TBytes 330 TBytes Areal Density: 1400 bit/in Gbit/in Gbit/in Gbit/in 2 Linear Density: 100 bit/in 485 kbit/in 510 kbit/in 818 kbit/in Track Density: 14 tracks/in 8.86 ktracks/in ktracks/in ktracks/in Areal Density >6.8M X 19.8 cm 5

6 HDD Areal Density Scaling: Areal density/capacity scaling achieved by shrinking the same basic technology to write smaller and smaller bits on disk Ref: IBM Corporation 2016 IBM Corporation

7 Noise and Magnetic Media Structure ~10 nm 7 Information is encoded in transition edge. Large grains à media noise To shrink the size of a bit, we need to shrink the size of the grains If grains become too small, magnetic state is unstable à superparamagnetic effect

8 The Superparamagnetic Limit Magnetic Media Trilemma : Small particles (V) SNR µ V HDD has reached the limit of (known) materials to produce larger write fields. Thermal Stability EB µ Ku V Writability H 0 µ K u H 0 < Head Field Technologies to go beyond the superparamagnetic limit: Two dimensional magnetic recording (TDMR) Heat Assisted Magnetic Recording (HAMR) Microwave Assisted Magnetic Recording (MAMR) Bit Patterned Media (BPM) 8

9 Areal Density Scaling 2015: IBM-FujiFilm demonstration of 123 Gb/in 2 on BaFe tape 2017: IBM-Sony demonstration of 201 Gb/in 2 on Sputtered Tape 201 Gbit/in 2 sputtered tape

10 2017 Storage Bit Cells and Extendibility Scaled bit cells: Magnified 25x: NAND Flash (3 bits) 2150 Gb/in nm x 17.3 nm HDD 1000 Gb/in 2 47 nm x 13 nm LTO7 Tape ~4.3 Gb/in nm x 52 nm Jag5A Tape ~9.6 Gb/in nm x 50 nm BaFe Demo 123 Gb/in nm x 37 nm Sputtered Demo 201 Gb/in nm x 31 nm àtremendous potential for future scaling of tape track density àkey technologies: improved track follow servo control improved media, reader, data channel 10

11 Demo Technologies Focus on aggressive track density scaling Require: dramatic improvement in track following à enables track width reduction reduce reader width from a few microns to < 100 nm Ultra narrow reader results in a dramatic loss in read back signal that must be compensated for with improved media technology à require improved writer technology improved signal processing and coding improved reader technology 11

12 Servo pattern design for high areal density demo Increased azimuth angle ð increased resolution Increased pattern density ð increased servo bandwidth and resolution t s H α LTO7 / Jag4 / Jag5 Pattern H = 93 µm, t = 1.25 µm, s = 3 µm α = 12, d = 76 µm 2x angle 1.46x rate d Demo Pattern H = µm, t = 1.0 µm, s = 2.4 µm α = 24, d = 52 µm Servo read-back signal IBM Corporation 2016 IBM Corporation

13 Synchronous servo channel Servo channel decodes the readback signal from the servo pattern and provides position information to the track follow control system Servo channel optimized for p-bafe à improved resolution Optimized servo channel in combination with prototype high SNR media with the 24 demo servo pattern provides nanoscale position information Servo readback signal Servo channel Servo signal ADC Fixed clock frequency Interpolation/ correlation Timingbase reference Optimum symbol detection LPOS symbols Reliability estimate Acquisition, monitoring, and control Lateral-position estimate Tape velocity estimate 13

14 New H track-follow control system Key features Prototype high bandwidth head actuator A speed dependent model of the system delay is used for control design The tape speed is used as a parameter to select the controller coefficients Disturbance rejection is enhanced at the frequencies of the tape path disturbances High Bandwidth Actuator Track-follow control system Actuator Response v tape + PES - v tape K Track-follow controller u y Track-follow actuator G + d LTM - D v tape Delay 14

15 Prototype tape transport & hardware platform Precision flangeless tape path with grooved rollers & pressured air bearings to minimize disturbances TS1140 electronics card for reel-to-reel control and analog front end FPGA Board: System-on-Chip (SoC) -> Servo channels -> Microprocessor for synchronous trackfollow (TF) servo controller FPGA Board Current Driver Servo Readback (LVDS) FPGA Board D/A Serial TF Servo Controller Microprocessor FPU FPGA SoC Servo Servo Servo Servo Channels Channels Channels Channels USB/ ETH Host PC IBM Corporation 2016 IBM Corporation

16 Track-follow performance on BaFe tape Track width computation based on measured position error signal: PES (INSIC method) σpes = standard deviation of position error signal: measure of track following fidelity Track width = 2* 2 * 3*σPES + Reader Width (Reader Width = 90nm) s-pes (nm) σ-pes 5.9 nm over TS1140 speed range tape speed (m/s) Reader Width = 90nm σ-pes 5.9 nm Track width = 140 nm Track density = 181 ktpi 16

17 Advanced BaFe Media Technology: 123 Gb/in 2 demo Key technologies for advanced tape media 1. Fine magnetic particles with high coercivity à archival lifetime 2. Smooth surface 3. Perpendicular orientation of magnetic particles 17

18 SEM Image of tape surface Latest MP tape 123Gb/in 2 demo tape Barium ferrite particles are well isolated and packed with high density IBM Corporation 2016 IBM Corporation

19 Surface profile Optical interferometry roughness Latest MP tape TS1150 JD tape 123Gb/in 2 Demo tape 180 µm 240 µm Ra 2.0nm Ra 1.6nm Ra 0.9nm AFM 40 µm 40 µm Ra 2.4nm Rz 40nm Ra 2.0nm Rz 34nm Ra 1.8nm Rz 27nm Reduced surface roughness of demo tape increases the media SNR IBM Corporation 2016 IBM Corporation

20 Perpendicular orientation Longitudinal orientation (MP tape) Random orientation (TS1140 JC and TS1150 JD tape) Highly perpendicular orientation (123Gb/in 2 demo tape) The perpendicular orientation of BaFe particle provides a strong increase in SNR IBM Corporation 2016 IBM Corporation

21 Read/write performance Gb/in 2 demo tape SNR (db) Latest MP tape TS1150 JD tape Linear density (kfci) The combination of small particle volume, smooth surface and perpendicular BaFe particle orientation provide a major increase in SNR IBM Corporation 2016 IBM Corporation

22 Enhanced Write Field Head Technology Magnetic Media Trilemma : Small particles (V) SNR µ V Thermal Stability EB µ Ku V Writability H 0 µ K u H 0 < Head Field SNRa (db) Std Writer HM Writer IBM developed a new high moment (HM) layered pole write head that produces much larger magnetic fields enabling the use of smaller magnetic particles Coercivity (ka/m) Increasing media coercivity 22

23 Iterative decoding Data C1 Parity Read channel C1 ECC Decoder C2 ECC Decoder C2 Parity 10 0 A user byte-error rate of can be achievable using two C1-C2 iterations with a byte error rate of» at the output of the detector With EPR4 detection byte error rate» 10-2 bit error rate Require SNRa» 10.5 db at the input of the detector to achieve a raw bit error rate < 10-2 at the output of the detector byte error rate N1=240 t1=5 N2=192 t2=12 dr=0 er= channel byte error rate undec C1-o1 C2-o1 C1-o2 C2-o2 C2-o3 capacity

24 Recording performance of BaFe with High moment writer & 90 nm GMR Reader SEM image of GMR reader SNR limit Reader Width = 90nm Byte error rate limit Advanced BaFe supports a linear density of 680kbpi with a 90nm reader and provides an operating margin of ~ 0.5dB SNR 24

25 Summary of BaFe demo results Ø Advanced Perpendicular BaFe medium Ø Linear density = 680 kbpi w/ 90 nm reader (single-channel recording) Ø 1-sigma PES = 5.9 nm, Ø Track density = 181 ktpi (track width = 140 nm) Areal recording density : 123 Gb/in x TS1155 areal density à 220 TB cartridge capacity (*) This demonstration shows that tape technology has the potential for significant capacity increase for years to come! (*) 220 TB cartridge capacity, assuming LTO6 format overheads and taking into account the 48% increase in tape length enabled by the thinner Aramid tape substrate used 25

26 Sputtered Media Technology: 201 Gb/in 2 Demo Key technologies for advanced sputtered tape media 1. Ultra-small grain size with high coercivity à archival lifetime 2. Perpendicular orientation of magnetic grains 3. Advanced sputter technique for low defect density 4. Smooth surface with new lubricant 26

27 Sputtered Media Technology Multi-layer metal film sputtered deposited on base film under vacuum Substrate outgassing and cooling reduce media defects Reel-to-reel sputter coating system TEM X-section of media stack wind-on system polymer film wind-off system to vacuum pump sputtering target (cathode) vacuum chamber cooling drum 10nm DLC CoPtCr-SiO 2 Ru#2 Ru#1 NiW TiCr CoZrNb 27

28 Sputtered Media Technology SEM Top View Atomic force microscope images tape surface roughness: CoPtCr-SiO 2 Ru#2 Ru#1 10nm Average grain size: 6.6 nm, s = 1.2 nm Ra = 0.9 nm Rz = 16 nm Good runability / low friction achieved on ultra smooth sputtered media using bonded lubricant technology 28

29 Recording Performance of Sputtered Media with High moment writer & 48 nm TMR Reader SEM image of TMR reader 13 SNR (db) dB linear density (kbpi) byte-error rate Reader Width = 48nm EPR4 NPML DD-NPML D3-NPML Byte error rate limit w/ iterative decoding linear density (kbpi) Sputtered tape supports a linear density of 818kbpi with a 48nm reader and provides an operating margin of ~ 0.8dB SNR 2013IBM IBMCorporation Corporation 2016

30 Track-follow performance on sputtered tape à 24 servo pattern, synchronous servo channel & demo tape path (a) PES (nm) (b) sigma PES (nm) time (s) σ-pes 6.5 nm over TS1140 speed range tape speed (m/s) Track width = 2* 2 * 3*σPES + Reader Width (INSIC method) 30 Reader Width = 48nm σ-pes 6.5 nm Track width = 103 nm Track density = ktpi

31 Summary of sputtered media demo results Ø Advanced perpendicular sputtered tape Ø Linear density = 818 kbpi w/ 48 nm reader (single-channel recording) Ø 1-sigma PES = 6.5 nm Ø Track density = ktpi (track width = 103 nm) Areal recording density : 201 Gb/in 2 20x TS1155 areal density à 330 TB cartridge capacity (*) This demonstration shows that tape technology has the potential for significant capacity increase for years to come! (*) 330 TB cartridge capacity, assuming TS1155 format overheads and taking into account the 6% increase in tape length enabled by the thinner tape 31

32 INSIC Tape Roadmap Parameter/Year Capacity (TB) % 2. Data rate per channel (MB/sec) % 3. Total data rate (MB/sec) % 4. FC Speed Roadmap (MB/sec)** Number of channels % 6. Tape thickness (um) % 7. Data capacity reserve 3.0% 3.0% 3.0% 3.0% 3.0% 3.0% 8. Tape length that is recordable (meters)*** % 9. Tape length total (meters)**** % 10. Track density (TPI) 10,686 15,652 22,868 33,543 49,372 72, % track pitch = 25400/tpi (um) Linear bit density (kfci)***** % fcmm = kfci/ ,898 22,866 27,668 33,478 40,509 49, Areal density (Gbits/inch2) % 13. Tape speed (m/sec) % 14. Tape width in mm ECC and formatting overhead 22.00% 20.28% 18.69% 17.22% 15.87% 14.63% -4.00% 16. Servo track and layout overhead ****** 15.80% 14.84% 13.46% 12.24% 11.17% 10.22% -6.00% 17. Number of passes to write a tape Number of passes to end-of-life (media) % 19. Time to fill a tape in mins % 20. Number of data tracks 4,481 6,639 9,856 14,660 21,842 32, % 21. Number of data bands % overall head span (um) 3,000 2,268 1,715 1, ~161 Gb/in 2 in 2027 ~215 Gb/in 2 in 2028 INSIC Roadmap available at: 32

33 Summary: The era of big data is creating demand for cost effective storage solutions Tape remains the most cost-efficient and greenest technology for archival storage and active archive applications Tape has a sustainable roadmap for at least another decade BaFe and Sputtered media demos show feasibility of multiple future tape generations Potential exists for the continued of scaling of tape beyond 201 Gbit/in 2 The cost advantage of tape over HDD (and optical disk) will continue to grow 33