AudioVideo Network Convergence with High Capacity Systems

1
Audio/Video Network Convergence with High
Capacity Systems

• Voice Video Engine for Mass Distribution

Amir Zmora VP Business Development Surf
Communication Solutions
2
Presentation Agenda

• Surf Overview
• The Market -- and the Technology that Drives the
  Trends
• Triple-Play Development Considerations
• Hardware Trends

3
Surf Overview

• Surf provides Hardware and Software that enable

large-scale propagation of voice video
applications
large-scale propagation of voice video
applications
4
Surf is the ONLY company in the world that offers
Triple-Play (Voice, Video, Data) media processing
technology simultaneously on a single platform
Uniqueness
Uniqueness
Surf is the first company ever to offer complete
media processing solutions on an AMC platform,
enabling the creation of a Voice and Video engine
driving mass distribution of high-capacity
applications
Surf is the ONLY company in theworld capable of
providing its solutionon ANY TI C-64x DSP
VERY High Barrier To Entry
5
SurfGeneration
6
The Market -- and the Technology that Drives the
Trends
7
Todays Technology
Supporting Infrastructure
Users / Application
enabling
enabling
Conferencing Server
8
Todays Technology
Supporting Infrastructure
Users / Application
enabling
enabling
Conferencing Server
9
Tomorrows / Future Technology
Supporting Infrastructure
Users / Application
Video Calls
Carrier Infrastructure
Video Conference
Voice Video Media Gateway Media Server
IPTV
Mobile Presence
C-6x
Content Streaming
Security/ Surveillance
10
Cellular 3G Services (1)

• Video Download Streaming
• Download is the most common method for video
  entertainment
• Content saved on phone
• Solves QoS issues
• Streaming requires higher bandwidth but has also
  advantages
• More real-time
• Easier in DRM perspective content is not saved
• Phone memory used only for buffering
• Service examples
• V-Live of NTT DoCoMo Real-Time TV
• Video on demand
• Home/Business/Homeland security Video
  surveillance

11
Cellular 3G Services (2)

• Video Streaming Circuit-Switched vs. IP
• IP
• Cost effective
• Easier integration with broadband networks
• Circuit-Switched
• QoS is fixed and predictable
• Well-defined mechanism for addressing and billing
• Easy access Dial number and use DTMF to choose
  clip
• It works

12
Cellular 3G Services (3)

• Messaging Presence
• Evolution IM ? MMS ? Video mail
• PoC
• Presence Not only online/offline but if
  available for service X with person Y
• Self created contentNo DRM issues

13
Cellular 3G Services (4)

• PoC Half-duplex communication (Push-to-Talk,
  Show, View)
• Implemented over IP using SIP
• Many implementations are still proprietary
• Video Telephony
• A 3G service that uses 64Kbps for Control, Voice
  and Video 3G-324M
• Video MPEG-4
• Audio NB-AMR
• Adopted by 3Gpp and 3GPP2 but deployed currently
  on W-CDMA networks only

14
Cellular 3G Services (5)

• Video Telephony Issues
• User adoption is still limited
• Network affect will make it more popular in the
  future (need someone to call)
• Quality is not high 10FPS, QCIF
• Call setup delay

15
Cellular 3G Services (6)

• Video Sharing
• Feature introduced by Nokia
• While in voice circuit switch session, opens a
  parallel SIP session
• Shares live video, clips, pictures
• Effective bandwidth usage (IP vs. CS)

16
Nokia Video Sharing Demo
Ongoing CS-call
User A
User B
Initiate sharing session
User accepts invitation. Video sharing session
is established.
User A camera view displayed
Select Share live video from options
Pressing Start begins video sharing
Pressing Stop ends sharing
17
(No Transcript)
18
Conclusions

• Voice price is shrinking
• Infrastructure Voice port price
• Operators Voice revenue (International/Long
  distance/Local)
• Moving to one box solution for voice/video,
  Mobile/Wireless/Wireline - All in one - Siemens
  Wireline Wireless merge (ICNICMCom)
• Moving into the video market is essential for
  telecom equipment manufacturers
• Video deployment will dramatically grow due to
• Competition between service providers/operators/ca
  bles
• Cross network services
• Network effect

19
Triple-Play Development Considerations
20
Typical Media Gateway / Media Server
Signaling
Transport
Media Processing
21
Voice and Video on the Same DSP

• Easy migration path to any video/voice usage
  model.
• Lower system delay generates a higher overall
  quality
• No need to separate voice from video and send
  each to a separate chip/module
• Easier to establish lip-sync between voice and
  video when packets are lost or delayed
• Lower operational expenses due to the need for
  support of only one chip.
• Lower integration efforts
• Single API for voice and video
• No need for an aggregation device to combine or
  split voice and video

22
The benefits of Running 3G-324M (H.223) on the
DSP rather than on the Host ?

• Lower delay generates higher overall quality
  gateways
• Same DSP runs 3G-324M, voice transcoding, video
  transcoding and sends voice/video RTP directly to
  IP network, hence, no bottlenecks.
• Scalability
• When H.223 is running on the DSP, Host processor
  performance is not a limiting factor for moving
  to higher densities just add more DSPs/boards.
• Lower operational expenses
• Less components in the system. No need for
• DSP board/module for voice transcoding
• DSP board/module for video transcoding
• Controller module/board for 3G-324M

23
The benefits of building a Gateway rather than
buying a closed box

• Flexibility -The customer can determine what
  features to develop and can better differentiate
  his offerings.
• Better response to the end user since no need to
  wait for the closed box developer to provide his
  service.
• Open solution.
• Price.

24
Hardware Trends
25
Relevant Hardware History

• Once there were PC cPCI systems
• Parallel passive backplanes
• Then came cPCI 2.16, 2.17, 2.18
• Combination of parallel bus multiple
  packet-switching serial busses
• Then came ATCA (Advanced Telecom Computer
  Architecture)
• Packet-switching serial busses only
• Late 2001 - Suppliers, OEMs and SPs define the
  AdvancedTCA (PICMG 3.0)
• Late 2002 Start work on AdvancedMC (PICMG
  AMC.0)
• Late 2005 First MicroTCA prototypes

26
Hardware Forecasts

• Varying forecasts for ATCA AMC market size
• ATCA market to reach 3.7B by 2007 (RHK)
• ATCA market to reach 42B by 2009 (Crystal Cube)
• AMC market to reach 14B by 2009 (Crystal Cube)
• Target cost per board to be reduced by 60 in
  four years, using ATCA (not proprietary)
• Cheaper RD
• Open competition
• Mass production
• Maintenance
• etc.

27
Hardware History
1981
First PC
1994
cPCI introduced
2001
cPCI 2.16 introduced First ATCA definition
2002
First AMC definition
2005
First ATCA prototypes First MicroTCA
prototypes First AMC prototypes

• Surfs Current Offering
• PCI solutions
• PTMC solutions
• AMC solutions
• cPCI integrated solutions

28
Market Hardware Trends Carriers
PCI
cPCI 2.16
ATCA
100Mbs per board
100Mbs per board
1Gbit/sec per link
1Gbit/sec per link
10Gbit/sec per link
PCI
PCI
PCI, Ethernet
PCI, Ethernet
GbE, PCI_Express, Rapid_IO, Infiniband,FiberChanne
l Advanced switching etc
N/A
N/A
IPMI
IPMI
I2C based redundant IPMI
Not supported
Not supported
Blade only
Blade only
Blade and mezzanine (AMC)
25W per blade
25W per blade
70W per blade
70W per blade
200W per blade
N/A
N/A
3-5 9S
5 9S
3-5 9S
1 mezzanine board
1 mezzanine board
2 mezzanine boards
8 mezzanine boards
2 mezzanine boards
PCI
PCI
3U/6U x 160mm
8U x 280mm
3U/6U x 160mm
29
Market Hardware Trends Mezzanines
PTMC
AMC
100Mbit/sec
100Mbit/sec
10Gbit/sec per link
PCI, TDM, RMII
PCI, TDM, RMII
GbE, PCI_Express, Rapid_IO, Infiniband,FiberChanne
l Advanced switching etc
Not supported
Not supported
I2C based IPMI
Not supported
Not supported
Supported
15W max
15W max
60W max
3 9S to 5 9S
3 9S to 5 9S
5 9S
Up to 8 per ATCA
Up to 4 per ATCA
Up to 4 per ATCA
PTMC
PTMC
AMC
30
Market Hardware TrendsSystem
ATCA
PC
cPCI
100Mbs per board
100Mbs per board
Up to 40Gb/s per chassis
Up to 10 Tera per chassis
PCI
PCI
PCI, Ethernet
PCI, Ethernet
GbE, PCI_Express, Rapid_IO, Infiniband,FiberChanne
l Advanced switching etc
Not supported
Not supported
IPMI
IPMI
I2C based redundant IPMI
Not supported
Not supported
Blade only
Blade only
Blade and mezzanine (AMC)
25W per blade
25W per blade
50W per blade
50W per blade
200W per blade
5 9S
3 9S to 5 9s
3 9S to 5 9s
N/A
N/A
Up to 19 blades
Up to 19 blades
Up to 16 blades
8 boards
8 boards
6U
6U
8U
PC
PC
31
AMC System for the Poor MicroTCA
ATCA
MicroTCA
Up to 40Gbits/s per chassis
Up to 40Gbits/s per chassis
Up to 10 TeraBits/s per chassis
GbE, PCI_Express, Rapid_IO, Infiniband,FiberChanne
l Advanced switching etc
GbE, PCI_Express, Rapid_IO, Infiniband,FiberChanne
l Advanced switching etc
GbE, PCI_Express, Rapid_IO, Infiniband,FiberChanne
l Advanced switching etc
I2C based redundant IPMI
I2C based redundant IPMI
I2C based redundant IPMI
Supported per blade
Supported per blade
Supported per blade
Up to 60W per blade
Up to 60W per blade
Up to 200W per blade
Scalable for three 9s to five 9s
Scalable for three 9s to five 9s
five 9S
Up to 12 AMC modules
Up to 12 AMC modules
Up to 16 ATCA blades (128 AMCs)
4Ux19 and only 300mm deep
4Ux19 and only 300mm deep
12Ux19
Entry point of 500
Entry point of 500
Entry point of several 1000s
32
Bottom Line

• ATCA/MicroTCA platforms with AMC modules are the
  future
• Surf is aligned with the market hardware needs
  for high-capacity DSP media processing solutions

33
Thank You
www.surf-com.com