SlideShare a Scribd company logo
1 of 17
○ May 9, 2016
Enhancing Data Center Flash Storage
Endurance Using Network-on-Chip
Interconnect IP
ChipEx2016, Israel
Xavier van Ruymbeke
Application Engineering Manager, Arteris
xavier.van-ruymbeke@arteris.com@arteris.com
Copyright © 2016 Arteris
○ May 9, 2016
Arteris Snapshot
Copyright © 2016 Arteris
Founded 2003; Headquarters in Silicon Valley
Global PresenceInterconnect IP Leadership
Large&GrowingCustomerBase
Customersshipped ~600M unitsin 2014
Connected byArteris Ecosystem
Shanghai
TokyoCampbell
Seoul
TaipeiBangalore
Paris
Israel
Strong R&D and Support
R&D
Engineering
48%App
Engineering
29%
Sales
14%
G&A
9%
Marketing
2%
Top Semis use Arteris
Publicly Disclosed Customers
Toshiba
In Leading Systems
Over 100
NoC-based
SoCs in
production
1
6 9
13
20
41
52
58
67
76 80
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
2
NoCSolution 2006
Main Interconnect 1st
Generation
FlexNoC 2010
Main Interconnect 2nd
Generation
FlexWay 2010 Subsystem Interconnect
FlexLLI 2011 Interchip Link
FlexPSI 2013 Interchip Link
FlexNoC
Resilience
2014 For Mission Critical
Electronics
FlexNoC
Physical
2015
Physical Awareness
w/FlexNoC V3
○ May 9, 2016
Need for on-chip data protection
emerging in data center flash controllers
Copyright © 2016 Arteris
Data Center SSD
○ Huge bandwidth – Logical Block Addressing (LBA) and Cascading Slices
to access 10+ TB
○ Low latency – NVMe replaces AHCI
○ Updated standards: JEDEC JESD218B (Endurance), SNIA Solid State
Storage (SSS) Performance Test Specification (PTS)
○ Result: More complexity, more gates…
• But on-chip errors reduce SSD endurance and reliability
More complex systems = Data protection required!
3
○ May 9, 2016
Data center SSD controller complexity
requires larger, more complex interconnect
Copyright © 2016 Arteris
Data RAM Subsystem
(Buffers, Parity,
Sequence)
DMA Engine
FlexNoC® Top Level Interconnect
Peripheral
Subsystem
.
SPI
I2C
UART
GPIO
ROM
On-chip interconnect IP
DRAM Subsystem
DDR3/4 DDR3/4
PHY PHY
Memory Scheduler
DRAM Controller
XOR Engine ECC
Flash
Control
CPU Subsystem
R5/7
R5/7
R5/7
R5/7
R5/7
R5/7
Cache
LLPP
NVMe
PCIe Subsystem
.
.
Interconnect
Timer
Debug
PCIePHY
Data
RAM
Data
RAM
Data
RAM
Data
RAM
SRAM RAID Flash
4
○ May 9, 2016
Controller complexity and size increase
probability of digital logic failures
Copyright © 2016 Arteris
Particle hitting
the gate and
changing its
output
Wire broken:
Un-driven input
Destroyed Gate: Stuck at
Data center flash controllers must adapt to inevitable errors
○ Digital integrated circuits can fail in
many ways:
• Transient electrical problems –
glitches on power supply, clock
supply
• Soft errors – Alpha particles,
cosmic rays, thermal neutrons
• Physical damage – shorts or open
wires, destroyed logic gates
○ Designing for data protection requires
covering all of them
• Not just Soft Errors
5
○ May 9, 2016
Why implement data protection in the
SoC interconnect?
Implementing protection in the interconnect is better because…
1. Complexity – SW is much more effort to develop and maintain than using
certifiable HW IP
2. Quality – HW IP is pre-tested and can be pre-certified while future SW
quality is a risk
3. Control – Semi vendors can set a baseline of data protection features that
must be implemented, so they are at less risk of poor quality software
developed without their knowledge, input or feedback.
Copyright © 2016 Arteris
Implementing data protection in hardware is less risky than
implementing solely in software!
6
○ May 9, 2016
Protecting on-chip interconnect data
makes the SoC more reliable
Each feature is optional and configurable:
 Parity on packet header in transport
• Cover faults on packet content in transport
layer
 Packet Consistency checkers
• Capture bad routing or partial packet
deletion/duplication in transport
 User defined ECC/parity on payload
• Cover all faults affecting payload
• Can be input/output from socket or
generate/terminate
 Initiator timeout
• Capture single packet deletion or Initiator
misbehavior
 Target timeout
• Capture slave misbehavior
 ARM Cortex® R5/R7 support
 Register parity bit
• Cover Faults on NoC configuration register
Copyright © 2016 Arteris
NoC=
?
=
?
=
?
=
?
=
?
=
?
=
?
=
? NIUNIUNIUNIU
NIUNIUNIUNIU
Master
IP
Master
IP
Master
IP
Slave
IP
Slave
IP
Slave
IP
Master
IP
Slave
IP
NIUNIUNIU
NIUNIUNIU
NIU
NIU
SafetyController
Fault
cc cc cc cc
cc cc cc cc
ECC/P ECC/P ECC/P ECC/P
ECC/P ECC/P ECC/P ECC/P
7
○ May 9, 2016
Data protection advanced features
 Unit duplication and checkers
• Cover faults in NIU, firewall, power
disconnect elements able to transform
the access content
 Safety Controller
• Fault reporting and BIST checking for
duplicated unit checkers
Copyright © 2016 Arteris
NoC=
?
=
?
=
?
=
?
=
?
=
?
=
?
=
? NIUNIUNIUNIU
NIUNIUNIUNIU
Master
IP
Master
IP
Master
IP
Slave
IP
Slave
IP
Slave
IP
Master
IP
Slave
IP
NIUNIUNIU
NIUNIUNIU
NIU
NIU
SafetyController
Fault
cc cc cc cc
cc cc cc cc
ECC/P ECC/P ECC/P ECC/P
ECC/P ECC/P ECC/P ECC/P
These are effective for unique
use cases, but are more
costly in power consumption
and die area.
8
○ May 9, 2016
Protect SoC data with end-to-end data
protection
○ Core Interface Protection supports
• User-defined protection schemes
○ To ensure safe connection from core to
NoC, core-side protection logic
• Detects and corrects errors upon
transaction requests
• Generates protection semantics to IP
core upon transaction responses
○ Packet transport protection logic
 Detects and corrects transport-level
errors
Copyright © 2016 Arteris
Protected IP Core
Resilient
NoC
Core-side
Error Detection
& Correction
Protected
Core
Interface
Transport
Parity/ECC
Transport
Parity/ECC
Core-side
Protection
Generation
Request Response
9
○ May 9, 2016
Simplest, lowest cost option is parity
protection
○ Inserted and terminated by NIUs and units that modify packets
○ Packet Header protection
• One parity bit per group of header bits
• EVEN parity
• Group size determined by “TimingImpact” parameter
 Low: 8 bits (more number of resilience bits in header)
 Medium: 16 bits
 High: 32 bits (less number of resilience bits in header)
○ Packet payload protection
• One parity bit per data byte (including byte enable)
Copyright © 2016 Arteris
Group N
N Groups
Group N-1Group N-2Group 2Group 1Group 0
0/1 0/1 0/1 0/1 0/1 0/1
Header
Width = N Groups
even parity bit
Calculation
one bit per Group
…
10
○ May 9, 2016
Protect ARM Cortex-R5/R7 transactions
with built-in interconnect ECC support
○ Control
• All command redundancy is generated/terminated
• 32/64 bits AXI socket support
• ODD/EVEN parity support
○ Data
• Cortex ECC terminated/generated in the Initiator
NIU
• Optional byte-level ECC generation inside NoC
○ Option to do network interface unit (NIU)
duplication
NIUinitiator
Cortex
NIUinitiator
S2G
G2T
ECC
checks
NIUTarget
G2T
S2G
ECC
checks
transport
Cortex ECC
terminated, Byte
ECC generation
Byte ECC
terminated, custom
ECC generation
Copyright © 2016 Arteris 11
○ May 9, 2016
Implement user-defined payload ECC to
protect other data
○ Per-socket, user-defined number of DataInfo bits per
bytes
○ NoC wide, transport level user-defined number of
DataInfo bits per bytes
○ Conversion to/from socket DataInfo semantics from/to
Transport DataInfo semantics
• Done inside the NIUs at generic level
• If socket and transport DataInfo are identical (same number of bits
per bytes, same semantic): can be just copied transparently
• If socket and transport DataInfo are different (different number of
bits per bytes, and/or different semantic): conversion logic need to
be provided by the user and will be located in NIU placeholders
○ Socket Protocols supporting ECC:
• AXI – with narrow bursts split to single word transactions
• AHB – with AHB target using Byte Enables
• OCP – with targets providing correct ECC on SData/SDataInfo
for write responses
NIUinitiator
NIUinitiator
S2G
G2T
Socket <>
Transport
NIUTarget
G2T
S2G
Transport
<> Socket
transport
Socket defined
DataInfo bits per
byte
Socket defined
DataInfo bits per
bytes
NoC transport
number of bits per
payload byte
Conversion
placeholder
Conversion
placeholder
Copyright © 2016 Arteris 12
○ May 9, 2016
Duplicate parts of the interconnect to achieve
higher reliability
○ Generic and specific NIUs are duplicated
• Includes reorder buffers, scheduler – no restrictions
○ Units modifying packets are duplicated
• Target power disconnect, Firewall
○ Duplicate logic delayed by one or two cycle
• Using separate clock input and reset input
• Avoid common faults due to power glitches and clock
branch defects
○ A Checker is used to compare outputs
• Checker has separate clock gater
○ Differences in output are reported as faults
○ Some unit configurations cannot be duplicated
• NIUs with multiple generic interfaces (LLI, OCP MT)
• Target power disconnect with asynchronous crossing
• These units can be present in the system but won’t be
duplicated
Reference
Unit
Functional
Unit
Delay
Inputs
Checker
Outputs
ClkClk’
FaultBIST
Control
Copyright © 2016 Arteris 13
○ May 9, 2016
Need checkers, fault reporting and BIST
to ensure data protection
Copyright © 2016 Arteris
• Sequence testing
patterns through
delay registers
• Report issues in
the comparison
logic (stuck-at
faults typically)
○ One Checker per duplicated unit
○ In Mission mode: Compares Unit outputs
• Early Unit outputs delayed by one cycle
• Early/Late Unit regular outputs bit-compared
• Logic fault outputs from both units compared to 0
• Redundant report of packet and/or NIU check faults
○ In BIST mode: Checks comparison logic)
Safety Controller
Checker
Checker
Clk
Clk’
Fault
BIST control
BIST control
Mission
Interrupt
Latent
Interrupt
Clk
….
14
○ May 9, 2016
Protect only the parts of controller that affect
system reliability the most
Accelerators
Peripheral
Data-Protected
NoC
I/OMem Ctrl Peripheral
DRAM
ROM/Flash
CPUs
Peripheral
I/O
ECC - Core
SafetyController
=?
=?
=?
=? =?
Fault
cccccc
cc cc
cc Consistency checker Equality checker
Fault
NoC
Without
Protection
Firewall (SW programmable)
=?
Copyright © 2016 Arteris 15
○ May 9, 2016
Data center and SSD flash controllers
require on-chip data protection
○ Interconnect size as a percentage of die size, and complexity,
increase as the number of functional IP blocks increases
○ Data traveling on the interconnect is at risk of corruption
○ Technologies exist today to protect data traveling over the on-chip
interconnect, which increases system reliability and endurance
Copyright © 2016 Arteris
Meeting data center storage reliability goals requires on-chip
interconnect data protection
16
○ May 9, 2016
Copyright © 2016 Arteris 17

More Related Content

Similar to Xavier van Ruymbeke, App. Engineer, Arteris

Building an open memory-centric computing architecture using intel optane
Building an open memory-centric computing architecture using intel optaneBuilding an open memory-centric computing architecture using intel optane
Building an open memory-centric computing architecture using intel optaneUniFabric
 
The Considerations for Internet of Things @ 2017
The Considerations for Internet of Things @ 2017The Considerations for Internet of Things @ 2017
The Considerations for Internet of Things @ 2017Jian-Hong Pan
 
Singapore_IEEE_power_self-supply_based_on_an_io_t_driven_ platform
Singapore_IEEE_power_self-supply_based_on_an_io_t_driven_ platformSingapore_IEEE_power_self-supply_based_on_an_io_t_driven_ platform
Singapore_IEEE_power_self-supply_based_on_an_io_t_driven_ platformFrank Alexander Reusch
 
Improving performance and efficiency with Network Virtualization Overlays
Improving performance and efficiency with Network Virtualization OverlaysImproving performance and efficiency with Network Virtualization Overlays
Improving performance and efficiency with Network Virtualization OverlaysAdam Johnson
 
Affordable trustworthy-systems
Affordable trustworthy-systemsAffordable trustworthy-systems
Affordable trustworthy-systemsmicrokerneldude
 
PoC Requirements and Use Cases
PoC Requirements and Use CasesPoC Requirements and Use Cases
PoC Requirements and Use Casesjennimenni
 
ODSA - PoC Requirements and Use Cases
ODSA - PoC Requirements and Use CasesODSA - PoC Requirements and Use Cases
ODSA - PoC Requirements and Use CasesODSA Workgroup
 
DPDK Summit - 08 Sept 2014 - NTT - High Performance vSwitch
DPDK Summit - 08 Sept 2014 - NTT - High Performance vSwitchDPDK Summit - 08 Sept 2014 - NTT - High Performance vSwitch
DPDK Summit - 08 Sept 2014 - NTT - High Performance vSwitchJim St. Leger
 
Achieving Network Deployment Flexibility with Mirantis OpenStack
Achieving Network Deployment Flexibility with Mirantis OpenStackAchieving Network Deployment Flexibility with Mirantis OpenStack
Achieving Network Deployment Flexibility with Mirantis OpenStackEric Zhaohui Ji
 
“Using a Neural Processor for Always-sensing Cameras,” a Presentation from Ex...
“Using a Neural Processor for Always-sensing Cameras,” a Presentation from Ex...“Using a Neural Processor for Always-sensing Cameras,” a Presentation from Ex...
“Using a Neural Processor for Always-sensing Cameras,” a Presentation from Ex...Edge AI and Vision Alliance
 
NSA Capstone Presentation
NSA Capstone PresentationNSA Capstone Presentation
NSA Capstone PresentationMinh Vu
 
CIF16: Building the Superfluid Cloud with Unikernels (Simon Kuenzer, NEC Europe)
CIF16: Building the Superfluid Cloud with Unikernels (Simon Kuenzer, NEC Europe)CIF16: Building the Superfluid Cloud with Unikernels (Simon Kuenzer, NEC Europe)
CIF16: Building the Superfluid Cloud with Unikernels (Simon Kuenzer, NEC Europe)The Linux Foundation
 
Open vSwitch Implementation Options
Open vSwitch Implementation Options Open vSwitch Implementation Options
Open vSwitch Implementation Options Netronome
 
Webinar: Faster, Smarter, Simpler – The New Requirements in Storage Networking
Webinar: Faster, Smarter, Simpler – The New Requirements in Storage NetworkingWebinar: Faster, Smarter, Simpler – The New Requirements in Storage Networking
Webinar: Faster, Smarter, Simpler – The New Requirements in Storage NetworkingStorage Switzerland
 
System On Chip (SOC)
System On Chip (SOC)System On Chip (SOC)
System On Chip (SOC)Shivam Gupta
 
"Accelerating Deep Learning Using Altera FPGAs," a Presentation from Intel
"Accelerating Deep Learning Using Altera FPGAs," a Presentation from Intel"Accelerating Deep Learning Using Altera FPGAs," a Presentation from Intel
"Accelerating Deep Learning Using Altera FPGAs," a Presentation from IntelEdge AI and Vision Alliance
 
Sagar Kadam, Lead Software Engineer, Open-Silicon
Sagar Kadam, Lead Software Engineer, Open-SiliconSagar Kadam, Lead Software Engineer, Open-Silicon
Sagar Kadam, Lead Software Engineer, Open-Siliconchiportal
 
CisCon 2018 - Analytics per Storage Area Networks
CisCon 2018 - Analytics per Storage Area NetworksCisCon 2018 - Analytics per Storage Area Networks
CisCon 2018 - Analytics per Storage Area NetworksAreaNetworking.it
 

Similar to Xavier van Ruymbeke, App. Engineer, Arteris (20)

Building an open memory-centric computing architecture using intel optane
Building an open memory-centric computing architecture using intel optaneBuilding an open memory-centric computing architecture using intel optane
Building an open memory-centric computing architecture using intel optane
 
The Considerations for Internet of Things @ 2017
The Considerations for Internet of Things @ 2017The Considerations for Internet of Things @ 2017
The Considerations for Internet of Things @ 2017
 
Singapore_IEEE_power_self-supply_based_on_an_io_t_driven_ platform
Singapore_IEEE_power_self-supply_based_on_an_io_t_driven_ platformSingapore_IEEE_power_self-supply_based_on_an_io_t_driven_ platform
Singapore_IEEE_power_self-supply_based_on_an_io_t_driven_ platform
 
Improving performance and efficiency with Network Virtualization Overlays
Improving performance and efficiency with Network Virtualization OverlaysImproving performance and efficiency with Network Virtualization Overlays
Improving performance and efficiency with Network Virtualization Overlays
 
Affordable trustworthy-systems
Affordable trustworthy-systemsAffordable trustworthy-systems
Affordable trustworthy-systems
 
PoC Requirements and Use Cases
PoC Requirements and Use CasesPoC Requirements and Use Cases
PoC Requirements and Use Cases
 
ODSA - PoC Requirements and Use Cases
ODSA - PoC Requirements and Use CasesODSA - PoC Requirements and Use Cases
ODSA - PoC Requirements and Use Cases
 
DPDK Summit - 08 Sept 2014 - NTT - High Performance vSwitch
DPDK Summit - 08 Sept 2014 - NTT - High Performance vSwitchDPDK Summit - 08 Sept 2014 - NTT - High Performance vSwitch
DPDK Summit - 08 Sept 2014 - NTT - High Performance vSwitch
 
s2000nano
s2000nanos2000nano
s2000nano
 
Achieving Network Deployment Flexibility with Mirantis OpenStack
Achieving Network Deployment Flexibility with Mirantis OpenStackAchieving Network Deployment Flexibility with Mirantis OpenStack
Achieving Network Deployment Flexibility with Mirantis OpenStack
 
“Using a Neural Processor for Always-sensing Cameras,” a Presentation from Ex...
“Using a Neural Processor for Always-sensing Cameras,” a Presentation from Ex...“Using a Neural Processor for Always-sensing Cameras,” a Presentation from Ex...
“Using a Neural Processor for Always-sensing Cameras,” a Presentation from Ex...
 
NSA Capstone Presentation
NSA Capstone PresentationNSA Capstone Presentation
NSA Capstone Presentation
 
CIF16: Building the Superfluid Cloud with Unikernels (Simon Kuenzer, NEC Europe)
CIF16: Building the Superfluid Cloud with Unikernels (Simon Kuenzer, NEC Europe)CIF16: Building the Superfluid Cloud with Unikernels (Simon Kuenzer, NEC Europe)
CIF16: Building the Superfluid Cloud with Unikernels (Simon Kuenzer, NEC Europe)
 
Open vSwitch Implementation Options
Open vSwitch Implementation Options Open vSwitch Implementation Options
Open vSwitch Implementation Options
 
Webinar: Faster, Smarter, Simpler – The New Requirements in Storage Networking
Webinar: Faster, Smarter, Simpler – The New Requirements in Storage NetworkingWebinar: Faster, Smarter, Simpler – The New Requirements in Storage Networking
Webinar: Faster, Smarter, Simpler – The New Requirements in Storage Networking
 
System On Chip (SOC)
System On Chip (SOC)System On Chip (SOC)
System On Chip (SOC)
 
"Accelerating Deep Learning Using Altera FPGAs," a Presentation from Intel
"Accelerating Deep Learning Using Altera FPGAs," a Presentation from Intel"Accelerating Deep Learning Using Altera FPGAs," a Presentation from Intel
"Accelerating Deep Learning Using Altera FPGAs," a Presentation from Intel
 
Sagar Kadam, Lead Software Engineer, Open-Silicon
Sagar Kadam, Lead Software Engineer, Open-SiliconSagar Kadam, Lead Software Engineer, Open-Silicon
Sagar Kadam, Lead Software Engineer, Open-Silicon
 
CisCon 2018 - Analytics per Storage Area Networks
CisCon 2018 - Analytics per Storage Area NetworksCisCon 2018 - Analytics per Storage Area Networks
CisCon 2018 - Analytics per Storage Area Networks
 
HOW Series: Knights Landing
HOW Series: Knights LandingHOW Series: Knights Landing
HOW Series: Knights Landing
 

More from chiportal

Prof. Zhihua Wang, Tsinghua University, Beijing, China
Prof. Zhihua Wang, Tsinghua University, Beijing, China Prof. Zhihua Wang, Tsinghua University, Beijing, China
Prof. Zhihua Wang, Tsinghua University, Beijing, China chiportal
 
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...chiportal
 
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...chiportal
 
Prof. Uri Weiser,Technion
Prof. Uri Weiser,TechnionProf. Uri Weiser,Technion
Prof. Uri Weiser,Technionchiportal
 
Ken Liao, Senior Associate VP, Faraday
Ken Liao, Senior Associate VP, FaradayKen Liao, Senior Associate VP, Faraday
Ken Liao, Senior Associate VP, Faradaychiportal
 
Prof. Danny Raz, Director, Bell Labs Israel, Nokia
 Prof. Danny Raz, Director, Bell Labs Israel, Nokia  Prof. Danny Raz, Director, Bell Labs Israel, Nokia
Prof. Danny Raz, Director, Bell Labs Israel, Nokia chiportal
 
Marco Casale-Rossi, Product Mktg. Manager, Synopsys
Marco Casale-Rossi, Product Mktg. Manager, SynopsysMarco Casale-Rossi, Product Mktg. Manager, Synopsys
Marco Casale-Rossi, Product Mktg. Manager, Synopsyschiportal
 
Dr.Efraim Aharoni, ESD Leader, TowerJazz
Dr.Efraim Aharoni, ESD Leader, TowerJazzDr.Efraim Aharoni, ESD Leader, TowerJazz
Dr.Efraim Aharoni, ESD Leader, TowerJazzchiportal
 
Eddy Kvetny, System Engineering Group Leader, Intel
Eddy Kvetny, System Engineering Group Leader, IntelEddy Kvetny, System Engineering Group Leader, Intel
Eddy Kvetny, System Engineering Group Leader, Intelchiportal
 
Dr. John Bainbridge, Principal Application Architect, NetSpeed
 Dr. John Bainbridge, Principal Application Architect, NetSpeed  Dr. John Bainbridge, Principal Application Architect, NetSpeed
Dr. John Bainbridge, Principal Application Architect, NetSpeed chiportal
 
Asi Lifshitz, VP R&D, Vtool
Asi Lifshitz, VP R&D, VtoolAsi Lifshitz, VP R&D, Vtool
Asi Lifshitz, VP R&D, Vtoolchiportal
 
Zvika Rozenshein,General Manager, EngineeringIQ
Zvika Rozenshein,General Manager, EngineeringIQZvika Rozenshein,General Manager, EngineeringIQ
Zvika Rozenshein,General Manager, EngineeringIQchiportal
 
Lewis Chu,Marketing Director,GUC
Lewis Chu,Marketing Director,GUC Lewis Chu,Marketing Director,GUC
Lewis Chu,Marketing Director,GUC chiportal
 
Kunal Varshney, VLSI Engineer, Open-Silicon
Kunal Varshney, VLSI Engineer, Open-SiliconKunal Varshney, VLSI Engineer, Open-Silicon
Kunal Varshney, VLSI Engineer, Open-Siliconchiportal
 
Gert Goossens,Sen. Director, ASIP Tools, Synopsys
Gert Goossens,Sen. Director, ASIP Tools, SynopsysGert Goossens,Sen. Director, ASIP Tools, Synopsys
Gert Goossens,Sen. Director, ASIP Tools, Synopsyschiportal
 
Tuvia Liran, Director of VLSI, Nano Retina
Tuvia Liran, Director of VLSI, Nano RetinaTuvia Liran, Director of VLSI, Nano Retina
Tuvia Liran, Director of VLSI, Nano Retinachiportal
 
Ronen Shtayer,Director of ASG Operations & PMO, NXP Semiconductor
Ronen Shtayer,Director of ASG Operations & PMO, NXP SemiconductorRonen Shtayer,Director of ASG Operations & PMO, NXP Semiconductor
Ronen Shtayer,Director of ASG Operations & PMO, NXP Semiconductorchiportal
 
Prof. Emanuel Cohen, Technion
Prof. Emanuel Cohen, TechnionProf. Emanuel Cohen, Technion
Prof. Emanuel Cohen, Technionchiportal
 
Prof. Ramez Daniel, Technion
Prof. Ramez Daniel, TechnionProf. Ramez Daniel, Technion
Prof. Ramez Daniel, Technionchiportal
 
Rotem Ben-Hur,Graduate Student,Technio
Rotem Ben-Hur,Graduate Student,TechnioRotem Ben-Hur,Graduate Student,Technio
Rotem Ben-Hur,Graduate Student,Techniochiportal
 

More from chiportal (20)

Prof. Zhihua Wang, Tsinghua University, Beijing, China
Prof. Zhihua Wang, Tsinghua University, Beijing, China Prof. Zhihua Wang, Tsinghua University, Beijing, China
Prof. Zhihua Wang, Tsinghua University, Beijing, China
 
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...
 
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...
Prof. Steve Furber, University of Manchester, Principal Designer of the BBC M...
 
Prof. Uri Weiser,Technion
Prof. Uri Weiser,TechnionProf. Uri Weiser,Technion
Prof. Uri Weiser,Technion
 
Ken Liao, Senior Associate VP, Faraday
Ken Liao, Senior Associate VP, FaradayKen Liao, Senior Associate VP, Faraday
Ken Liao, Senior Associate VP, Faraday
 
Prof. Danny Raz, Director, Bell Labs Israel, Nokia
 Prof. Danny Raz, Director, Bell Labs Israel, Nokia  Prof. Danny Raz, Director, Bell Labs Israel, Nokia
Prof. Danny Raz, Director, Bell Labs Israel, Nokia
 
Marco Casale-Rossi, Product Mktg. Manager, Synopsys
Marco Casale-Rossi, Product Mktg. Manager, SynopsysMarco Casale-Rossi, Product Mktg. Manager, Synopsys
Marco Casale-Rossi, Product Mktg. Manager, Synopsys
 
Dr.Efraim Aharoni, ESD Leader, TowerJazz
Dr.Efraim Aharoni, ESD Leader, TowerJazzDr.Efraim Aharoni, ESD Leader, TowerJazz
Dr.Efraim Aharoni, ESD Leader, TowerJazz
 
Eddy Kvetny, System Engineering Group Leader, Intel
Eddy Kvetny, System Engineering Group Leader, IntelEddy Kvetny, System Engineering Group Leader, Intel
Eddy Kvetny, System Engineering Group Leader, Intel
 
Dr. John Bainbridge, Principal Application Architect, NetSpeed
 Dr. John Bainbridge, Principal Application Architect, NetSpeed  Dr. John Bainbridge, Principal Application Architect, NetSpeed
Dr. John Bainbridge, Principal Application Architect, NetSpeed
 
Asi Lifshitz, VP R&D, Vtool
Asi Lifshitz, VP R&D, VtoolAsi Lifshitz, VP R&D, Vtool
Asi Lifshitz, VP R&D, Vtool
 
Zvika Rozenshein,General Manager, EngineeringIQ
Zvika Rozenshein,General Manager, EngineeringIQZvika Rozenshein,General Manager, EngineeringIQ
Zvika Rozenshein,General Manager, EngineeringIQ
 
Lewis Chu,Marketing Director,GUC
Lewis Chu,Marketing Director,GUC Lewis Chu,Marketing Director,GUC
Lewis Chu,Marketing Director,GUC
 
Kunal Varshney, VLSI Engineer, Open-Silicon
Kunal Varshney, VLSI Engineer, Open-SiliconKunal Varshney, VLSI Engineer, Open-Silicon
Kunal Varshney, VLSI Engineer, Open-Silicon
 
Gert Goossens,Sen. Director, ASIP Tools, Synopsys
Gert Goossens,Sen. Director, ASIP Tools, SynopsysGert Goossens,Sen. Director, ASIP Tools, Synopsys
Gert Goossens,Sen. Director, ASIP Tools, Synopsys
 
Tuvia Liran, Director of VLSI, Nano Retina
Tuvia Liran, Director of VLSI, Nano RetinaTuvia Liran, Director of VLSI, Nano Retina
Tuvia Liran, Director of VLSI, Nano Retina
 
Ronen Shtayer,Director of ASG Operations & PMO, NXP Semiconductor
Ronen Shtayer,Director of ASG Operations & PMO, NXP SemiconductorRonen Shtayer,Director of ASG Operations & PMO, NXP Semiconductor
Ronen Shtayer,Director of ASG Operations & PMO, NXP Semiconductor
 
Prof. Emanuel Cohen, Technion
Prof. Emanuel Cohen, TechnionProf. Emanuel Cohen, Technion
Prof. Emanuel Cohen, Technion
 
Prof. Ramez Daniel, Technion
Prof. Ramez Daniel, TechnionProf. Ramez Daniel, Technion
Prof. Ramez Daniel, Technion
 
Rotem Ben-Hur,Graduate Student,Technio
Rotem Ben-Hur,Graduate Student,TechnioRotem Ben-Hur,Graduate Student,Technio
Rotem Ben-Hur,Graduate Student,Technio
 

Recently uploaded

Call Girls Miyapur 7001305949 all area service COD available Any Time
Call Girls Miyapur 7001305949 all area service COD available Any TimeCall Girls Miyapur 7001305949 all area service COD available Any Time
Call Girls Miyapur 7001305949 all area service COD available Any Timedelhimodelshub1
 
8447779800, Low rate Call girls in Uttam Nagar Delhi NCR
8447779800, Low rate Call girls in Uttam Nagar Delhi NCR8447779800, Low rate Call girls in Uttam Nagar Delhi NCR
8447779800, Low rate Call girls in Uttam Nagar Delhi NCRashishs7044
 
Call US-88OO1O2216 Call Girls In Mahipalpur Female Escort Service
Call US-88OO1O2216 Call Girls In Mahipalpur Female Escort ServiceCall US-88OO1O2216 Call Girls In Mahipalpur Female Escort Service
Call US-88OO1O2216 Call Girls In Mahipalpur Female Escort Servicecallgirls2057
 
8447779800, Low rate Call girls in Saket Delhi NCR
8447779800, Low rate Call girls in Saket Delhi NCR8447779800, Low rate Call girls in Saket Delhi NCR
8447779800, Low rate Call girls in Saket Delhi NCRashishs7044
 
(Best) ENJOY Call Girls in Faridabad Ex | 8377087607
(Best) ENJOY Call Girls in Faridabad Ex | 8377087607(Best) ENJOY Call Girls in Faridabad Ex | 8377087607
(Best) ENJOY Call Girls in Faridabad Ex | 8377087607dollysharma2066
 
Ten Organizational Design Models to align structure and operations to busines...
Ten Organizational Design Models to align structure and operations to busines...Ten Organizational Design Models to align structure and operations to busines...
Ten Organizational Design Models to align structure and operations to busines...Seta Wicaksana
 
APRIL2024_UKRAINE_xml_0000000000000 .pdf
APRIL2024_UKRAINE_xml_0000000000000 .pdfAPRIL2024_UKRAINE_xml_0000000000000 .pdf
APRIL2024_UKRAINE_xml_0000000000000 .pdfRbc Rbcua
 
FULL ENJOY Call girls in Paharganj Delhi | 8377087607
FULL ENJOY Call girls in Paharganj Delhi | 8377087607FULL ENJOY Call girls in Paharganj Delhi | 8377087607
FULL ENJOY Call girls in Paharganj Delhi | 8377087607dollysharma2066
 
Intro to BCG's Carbon Emissions Benchmark_vF.pdf
Intro to BCG's Carbon Emissions Benchmark_vF.pdfIntro to BCG's Carbon Emissions Benchmark_vF.pdf
Intro to BCG's Carbon Emissions Benchmark_vF.pdfpollardmorgan
 
Keppel Ltd. 1Q 2024 Business Update Presentation Slides
Keppel Ltd. 1Q 2024 Business Update  Presentation SlidesKeppel Ltd. 1Q 2024 Business Update  Presentation Slides
Keppel Ltd. 1Q 2024 Business Update Presentation SlidesKeppelCorporation
 
Digital Transformation in the PLM domain - distrib.pdf
Digital Transformation in the PLM domain - distrib.pdfDigital Transformation in the PLM domain - distrib.pdf
Digital Transformation in the PLM domain - distrib.pdfJos Voskuil
 
2024 Numerator Consumer Study of Cannabis Usage
2024 Numerator Consumer Study of Cannabis Usage2024 Numerator Consumer Study of Cannabis Usage
2024 Numerator Consumer Study of Cannabis UsageNeil Kimberley
 
Call Us 📲8800102216📞 Call Girls In DLF City Gurgaon
Call Us 📲8800102216📞 Call Girls In DLF City GurgaonCall Us 📲8800102216📞 Call Girls In DLF City Gurgaon
Call Us 📲8800102216📞 Call Girls In DLF City Gurgaoncallgirls2057
 
8447779800, Low rate Call girls in Kotla Mubarakpur Delhi NCR
8447779800, Low rate Call girls in Kotla Mubarakpur Delhi NCR8447779800, Low rate Call girls in Kotla Mubarakpur Delhi NCR
8447779800, Low rate Call girls in Kotla Mubarakpur Delhi NCRashishs7044
 
Call Girls In Connaught Place Delhi ❤️88604**77959_Russian 100% Genuine Escor...
Call Girls In Connaught Place Delhi ❤️88604**77959_Russian 100% Genuine Escor...Call Girls In Connaught Place Delhi ❤️88604**77959_Russian 100% Genuine Escor...
Call Girls In Connaught Place Delhi ❤️88604**77959_Russian 100% Genuine Escor...lizamodels9
 
Call Girls In Sikandarpur Gurgaon ❤️8860477959_Russian 100% Genuine Escorts I...
Call Girls In Sikandarpur Gurgaon ❤️8860477959_Russian 100% Genuine Escorts I...Call Girls In Sikandarpur Gurgaon ❤️8860477959_Russian 100% Genuine Escorts I...
Call Girls In Sikandarpur Gurgaon ❤️8860477959_Russian 100% Genuine Escorts I...lizamodels9
 
Future Of Sample Report 2024 | Redacted Version
Future Of Sample Report 2024 | Redacted VersionFuture Of Sample Report 2024 | Redacted Version
Future Of Sample Report 2024 | Redacted VersionMintel Group
 
Call Girls In Radisson Blu Hotel New Delhi Paschim Vihar ❤️8860477959 Escorts...
Call Girls In Radisson Blu Hotel New Delhi Paschim Vihar ❤️8860477959 Escorts...Call Girls In Radisson Blu Hotel New Delhi Paschim Vihar ❤️8860477959 Escorts...
Call Girls In Radisson Blu Hotel New Delhi Paschim Vihar ❤️8860477959 Escorts...lizamodels9
 
Cybersecurity Awareness Training Presentation v2024.03
Cybersecurity Awareness Training Presentation v2024.03Cybersecurity Awareness Training Presentation v2024.03
Cybersecurity Awareness Training Presentation v2024.03DallasHaselhorst
 
Lowrate Call Girls In Sector 18 Noida ❤️8860477959 Escorts 100% Genuine Servi...
Lowrate Call Girls In Sector 18 Noida ❤️8860477959 Escorts 100% Genuine Servi...Lowrate Call Girls In Sector 18 Noida ❤️8860477959 Escorts 100% Genuine Servi...
Lowrate Call Girls In Sector 18 Noida ❤️8860477959 Escorts 100% Genuine Servi...lizamodels9
 

Recently uploaded (20)

Call Girls Miyapur 7001305949 all area service COD available Any Time
Call Girls Miyapur 7001305949 all area service COD available Any TimeCall Girls Miyapur 7001305949 all area service COD available Any Time
Call Girls Miyapur 7001305949 all area service COD available Any Time
 
8447779800, Low rate Call girls in Uttam Nagar Delhi NCR
8447779800, Low rate Call girls in Uttam Nagar Delhi NCR8447779800, Low rate Call girls in Uttam Nagar Delhi NCR
8447779800, Low rate Call girls in Uttam Nagar Delhi NCR
 
Call US-88OO1O2216 Call Girls In Mahipalpur Female Escort Service
Call US-88OO1O2216 Call Girls In Mahipalpur Female Escort ServiceCall US-88OO1O2216 Call Girls In Mahipalpur Female Escort Service
Call US-88OO1O2216 Call Girls In Mahipalpur Female Escort Service
 
8447779800, Low rate Call girls in Saket Delhi NCR
8447779800, Low rate Call girls in Saket Delhi NCR8447779800, Low rate Call girls in Saket Delhi NCR
8447779800, Low rate Call girls in Saket Delhi NCR
 
(Best) ENJOY Call Girls in Faridabad Ex | 8377087607
(Best) ENJOY Call Girls in Faridabad Ex | 8377087607(Best) ENJOY Call Girls in Faridabad Ex | 8377087607
(Best) ENJOY Call Girls in Faridabad Ex | 8377087607
 
Ten Organizational Design Models to align structure and operations to busines...
Ten Organizational Design Models to align structure and operations to busines...Ten Organizational Design Models to align structure and operations to busines...
Ten Organizational Design Models to align structure and operations to busines...
 
APRIL2024_UKRAINE_xml_0000000000000 .pdf
APRIL2024_UKRAINE_xml_0000000000000 .pdfAPRIL2024_UKRAINE_xml_0000000000000 .pdf
APRIL2024_UKRAINE_xml_0000000000000 .pdf
 
FULL ENJOY Call girls in Paharganj Delhi | 8377087607
FULL ENJOY Call girls in Paharganj Delhi | 8377087607FULL ENJOY Call girls in Paharganj Delhi | 8377087607
FULL ENJOY Call girls in Paharganj Delhi | 8377087607
 
Intro to BCG's Carbon Emissions Benchmark_vF.pdf
Intro to BCG's Carbon Emissions Benchmark_vF.pdfIntro to BCG's Carbon Emissions Benchmark_vF.pdf
Intro to BCG's Carbon Emissions Benchmark_vF.pdf
 
Keppel Ltd. 1Q 2024 Business Update Presentation Slides
Keppel Ltd. 1Q 2024 Business Update  Presentation SlidesKeppel Ltd. 1Q 2024 Business Update  Presentation Slides
Keppel Ltd. 1Q 2024 Business Update Presentation Slides
 
Digital Transformation in the PLM domain - distrib.pdf
Digital Transformation in the PLM domain - distrib.pdfDigital Transformation in the PLM domain - distrib.pdf
Digital Transformation in the PLM domain - distrib.pdf
 
2024 Numerator Consumer Study of Cannabis Usage
2024 Numerator Consumer Study of Cannabis Usage2024 Numerator Consumer Study of Cannabis Usage
2024 Numerator Consumer Study of Cannabis Usage
 
Call Us 📲8800102216📞 Call Girls In DLF City Gurgaon
Call Us 📲8800102216📞 Call Girls In DLF City GurgaonCall Us 📲8800102216📞 Call Girls In DLF City Gurgaon
Call Us 📲8800102216📞 Call Girls In DLF City Gurgaon
 
8447779800, Low rate Call girls in Kotla Mubarakpur Delhi NCR
8447779800, Low rate Call girls in Kotla Mubarakpur Delhi NCR8447779800, Low rate Call girls in Kotla Mubarakpur Delhi NCR
8447779800, Low rate Call girls in Kotla Mubarakpur Delhi NCR
 
Call Girls In Connaught Place Delhi ❤️88604**77959_Russian 100% Genuine Escor...
Call Girls In Connaught Place Delhi ❤️88604**77959_Russian 100% Genuine Escor...Call Girls In Connaught Place Delhi ❤️88604**77959_Russian 100% Genuine Escor...
Call Girls In Connaught Place Delhi ❤️88604**77959_Russian 100% Genuine Escor...
 
Call Girls In Sikandarpur Gurgaon ❤️8860477959_Russian 100% Genuine Escorts I...
Call Girls In Sikandarpur Gurgaon ❤️8860477959_Russian 100% Genuine Escorts I...Call Girls In Sikandarpur Gurgaon ❤️8860477959_Russian 100% Genuine Escorts I...
Call Girls In Sikandarpur Gurgaon ❤️8860477959_Russian 100% Genuine Escorts I...
 
Future Of Sample Report 2024 | Redacted Version
Future Of Sample Report 2024 | Redacted VersionFuture Of Sample Report 2024 | Redacted Version
Future Of Sample Report 2024 | Redacted Version
 
Call Girls In Radisson Blu Hotel New Delhi Paschim Vihar ❤️8860477959 Escorts...
Call Girls In Radisson Blu Hotel New Delhi Paschim Vihar ❤️8860477959 Escorts...Call Girls In Radisson Blu Hotel New Delhi Paschim Vihar ❤️8860477959 Escorts...
Call Girls In Radisson Blu Hotel New Delhi Paschim Vihar ❤️8860477959 Escorts...
 
Cybersecurity Awareness Training Presentation v2024.03
Cybersecurity Awareness Training Presentation v2024.03Cybersecurity Awareness Training Presentation v2024.03
Cybersecurity Awareness Training Presentation v2024.03
 
Lowrate Call Girls In Sector 18 Noida ❤️8860477959 Escorts 100% Genuine Servi...
Lowrate Call Girls In Sector 18 Noida ❤️8860477959 Escorts 100% Genuine Servi...Lowrate Call Girls In Sector 18 Noida ❤️8860477959 Escorts 100% Genuine Servi...
Lowrate Call Girls In Sector 18 Noida ❤️8860477959 Escorts 100% Genuine Servi...
 

Xavier van Ruymbeke, App. Engineer, Arteris

  • 1. ○ May 9, 2016 Enhancing Data Center Flash Storage Endurance Using Network-on-Chip Interconnect IP ChipEx2016, Israel Xavier van Ruymbeke Application Engineering Manager, Arteris xavier.van-ruymbeke@arteris.com@arteris.com Copyright © 2016 Arteris
  • 2. ○ May 9, 2016 Arteris Snapshot Copyright © 2016 Arteris Founded 2003; Headquarters in Silicon Valley Global PresenceInterconnect IP Leadership Large&GrowingCustomerBase Customersshipped ~600M unitsin 2014 Connected byArteris Ecosystem Shanghai TokyoCampbell Seoul TaipeiBangalore Paris Israel Strong R&D and Support R&D Engineering 48%App Engineering 29% Sales 14% G&A 9% Marketing 2% Top Semis use Arteris Publicly Disclosed Customers Toshiba In Leading Systems Over 100 NoC-based SoCs in production 1 6 9 13 20 41 52 58 67 76 80 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2 NoCSolution 2006 Main Interconnect 1st Generation FlexNoC 2010 Main Interconnect 2nd Generation FlexWay 2010 Subsystem Interconnect FlexLLI 2011 Interchip Link FlexPSI 2013 Interchip Link FlexNoC Resilience 2014 For Mission Critical Electronics FlexNoC Physical 2015 Physical Awareness w/FlexNoC V3
  • 3. ○ May 9, 2016 Need for on-chip data protection emerging in data center flash controllers Copyright © 2016 Arteris Data Center SSD ○ Huge bandwidth – Logical Block Addressing (LBA) and Cascading Slices to access 10+ TB ○ Low latency – NVMe replaces AHCI ○ Updated standards: JEDEC JESD218B (Endurance), SNIA Solid State Storage (SSS) Performance Test Specification (PTS) ○ Result: More complexity, more gates… • But on-chip errors reduce SSD endurance and reliability More complex systems = Data protection required! 3
  • 4. ○ May 9, 2016 Data center SSD controller complexity requires larger, more complex interconnect Copyright © 2016 Arteris Data RAM Subsystem (Buffers, Parity, Sequence) DMA Engine FlexNoC® Top Level Interconnect Peripheral Subsystem . SPI I2C UART GPIO ROM On-chip interconnect IP DRAM Subsystem DDR3/4 DDR3/4 PHY PHY Memory Scheduler DRAM Controller XOR Engine ECC Flash Control CPU Subsystem R5/7 R5/7 R5/7 R5/7 R5/7 R5/7 Cache LLPP NVMe PCIe Subsystem . . Interconnect Timer Debug PCIePHY Data RAM Data RAM Data RAM Data RAM SRAM RAID Flash 4
  • 5. ○ May 9, 2016 Controller complexity and size increase probability of digital logic failures Copyright © 2016 Arteris Particle hitting the gate and changing its output Wire broken: Un-driven input Destroyed Gate: Stuck at Data center flash controllers must adapt to inevitable errors ○ Digital integrated circuits can fail in many ways: • Transient electrical problems – glitches on power supply, clock supply • Soft errors – Alpha particles, cosmic rays, thermal neutrons • Physical damage – shorts or open wires, destroyed logic gates ○ Designing for data protection requires covering all of them • Not just Soft Errors 5
  • 6. ○ May 9, 2016 Why implement data protection in the SoC interconnect? Implementing protection in the interconnect is better because… 1. Complexity – SW is much more effort to develop and maintain than using certifiable HW IP 2. Quality – HW IP is pre-tested and can be pre-certified while future SW quality is a risk 3. Control – Semi vendors can set a baseline of data protection features that must be implemented, so they are at less risk of poor quality software developed without their knowledge, input or feedback. Copyright © 2016 Arteris Implementing data protection in hardware is less risky than implementing solely in software! 6
  • 7. ○ May 9, 2016 Protecting on-chip interconnect data makes the SoC more reliable Each feature is optional and configurable:  Parity on packet header in transport • Cover faults on packet content in transport layer  Packet Consistency checkers • Capture bad routing or partial packet deletion/duplication in transport  User defined ECC/parity on payload • Cover all faults affecting payload • Can be input/output from socket or generate/terminate  Initiator timeout • Capture single packet deletion or Initiator misbehavior  Target timeout • Capture slave misbehavior  ARM Cortex® R5/R7 support  Register parity bit • Cover Faults on NoC configuration register Copyright © 2016 Arteris NoC= ? = ? = ? = ? = ? = ? = ? = ? NIUNIUNIUNIU NIUNIUNIUNIU Master IP Master IP Master IP Slave IP Slave IP Slave IP Master IP Slave IP NIUNIUNIU NIUNIUNIU NIU NIU SafetyController Fault cc cc cc cc cc cc cc cc ECC/P ECC/P ECC/P ECC/P ECC/P ECC/P ECC/P ECC/P 7
  • 8. ○ May 9, 2016 Data protection advanced features  Unit duplication and checkers • Cover faults in NIU, firewall, power disconnect elements able to transform the access content  Safety Controller • Fault reporting and BIST checking for duplicated unit checkers Copyright © 2016 Arteris NoC= ? = ? = ? = ? = ? = ? = ? = ? NIUNIUNIUNIU NIUNIUNIUNIU Master IP Master IP Master IP Slave IP Slave IP Slave IP Master IP Slave IP NIUNIUNIU NIUNIUNIU NIU NIU SafetyController Fault cc cc cc cc cc cc cc cc ECC/P ECC/P ECC/P ECC/P ECC/P ECC/P ECC/P ECC/P These are effective for unique use cases, but are more costly in power consumption and die area. 8
  • 9. ○ May 9, 2016 Protect SoC data with end-to-end data protection ○ Core Interface Protection supports • User-defined protection schemes ○ To ensure safe connection from core to NoC, core-side protection logic • Detects and corrects errors upon transaction requests • Generates protection semantics to IP core upon transaction responses ○ Packet transport protection logic  Detects and corrects transport-level errors Copyright © 2016 Arteris Protected IP Core Resilient NoC Core-side Error Detection & Correction Protected Core Interface Transport Parity/ECC Transport Parity/ECC Core-side Protection Generation Request Response 9
  • 10. ○ May 9, 2016 Simplest, lowest cost option is parity protection ○ Inserted and terminated by NIUs and units that modify packets ○ Packet Header protection • One parity bit per group of header bits • EVEN parity • Group size determined by “TimingImpact” parameter  Low: 8 bits (more number of resilience bits in header)  Medium: 16 bits  High: 32 bits (less number of resilience bits in header) ○ Packet payload protection • One parity bit per data byte (including byte enable) Copyright © 2016 Arteris Group N N Groups Group N-1Group N-2Group 2Group 1Group 0 0/1 0/1 0/1 0/1 0/1 0/1 Header Width = N Groups even parity bit Calculation one bit per Group … 10
  • 11. ○ May 9, 2016 Protect ARM Cortex-R5/R7 transactions with built-in interconnect ECC support ○ Control • All command redundancy is generated/terminated • 32/64 bits AXI socket support • ODD/EVEN parity support ○ Data • Cortex ECC terminated/generated in the Initiator NIU • Optional byte-level ECC generation inside NoC ○ Option to do network interface unit (NIU) duplication NIUinitiator Cortex NIUinitiator S2G G2T ECC checks NIUTarget G2T S2G ECC checks transport Cortex ECC terminated, Byte ECC generation Byte ECC terminated, custom ECC generation Copyright © 2016 Arteris 11
  • 12. ○ May 9, 2016 Implement user-defined payload ECC to protect other data ○ Per-socket, user-defined number of DataInfo bits per bytes ○ NoC wide, transport level user-defined number of DataInfo bits per bytes ○ Conversion to/from socket DataInfo semantics from/to Transport DataInfo semantics • Done inside the NIUs at generic level • If socket and transport DataInfo are identical (same number of bits per bytes, same semantic): can be just copied transparently • If socket and transport DataInfo are different (different number of bits per bytes, and/or different semantic): conversion logic need to be provided by the user and will be located in NIU placeholders ○ Socket Protocols supporting ECC: • AXI – with narrow bursts split to single word transactions • AHB – with AHB target using Byte Enables • OCP – with targets providing correct ECC on SData/SDataInfo for write responses NIUinitiator NIUinitiator S2G G2T Socket <> Transport NIUTarget G2T S2G Transport <> Socket transport Socket defined DataInfo bits per byte Socket defined DataInfo bits per bytes NoC transport number of bits per payload byte Conversion placeholder Conversion placeholder Copyright © 2016 Arteris 12
  • 13. ○ May 9, 2016 Duplicate parts of the interconnect to achieve higher reliability ○ Generic and specific NIUs are duplicated • Includes reorder buffers, scheduler – no restrictions ○ Units modifying packets are duplicated • Target power disconnect, Firewall ○ Duplicate logic delayed by one or two cycle • Using separate clock input and reset input • Avoid common faults due to power glitches and clock branch defects ○ A Checker is used to compare outputs • Checker has separate clock gater ○ Differences in output are reported as faults ○ Some unit configurations cannot be duplicated • NIUs with multiple generic interfaces (LLI, OCP MT) • Target power disconnect with asynchronous crossing • These units can be present in the system but won’t be duplicated Reference Unit Functional Unit Delay Inputs Checker Outputs ClkClk’ FaultBIST Control Copyright © 2016 Arteris 13
  • 14. ○ May 9, 2016 Need checkers, fault reporting and BIST to ensure data protection Copyright © 2016 Arteris • Sequence testing patterns through delay registers • Report issues in the comparison logic (stuck-at faults typically) ○ One Checker per duplicated unit ○ In Mission mode: Compares Unit outputs • Early Unit outputs delayed by one cycle • Early/Late Unit regular outputs bit-compared • Logic fault outputs from both units compared to 0 • Redundant report of packet and/or NIU check faults ○ In BIST mode: Checks comparison logic) Safety Controller Checker Checker Clk Clk’ Fault BIST control BIST control Mission Interrupt Latent Interrupt Clk …. 14
  • 15. ○ May 9, 2016 Protect only the parts of controller that affect system reliability the most Accelerators Peripheral Data-Protected NoC I/OMem Ctrl Peripheral DRAM ROM/Flash CPUs Peripheral I/O ECC - Core SafetyController =? =? =? =? =? Fault cccccc cc cc cc Consistency checker Equality checker Fault NoC Without Protection Firewall (SW programmable) =? Copyright © 2016 Arteris 15
  • 16. ○ May 9, 2016 Data center and SSD flash controllers require on-chip data protection ○ Interconnect size as a percentage of die size, and complexity, increase as the number of functional IP blocks increases ○ Data traveling on the interconnect is at risk of corruption ○ Technologies exist today to protect data traveling over the on-chip interconnect, which increases system reliability and endurance Copyright © 2016 Arteris Meeting data center storage reliability goals requires on-chip interconnect data protection 16
  • 17. ○ May 9, 2016 Copyright © 2016 Arteris 17

Editor's Notes

  1. JEDEC new standard: SOLID STATE DRIVE (SSD) REQUIREMENTS AND ENDURANCE TEST METHOD JESD218B Mar 2016
  2. If core is unprotected, can still use transport protection logic to ensure protection within the NoC Word based ECC is converted to bite based ECC in FlexNoC in the NIUs