Page 1
Company Public – NXP, the NXP logo, and NXP secure connections for a smarter world are trademarks of NXP
B.V. All other product or service names are the property of their respective owners. © 2019 NXP B.V.
David Frerichs, Solutions Engineer Alexa Voice Services
Leo Azevedo, Senior Director NXP i.MX Product Marketing
Driving Video, Audio and Voice Solutions with the i.MX 8M Family of Multimedia Applications Processors
June 2019 | Session #AMF-SMH-T3515
Page 2
COMPANY PUBLIC 1COMPANY PUBLIC 1
• Market Trends in IoT
• The i.MX 8 Series Roadmap
• The i.MX 8M Family Line-up
• Making It Easy for You
• Why You Should Choose NXP
Agenda
Page 3
COMPANY PUBLIC 2
Market Trends in IoT
Page 4
COMPANY PUBLIC 3
IoT Shifts to Consumer*
• As costs drop, consumer will dominate IoT
• Dependent on smart-home platforms:
− Smart speakers provide entry point for IoT
− Alexa going into alarms, appliances, door locks, lighting, thermostats and more
• In 2023, consumer totals 74% of market in connected appliances, smart home, wearables
Source: The Linley Group 2018
Smart
Home
Smart
Appliance
Wearables
Smart
Meters
Smart
Buildings
Other
Industrial
* Consumer IoT excludes traditional connected devices such as media players, smart TVs, smartphones, and PCs
IoT Market
Page 5
COMPANY PUBLIC 4
Decora Voice™ Dimmer with Amazon Alexa Built-inPowered by i.MX 8M Mini
Page 6
COMPANY PUBLIC 5
Six Most Popular Smart Home Devices*
• Networked cameras
• Video doorbells
− 25% of U.S. broadband households plan to buy a video doorbell in 2019
• Smart door locks
• Smart thermostat
• Smart lightbulbs
• Smart garage door opener
Owners of smart speakers with voice assistants (Amazon Echo, Google Home) are 3xmore likely to own a smart home device than consumers who do not
Multiple IoT devices in home is becoming common— among consumers who own a connected entertainment or smart home device, more than 60% have more than one
Source: Parks Associates 2019 Connected Home & Entertainment Trends * Found in 22% of U.S. broadband households
Page 7
COMPANY PUBLIC 6
The i.MX 8 Series Roadmap
Page 8
COMPANY PUBLIC 7
i.MX 8M Family of Applications Processors
M4
i.MX 7 Family
i.MX 8 Family
Safety Certifiable & Efficient Performance
Flexible Efficient Connectivity
ARM® v7-A
i.MX 8M family
i.MX 8X Family
Advanced Computing, Audio/Video & Voice
Advanced Graphics, Vision & Performance
i.MX 7ULP Family
Ultra Low Power with Graphics
ARM ® v8-A (32-bit/ 64-bit)
ARM® v7-A (32-bit)
i.MX 6QuadPlus
i.MX 6Dual
i.MX 6Solo
i.MX 6DualLite
i.MX 6SoloLite
i.MX 6SoloX
i.MX 6UltraLite
i.MX 6DualPlus
i.MX 6Quad
i.MX 6SLL
i.MX 6ULL
M4
A53
A72P
in-t
o-p
in C
om
pa
tib
le
So
ftw
are
Co
mp
atib
le
M4
A9
A9
A7
i.MX 8M
• 8M Quad
• 8M QuadLite
• 8M Dual
M4/M7
A53
A35
M4A7
So
ftw
are
Co
mp
atib
le
i.MX 6ULZ
ANNOUNCED
NOW
NOW
i.MX 8M Plus
In design
IN
DESIGN
i.MX 8M Mini
• 8M Mini Quad
• 8M Mini QuadLite
• 8M Mini Dual
• 8M Mini DualLite
• 8M Mini Solo
• 8M Mini SoloLite
i.MX 8M Nano
• 8M Nano Quad
• 8M Nano QuadLite
• 8M Nano Dual
• 8M Nano DualLite
• 8M Nano Solo
• 8M Nano SoloLite
i.MX 8M PicoIn design
CONCEPT
Pin
-to
-pin
Co
mp
atib
le
Page 9
COMPANY PUBLIC 8
i.MX 8 Series: Target ApplicationsAdvanced graphics, video, image processing, vision, audio and voice
i.MX 8 Familyi.MX 8X Familyi.MX 8M Family M4
A53
GPU
Advanced Graphics,
Vision & Performance
Advanced Computing,
Audio/Video & VoiceSafety Certifiable &
Efficient Performance
M4A35
DSPGPU M4
A53A72
DSP
GPU
Page 10
COMPANY PUBLIC 9
i.MX 8M Family: Expanding Performance, Efficiency,
Connectivity
9mW
Sleep Mode
2x CPU
Performance
14LPC FinFET: more performance, better efficiency
Applications <2-3W
75% less power in
deep sleep mode
60% less power
in run mode
Graphics
OpenGL ES 3.1, Vulkan
OpenCL 1.2
Quad Cortex-A53 up to 1.8MHz
plus Cortex-M4/M7 for task offload
and power optimizations
Video
4Kp60 decoder (HDR, Dolby Vision)
1080p60 encoder
Performance benchmark in Dhrystone with 6QP.
Support for new memory
types: LPDDR4 (12GB/s),
DDR4 plus DDR3L
New interfaces: USB 3.0, eDP, eARC
Immersive audio experience:
20+ channel high-fidelity
audio, multi-room music,
Dolby ATMOS certification
Quad-A53 plus Cortex-M4 and
Cortex-M7 for task offload and
power optimizations
Up to 2 TOPS ML acceleration
Page 11
COMPANY PUBLIC 10
The i.MX 8M Family Line-up
Page 12
COMPANY PUBLIC 11
i.MX 8M Family: Scalable Solutions for the Broad Market…Scalable series of three Arm V8 64-bit (/32-bit) based SoC Families in Production
(i.MX 8M Quad / i.MX 8M Mini) and Announced (i.MX 8M Nano)
Software Compatible (including GPU Tools)
i.MX 8M Nanoi.MX 8M Minii.MX 8M Quad
M4
A53
GPU
• Advanced Audio, Voice
• 4K Video, HDR
• 3D GPU with OpenCL
• Multi-Display and Camera
• USB 3.0, Multiple PCIe
• Advanced Audio, Voice
• 1080p Video Encode and Decode
• 3D GPU
• Single Display & Camera
• USB 2.0, PCIe
• Performance and Power-efficient
Pin Compatible
i.MX 8M Quad
i.MX 8M Dual
4Q’19
i.MX 8M
QuadLite
Pin Compatible 14x14 Package: Design Compatibility
i.MX 8M Mini
Quad/Dual/Solo
i.MX 8M Mini
QuadLite/DualLite/
SoloLite
M4
A53
GPU
High-end ($$$) => Mid-end ($$) => Low-end ($)
i.MX 8M Nano
Quad/Dual/Solo
M7
A53
GPU
• Low Cost Audio, Voice
• Video on Cortex-A53
• 3D GPU with OpenCL
• Single Display & Camera
• USB 2.0
• Performance and Power Efficient
NOW NOW
NANONANO
i.MX 8M Nano
QuadLite/DualLite/S
oloLite
Page 13
COMPANY PUBLIC 12
W.I.N.N.E.R.S.! – The Championship Line Up
Winner 2019 Product of the year, Active Components NXP i.MX 8M
✓ Enriched User Experience
✓ Performance and Versatility
✓ High Speed Interfaces
More info: NXP Blog
Page 14
COMPANY PUBLIC 13
Embedded World 2019: NXP Introduces i.MX 8M Nano Links to Coverage
NikkeiXTech: Low-power app processor with strong security,
NXP strong on audio
March 13, 2019
EETimes Japan: Conscious of edge inference, embedded
processor "i.MX 8M Nano”
March 8, 2019
EETimes, EETimes Asia: Edge Intelligence Grabs the
Spotlight at Embedded World
March 6, 2019
Lembarque: NXP completes i.MX 8M family of multimedia
processors with eco-efficient i.MX 8M Nano models
March 6, 2019
Forbes: NXP Proves That Smaller and More Secure Is
Better
March 5, 2019
ElectroniqueS: NXP Lowers the Consumption of its i.MX 8M
Processors
March 1, 2019
LinuxGizmos: NXP launches quad -A53 i.MX8M Nano SoC
with 600MHz Cortex-M7 MCU
February 28, 2019
ElectronicScienceKorea: NXP Launches i.MX 8M Family to
Accelerate Edge Computing Innovation
February 28, 2019
KoreaIndustryPost: You do not need to change processors
every time you develop new products
February 28, 2019
CNXSoftware.com: NXP i.MX 8M Nano is a Power-
optimized Arm Cortex-A53/M7 Processor
February 27, 2019
Page 15
COMPANY PUBLIC 14
i.MX 8M Family Target ApplicationsConsumer &
Pro Audio Systems
• Portable audio devices
• Wireless or networked speakers
• Surround sound and sound bars
• Audio/video receiver
• Public address systems
• Smart Remote Controls
• Voice>Music>Video
Convergence
• Video doorbell
• Security and surveillance system
• Smart light control
• HVAC climate control
• IoT gateway
• Two-way video conferencing
• Smart appliances
• Control panel
• Digital sign
• Barcode or image scanner
• Industrial printer
• Ruggedized HMI
• Kiosk
• Machine visual inspection
• Image analytics
• Factory automation
• Test and measurement
• Two-way radio
• Mobility and logistics
• High-speed i/f conversion
• Smart appliance
• Service robot, e.g. vacuum,
mower, cleaner
• Camera and LCD
• Mobile patient care, e.g.
infusion pump, respirator
• Blood pressure monitor
• Activity/wellness monitor
• Exercise equipment
• IP phones
Consumer,
Healthcare, Other
Industrial HMI,
Vision and Automation
Home &
Building Automation
v
Page 16
COMPANY PUBLIC 15
The i.MX 8M Quad Family is a Game Changer!• Industry-Leading Video and Audio
− Video quality with full 4K UltraHD resolution and HDR
▪ Dolby Vision, HDR10, and HLG
− Highest levels of pro audio fidelity, up to 20 channels
• Performance and Versatility− 1x-4x Cortex-A53 processors at speeds of up to 1.5
GHz per core, 1x Cortex-M for task offload
− Flexible memory options (speed, power, cost)
− Newest high-speed interfaces for flexible connectivity
− Configurations less than 2 to 3 Watts
• Basic to Advanced HMI Solutions− Industrial and consumer HMI
− Dual displays and camera inputs
− Immersive Audio and Video processing
− Voice Solutions (local and cloud connected)
− Interconnected Devices (smarter edge devices)
Page 17
COMPANY PUBLIC 16
i.MX 8M Quad Key Features
• Scalable CPU Performance
− Dual, Quad Cortex-A53 up to 1.5 GHz
− Cortex-M4 up to 266 MHz
▪ Offload Tasks, Optimize Power, Increase Security
− HW based resource isolation & safe sharing
between Cortex-A and Cortex-M
− Enables concurrent execution of multiple software
environments to provide high performance with
real time responsiveness
• High Performance at Low power
- Scalable power using mobile power architecture
• Independent power gate on each core
• Multiple Power Islands / Flexible clock gating
• DVFS support
- Support Cortex-M4 running with A53 and high-
speed modules power gated
- DDR3L, DDR4 and LPDDR4 high performance /
lower power memories
• Best In Class Multimedia Performance
− Up to 4K video decoding in h.265, h.264, VP9
− High Dynamic Range (HDR) support
− High Performance 3D Graphics Acceleration
▪ 32 GFLOPS
▪ OpenGL ES 3.1, Vulkan, OpenCL 1.2, and Open VG
support
- 20+ audio channel inputs/outputs
- 32-bit @ 384KHz interfaces (up to 768 KHz)
• Multiple High-Speed Interfaces
- Enabling Vision: 2x MIPI-CSI (4-lanes)
▪ Support 1080p30, 720p60, VGA at 60 fps
- Display
• HDMI 2.0a (up 4K), MIPI-DSI (4 lanes) (720p60)
- Connectivity
• 1x Gb Ethernet (MAC): AVB & IEEE 1588 for sync, and EEE for low power
• 2x USB 3.0 Type C/dual role with PHY (works as USB 3.0 or USB 2.0), 5Gbps
• 2x PCIe 2.0 (1-lane) with L1 substates (low power, fast wakeup)
Page 18
COMPANY PUBLIC 17
i.MX 8M Family – In Production NowApplications Processors for Consumer and Industrial Applications
Video
ARM CPU
Cortex-A53 | Cortex-M4U
p t
o 1
3,8
00 D
MIP
S
So
ftwa
re C
om
pa
tibility
Dis
cre
te P
acka
ge
Op
tion
Pin
Com
patib
ility
• Single core GPU
• 4 Vec4 Shader
• Up to 64 GFLOPS
• OpenGL ES 3.0
• OpenCL 1.2
• 267 MTri / sec
• 1.6 GPix / sec
Connectivity I/ODisplay / Camera
• 20-channels
• 32-bits @
384KHz
• DSD512, TDM
• SPDIF Tx& Rx
• HDMI ARC
Audio I/O
• 2x USB3.0
(Type C)
• 2x PCIe
• 2x SDIO
• 1x GbE
3D GPU
• Decode:
4Kp60
H.265,
VP9
H.264,
VP8
• MIPI-DSI 4-lane
• 2x MIPI-CSI 4-lane
Streaming Media Voice Assistants Industrial IoT Edge Compute Machine VisionAI, Machine Learning
i.MX 8M Quad4x A53
ARM CPU
Cortex-A53 | Cortex-M4
i.MX 8M Dual 2x A53
Up
to
6,9
00 D
MIP
S• HDMI
ARM CPU
Cortex-A53 | Cortex-M4
4x A53
Up
to
13,8
00 D
MIP
S
i.MX 8M QuadLite
Video Connectivity I/ODisplay / Camera
Audio I/O3D GPU
Video Connectivity I/ODisplay / Camera
Audio I/O3D GPU
SAME AS
8M Quad
SAME AS
8M Quad
SAME AS
8M Quad
SAME AS
8M Quad
SAME AS
8M Quad
SAME AS
8M Quad
SAME AS
8M Quad
SAME AS
8M Quad
SAME AS
8M QuadNo Video
acceleration
Page 19
COMPANY PUBLIC 18
i.MX 8M Quad
Planned to be added in the 10-year longevity program
Quad ARM Cortex-A53 @ 1.5+ GHz (13,800 DMIPS)
• ARM v8 Fully 64-bit capable
ARM Cortex-M4 @ 266 MHz for Low Power, Security
Package: FCBGA 17x17mm, 0.65mm pitch
Operating System targets: Linux OS, Android OS, FreeRTOS
Qualification for Consumer and Industrial applications
Feature Highlights:
• Security: DRM support for RSA, AES, 3DES, DES
• GC7000Lite 3D Graphics GPU (OpenGL ES 1.0, 2.0, 3.0, 3.1, Vulkan, and OpenCL 1.2)
• 4Kp60 Main, Main10 H.265/HEVC and VP9, 4Kp30 H.264 decoder
• High Dynamic Range Support: HDR10, Dolby Vision, HLG
• 1080p60 MPEG-2, MPEG-4p2, VC-1, VP8, RV9, AVS/AVS+, MJPEG, H.263 decoder
• High quality image resizing and graphics overlay
• Audio: S/PDIF Rx & Tx, 6x I2S/SAI (up to 20ch 32bit @ 384Khz support)
• Display Interfaces: 1x MIPI DSI (4-lane); 1x HDMI 2.0a Output with HDCP 2.2
• Camera Interfaces: 2x MIPI CSI2 input (4-lane each) with PHY
Connectivity
• 2x USB 3.0 Type C/dual role with PHY (works as USB 3.0 or USB 2.0)
• 1x Gb Ethernet (MAC): AVB & IEEE 1588 for sync, and EEE for low power
• 2x PCIe 2.0 (1-lane) with L1 substates (low power, fast wakeup)
• 4x UART, 4x I2C, 3x SPI
Memory
• x16, x32 LPDDR4/DDR4/DDR3(L) (up to 3200 Mtps)
• 2x SDIO3.0 / eMMC5.0 / SD 4
• Raw NAND controller (BCH62)
• Quad-SPI for fast boot from SPI NOR; with Execute in Place (XIP)
Page 20
COMPANY PUBLIC 19
Power Efficient and Scalable
• 14LPC FinFET: higher performance at lower power.
• Cortex-A53: 1x, 2x or 4x cores, up to 1.8 GHz, optional hardware video codecs.
• Cortex-M: heterogeneous processing for task offload and power optimizations.
• Pin-compatibility provides drop-in scalable product performance using the same hardware design and minimal software changes.
Cost-optimized System Design
• Full-featured packages with cost-effective PCB design references (6-8 layer board design and no microvias).
• LPDDR4 for highest performance and lowest power, or DDR3L and DDR4 for lowest system cost.
• Direct connection of PDM microphones (no CODEC).
• Linux/Android/FreeRTOS BSP and solutions software.
Machine Learning, Multi-Media
• Cost-effective machine-learning & inference using Cortex-A53 or GPU.
• Up to 1080p60 video decoding (H.265, H.264, VP8/9) and 1080p60 video encoding (H.264, VP8).
• Integrated GPUs for 1080p media UI.
• 20+ audio channels and audio HW enhancements on some devices (e.g. ASRC).
Industrial Reliability, Supply
• -40C to 105C (junction) and Industrial-tier qualification (10yrs, continuous ON).
• Minimum 10 year supply longevity on commercial and industrial devices.
• Gigabit Ethernet with IEEE1588 and AVB support.
• Fanless design for reduced cost in high temperature environments.
Extreme, Cost-Effective Scalability for Media and Industrial IoT
NANO
Page 21
COMPANY PUBLIC 20
i.MX 8M Mini & 8M Nano Applications ProcessorsPower-optimized, scalable, multi-core applications processors built in advanced
14nm LPC FinFET process
Video
Arm ® CPU
Cortex-A53 | Cortex-M4
So
ftwa
re C
om
pa
tibility
14
x1
4 0
.5m
m
Pin
Co
mp
atib
ility
Connectivity I/ODisplay / Camera
Audio I/O2D + 3D GPU
• 1080p60
Decode: H.265,
VP9 H.264, VP8
• 1080p60
Encode: H.264,
VP8
• MIPI-DSI 4-
lane
• MIPI-CSI 4-
lane
Streaming Media Voice Assistants Industrial and IoT Edge ComputeAI, Machine Learning
Lite - No HW
video acceleration
• 12Tx + 16Rx
I2S
up to 768kHz
• DSD512, TDM
• SPDIF Tx &
Rx
• 8-ch PDM MIC
• 2x USB 2.0
• 1x PCIe
• 3x SDIO
• 1x GbE
• x32
LPDDR4,
x16 DDR4,
x16 DDR3L
NANO
i.MX 8M Nano Quad
i.MX 8M Nano
QuadLite
i.MX 8M Nano Dual
i.MX 8M Nano
DualLite
i.MX 8M NanoSolo
i.MX 8M Nano
SoloLite
Video
Arm® CPU
Cortex-A53 | Cortex-M7
Connectivity I/ODisplay / Camera
Audio I/O3D GPU
• MIPI-DSI 4-
lane
• MIPI-CSI 4-
lane
• 10Tx + 10Rx
I2S
up to 768kHz
• DSD512, TDM
• SPDIF Tx &
Rx
• 8-ch PDM MIC
• ASRC
• 1x USB 2.0
• 3x SDIO
• 1x GbE
• x16
LPDDR4,
DDR4,
DDR3L
• 2 Vec4 Shader
• OpenGL ES
2.0/3.0/3.1
• OpenVG 1.1
• OpenCL 1.2
• Vulkan
• 1 Vec4 Shader
• OpenGL ES 2.0
• OpenVG 1.1
i.MX 8M Mini Quad
i.MX 8M Mini Dual
i.MX 8M Mini Solo
i.MX 8M Mini
QuadLite
i.MX 8M Mini DualLite
i.MX 8M Mini SoloLite
• A53 up to
1.8GHz
• Quad – 4x
A53
• Dual – 2x A53
• Solo – 1x A53
• M4 up to
400MHz
• No HW video
acceleration• A53 up to
1.5GHz
• Quad – 4x
A53
• Dual – 2x A53
• Solo – 1x A53
• M7 up to
600MHzLite - No HW
GPU graphics
acceleration
Page 22
COMPANY PUBLIC 21
i.MX 8M Mini
External MemoryMultimedia
i.MX 8M Mini
System Control
Security
Secure JTAG
XTAL
3D Graphics: GC NanoUltra
1 shader, OpenGL ES 2.0
Smart DMA x3
Random Number
TrustZone
Secure Clock
eFuse Key Storage
DRM Ciphers
Temperature Sensor
Connectivity & I/O
Dual-ch QuadSPI (XIP)
1x PCIe 2.0 – 1-lane
1Gb Ethernet
(IEEE 1588, EEE & AVB)
1080p60 VP8, VP9, H.264, H.265 decoder
Main CPU Platform
NEON
32KB D-cache
FPU
32KB I-cache
512KB L2 Cache
Quad/Dual Corte-A53
NEON
32KB D-cache
FPU
32KB I-cache
Quad/Dual ortex-A53
NEON
32KB D-cache
FPU
32KB I-cache
Quad Cortex-A53
NEON
32KB D-cache
FPU
32KB I-cache
2x USB2.0 OTG + PHY
1080p60 VP8, H.264 encoder
Low Power, Security CPU
Cortex-M4
16KB I-cache
256KB TCM (SRAM)
16KB D-cache
x16-x32 LPDDR4-3000
DDR4-2400, DDR3L-1600
3x SDIO3.0/eMMC5.1/SD3.0
S/PDIF Rx & Tx,
5x SAI (12Tx + 16Rx I2S lanes)
Up to 49.152MHz BCLK
TDM, DSD512
32KB Secure RAM
MIPI-CSI 4-lane with PHY
MIPI-DSI 4-lane with PHY
8ch PDM Input
NAND Controller (BCH62)
2D Graphics: GC328Timer x6
PWM x4
Watchdog x3
PLLs
Target Timeline: Production (14x14) 1Q’19
Package: FCBGA 14x14mm, 0.5mm pitch
Operating Systems: Linux OS, Android OS, FreeRTOS
Qualification: for consumer (0 to +95C) and industrial (-40C to +105C, 10yr 24/7) applications
Longevity: 10yr longevity program
On-Chip RAM (OCRAM)
256KB OCRAM
4x UART 5Mbps
4x I2C
3x SPI
Feature Highlights:• Arm Cortex-A53 MPCore platform:
− Quad/Dual/Solo 1.6-1.8GHz
− 32KB/32KB I/D cache, 512KB L2 cache with ECC
− Media Processing Engine (MPE) with NEON
− Floating Point Unit (FPU) supports VFPv4-D16 architecture
− 64-bit Armv8-A architecture
• Arm Cortex-M4 platform:
− Up to 400MHz
− 16KB/16KB I/D cache, 256KB TCM
• On-Chip RAM: 256KB OCRAM
• External Memory Interface:
− 32/16-bit DRAM interfaces (LPDDR4-3000, DDR4-2400, DDR3L-1600)
− 8-bit NAND Flash supports raw MLC/SLC (BCH62, ONFi3.2)
− eMMC 5.1 Flash, SPI NOR Flash, FlexSPI with support for XIP
• Multimedia:
− 1080p60 VP9 Profile 0,2 (10-bit) decoder, HEVC/H.265 decoder, AVC/H.264 Baseline,
Main High decoder, VP8 decoder
− 1080p60 AVC/H.264 encoder, VP8 encoder
− GC NanoUltra 3D GPU, 1x shader, OpenGL ES 2.0
− GC328 2D GPU
− LCDIF display controller supports up to 2 layers of overlay, 1080p60
− Display: 1x MIPI DSI (4-lane) with PHY
− Camera: 1x MIPI CSI (4-lane each) with PHY
− Audio: Supports over 20 channels of audio. 5x SAI (12Tx + 16Rx external I2S lanes);
4Tx + 4Rx support 49.152MHz BCLK for 768KHz audio, all others 24.576MHz BCLK
(32-bit, 2-ch 384KHz, up to 32-ch TDM); DSD512
• Security:
− DRM support for RSA, AES, 3DES, DES
• Connectivity:
− 1x Gb Ethernet (MAC): AVB & IEEE 1588 for sync, and EEE for low power
− 1x PCIe 2.0 (1-lane) with L1 substates (low power, fast wakeup)
− 2x USB 2.0 OTG with PHY
− 4x UART, 4x I2C, 3x SPI
− 3x SDIO3.0 / eMMC5.1 / SD3.0
Page 23
COMPANY PUBLIC 22
Feature Highlights:• Arm Cortex-A53 MPCore platform:
− Quad/Dual/Solo 1.2-1.5GHz
− 32KB/32KB I/D cache, 512KB L2 cache with ECC
− Media Processing Engine (MPE) with NEON
− Floating Point Unit (FPU) supports VFPv4-D16 architecture
− 64-bit Armv8-A architecture
• Arm Cortex-M7 platform:
− 600+ MHz
− 32KB/32KB I/D cache, 256KB TCM
• On-Chip RAM: 512KB OCRAM
• External Memory Interface:
− 16-bit DRAM interfaces (LPDDR4-3200, DDR4-2400, DDR3L-1600)
− 8-bit NAND Flash supports raw MLC/SLC (BCH62, ONFi3.2)
− eMMC 5.0/5.1 Flash, SPI NOR Flash, FlexSPI with support for XIP
• Multimedia:
− GC 7000UL 3D Graphics GPU, 2 shader core, OpenGL ES 3.1, Vulkan, OpenCL 1.2
− LCDIF display controller supports up to 2 layers of overlay, 1080p60
− Display: 1x MIPI DSI (4-lane) with PHY
− Camera: 1x MIPI CSI (4-lane each) with PHY
− Audio: Supports over 20 channels of audio. 5x SAI (10Tx + 10Rx external I2S lanes);
up to 49.152MHz BCLK for 768KHz audio (32-bit, 2-ch 768KHz, up to 32-ch TDM);
DSD512
− ASRC: 32 channels of very high quality, asynchronous audio sample rate conversion
• Security:
− DRM support for RSA, AES, 3DES, DES
• Connectivity:
− 1x USB 2.0 OTG with PHY
− 1x Gb Ethernet (MAC): AVB & IEEE 1588 for sync, and EEE for low power
− 4x UART, 4x I2C, 3x SPI
− 3x SDIO3.0 / eMMC5.1 / SD3.0
i.MX 8M Nano
Target Timeline: Production target 4Q’19 (14x14)
Package: FCBGA 14x14mm, 0.5mm pitch de-pop array
Operating Systems: Linux OS, Android OS, FreeRTOS
Qualification: for consumer (0 to +85C) and industrial (-40C to +105C, 10yr 24/7) applications
Longevity: 10yr longevity program
External MemoryMultimedia
i.MX 8M Nano
System Control
Security
Secure JTAG
XTAL
3D Graphics: GC 7000UL
2 shaders
OpenGL ES 2.0/3.0/3.1
Vulkan
OpenCL 1.2
Smart DMA x3
Random Number
TrustZone
Secure Clock
eFuse Key Storage
DRM Ciphers
Temperature Sensor
Connectivity & I/O
4x UART 5Mbps
Dual-ch QuadSPI (XIP)
1Gb Ethernet
(IEEE 1588, EEE & AVB)
Main CPU Platform
NEON
32KB D-cache
FPU
32KB I-cache
512KB L2 Cache
Quad/Dual Corte-A53
NEON
32KB D-cache
FPU
32KB I-cache
Quad/Dual ortex-A53
NEON
32KB D-cache
FPU
32KB I-cache
Quad Cortex-A53
NEON
32KB D-cache
FPU
32KB I-cache
1x USB2.0 OTG + PHY
Low Power, Security CPU
Cortex-M7
32KB I-cache
256KB TCM (SRAM)
32KB D-cache
x16 LPDDR4-3200
DDR4-2400, DDR3L-1600
3x SDIO3.0/eMMC5.0/5.1/SD3.0
S/PDIF Rx & Tx,
4x SAI (6Tx + 6Rx I2S lanes)
Up to 49.152MHz BCLK
TDM, DSD512
32KB Secure RAM
MIPI-CSI 4-lane with PHY
MIPI-DSI 4-lane with PHY
8ch PDM Input
NAND Controller (BCH62)ASRC
Timer x6
PWM x4
Watchdog x3
PLLs
Differences to i.MX 8M Mini
On-Chip RAM (OCRAM)
512KB OCRAM
4x I2C
3x SPI
Page 24
COMPANY PUBLIC 23
i.MX 8M Mini & 8M Nano
14x14 Package DesignOptimized for size and system cost
• 14 x 14 mm BGA, 0.5mm pin pitch, 486 pins
• Depopulated array for lower manufacturing cost
− Use low cost drilled vias; no high-cost laser microvias
− 3.2 mil trace and space escape
• Layout for 6 to 8 layer PCB
− Optimized for LPDDR4 routing
▪ LPDDR4 expects 10+ year product longevity to support wide adoption by automotive market
• Optimized power placement
− Accommodates decouple caps under the BGA
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
A VSSDRAM_D
QS0_P
DRAM_D
Q06
DRAM_D
M0
DRAM_D
Q00
ECSPI2_
SS0
ECSPI1_
MISO
ECSPI2_
MISO
MIPI_DSI
_D0_N
MIPI_DSI
_D1_N
MIPI_DSI
_CLK_N
MIPI_DSI
_D2_N
MIPI_DSI
_D3_N
MIPI_CSI
_D0_N
MIPI_CSI
_D1_N
MIPI_CSI
_CLK_N
MIPI_CSI
_D2_N
MIPI_CSI
_D3_N
PCIE_RX
N_N
PCIE_TX
N_N
PCIE_CL
K_N
USB1_D
N
USB2_D
N
PMIC_O
N_REQONOFF
RTC_XT
ALIVSS A
BDRAM_D
Q05
DRAM_D
QS0_NVSS
DRAM_D
Q07
DRAM_D
Q01
ECSPI1_
SS0
ECSPI1_
MOSI
ECSPI2_
MOSI
MIPI_DSI
_D0_P
MIPI_DSI
_D1_P
MIPI_DSI
_CLK_P
MIPI_DSI
_D2_P
MIPI_DSI
_D3_P
MIPI_CSI
_D0_P
MIPI_CSI
_D1_P
MIPI_CSI
_CLK_P
MIPI_CSI
_D2_P
MIPI_CSI
_D3_P
PCIE_RX
N_P
PCIE_TX
N_P
PCIE_CL
K_P
USB1_D
P
USB2_D
PPOR_B
RTC_XT
ALOVSS
24M_XTA
LIB
CDRAM_D
Q04VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS
24M_XTA
LO
JTAG_TR
ST_BC
DDRAM_D
Q03
DRAM_D
Q02
DRAM_A
C11
ECSPI1_
SCLK
I2C2_SD
A
I2C2_SC
L
I2C4_SC
L
MIPI_VR
EG_CAP
UART3_T
XD
PCIE_RE
SREFUSB1_ID USB2_ID
TEST_M
ODE
JTAG_M
ODD
EDRAM_D
Q14
DRAM_D
Q15VSS
DRAM_A
C10
ECSPI2_
SCLK
I2C1_SC
L
I2C3_SC
L
I2C4_SD
A
UART1_
RXD
UART2_T
XD
UART3_
RXD
USB1_TX
RTUNE
USB2_TX
RTUNE
PMIC_ST
BY_REQVSS
JTAG_TD
O
JTAG_TD
IE
FDRAM_D
M1
DRAM_D
Q08VSS
DRAM_A
C00
DRAM_A
C01
DRAM_A
C06
I2C1_SD
A
I2C3_SD
A
UART1_T
XD
UART2_
RXD
UART4_T
XD
UART4_
RXD
USB1_VB
US
USB2_VB
US
RTC_RE
SET_BVSS
JTAG_TC
K
JTAG_T
MSF
GDRAM_D
QS1_P
DRAM_D
Q09VSS VSS VSS VSS
VDD_PCI_
1P8VSS VSS VSS VSS
BOOT_M
ODE0
BOOT_M
ODE1G
HDRAM_D
QS1_NVSS
NVCC_ECS
PI
VDD_MIPI
_1P8
VDD_USB_
1P8VSS
CLKOUT
1CLKIN1 H
JDRAM_D
Q10
DRAM_D
Q11VSS
DRAM_A
C03
DRAM_A
C07
DRAM_A
C08VSS
VDD_DRA
MNVCC_I2C
NVCC_UAR
T
PVCC2_1P
8
VDD_MIPI
_0P9
VDD_MIPI
_1P2
VDD_PCI_
0P8
VDD_USB_
0P8NC_J18 VSS
NVCC_SNV
S_1P8
TSENSO
R_TEST_
OUT
TSENSO
R_RES_
EXT
VSSCLKOUT
2CLKIN2 J
KDRAM_D
Q13
DRAM_D
Q12VSS
DRAM_A
C02
DRAM_A
C14
DRAM_A
C09VSS
NVCC_DRA
M
NVCC_DRA
MVDD_VPU VDD_VPU VDD_SOC VDD_SOC
VDD_USB_
3P3VSS VSS
VDD_SNVS
_0P8
NAND_D
ATA02
NAND_D
ATA01VSS
NAND_D
ATA06
NAND_C
LEK
LDRAM_A
C05
DRAM_A
C04
NVCC_DRA
M
VDD_DRA
MVDD_VPU VSS VDD_VPU VDD_SOC VSS
VDD_ANA
_0P8VDD_SOC
NVCC_JTA
G
NAND_D
ATA05
NAND_C
E3_BL
MDRAM_A
C16
DRAM_A
C17
NVCC_DRA
MVSS VDD_VPU VDD_VPU VDD_SOC VSS NVCC_CLK
NAND_D
ATA04
NAND_C
E2_BM
NDRAM_A
C26
DRAM_A
C19VSS
DRAM_A
C12
DRAM_A
C13
DRAM_A
C15VSS
NVCC_DRA
M
NVCC_DRA
M
VDD_DRA
MVDD_VPU VSS VDD_SOC VDD_SOC VSS
VDD_ANA
_0P8VDD_SOC
VDD_24M
_XTAL_1P8
VDD_ANA
1_1P8VSS
NAND_A
LE
NAND_D
ATA03
NAND_C
E0_BVSS
NAND_D
ATA07
NAND_R
E_BN
PDRAM_V
REF
DRAM_Z
NVSS
VDD_DRA
M_PLL_1P
8
NVCC_DRA
M
VDD_DRA
M_PLL_0P
8
VDD_GPU VSS VSSVDD_ARM
_PLL_0P8
VDD_ANA
1_1P8VSS
NAND_D
ATA00VSS
NAND_R
EADY_B
NAND_C
E1_BP
RDRAM_R
ESET_N
DRAM_A
LERT_NVSS
DRAM_A
C32
DRAM_A
C33
DRAM_A
C27VSS
NVCC_DRA
M
NVCC_DRA
M
VDD_DRA
MVDD_GPU VSS VDD_ARM VDD_ARM VSS VDD_SOC VDD_SOC
VDD_ARM
_PLL_1P8VSS VSS
NAND_D
QS
SD1_RE
SET_B
SD1_STR
OBEVSS
NAND_W
E_B
NAND_W
P_BR
TDRAM_A
C34
DRAM_A
C35
NVCC_DRA
MVSS VDD_ARM VDD_ARM VDD_ARM VSS
PVCC1_1P
8
SD1_DAT
A3
SD1_DAT
A2T
UDRAM_A
C25
DRAM_A
C24
NVCC_DRA
M
VDD_DRA
MVDD_GPU VSS VDD_ARM VDD_ARM VSS VDD_SOC VDD_SOC
NVCC_NA
ND
SD1_DAT
A5
SD1_DAT
A4U
VDRAM_D
Q21
DRAM_D
Q20VSS
DRAM_A
C23
DRAM_A
C36
DRAM_A
C29VSS
NVCC_DRA
M
NVCC_DRA
MVDD_GPU VDD_ARM VDD_ARM VDD_ARM
NVCC_SAI
2NVCC_SD1 VSS NVCC_SD2
SD2_DAT
A3
SD2_DAT
A2VSS SD1_CLK
SD1_CM
DV
WDRAM_D
Q18
DRAM_D
Q19VSS
DRAM_A
C22
DRAM_A
C37
DRAM_A
C28VSS
VDD_DRA
MVDD_GPU
NVCC_GPI
O1VDD_ARM VDD_ARM VDD_ARM VDD_ARM
NVCC_SAI
5
NVCC_SAI
1VSS
NVCC_ENE
TSD2_CLK
SD2_CM
DVSS
SD1_DAT
A7
SD1_DAT
A6W
YDRAM_D
QS2_NVSS
NVCC_SAI
3VSS
VDD_ANA
0_1P8VSS
SD1_DAT
A1
SD1_DAT
A0Y
AADRAM_D
QS2_P
DRAM_D
Q17VSS VSS VSS VSS
VDD_ANA
0_1P8VSS VSS VSS VSS
SD2_CD_
BSD2_WP AA
ABDRAM_D
M2
DRAM_D
Q16VSS
DRAM_A
C20
DRAM_A
C21
DRAM_A
C38
GPIO1_I
O15
GPIO1_I
O12
PVCC0_1P
8
SAI5_RX
FS
SAI1_MC
LK
SAI1_TX
FS
SAI2_RX
C
SD2_DAT
A0
SD2_DAT
A1VSS
SD2_RE
SET_B
ENET_M
DIOAB
ACDRAM_D
Q22
DRAM_D
Q23VSS
DRAM_A
C30
SAI3_TX
FS
GPIO1_I
O14
GPIO1_I
O11
SAI5_RX
D3
SAI5_RX
D1
SAI5_RX
C
SAI1_TX
C
SAI2_RX
FS
SAI2_TX
D0
SAI2_RX
D0VSS
ENET_R
D3
ENET_M
DCAC
ADDRAM_D
Q27
DRAM_D
Q26
DRAM_A
C31
SAI3_MC
LK
GPIO1_I
O13
GPIO1_I
O10
SAI5_RX
D2
SAI5_MC
LK
SAI5_RX
D0
SAI2_MC
LK
SAI2_TX
C
SAI2_TX
FS
ENET_R
D2
ENET_R
D1AD
AEDRAM_D
Q28VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS
ENET_R
XC
ENET_R
D0AE
AFDRAM_D
Q29
DRAM_D
QS3_NVSS
DRAM_D
Q31
DRAM_D
Q25
SAI3_TX
D
SAI3_RX
D
SPDIF_E
XT_CLK
SPDIF_T
X
GPIO1_I
O09
GPIO1_I
O07
GPIO1_I
O05
GPIO1_I
O03
GPIO1_I
O01
SAI1_RX
D1
SAI1_RX
C
SAI1_RX
D3
SAI1_RX
D5
SAI1_RX
D7
SAI1_TX
D1
SAI1_TX
D3
SAI1_TX
D5
SAI1_TX
D7
ENET_T
X_CTL
ENET_T
D3
ENET_T
D1
ENET_R
X_CTLAF
AG VSSDRAM_D
QS3_P
DRAM_D
Q30
DRAM_D
M3
DRAM_D
Q24
SAI3_TX
C
SAI3_RX
C
SAI3_RX
FS
SPDIF_R
X
GPIO1_I
O08
GPIO1_I
O06
GPIO1_I
O04
GPIO1_I
O02
GPIO1_I
O00
SAI1_RX
D0
SAI1_RX
FS
SAI1_RX
D2
SAI1_RX
D4
SAI1_RX
D6
SAI1_TX
D0
SAI1_TX
D2
SAI1_TX
D4
SAI1_TX
D6
ENET_T
XC
ENET_T
D2
ENET_T
D0VSS AG
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
Page 25
COMPANY PUBLIC 24
i.MX 8M Mini Preliminary Power Measurements
Assumptions
• Power consumption use cases
have been measured using 4.9
GA Linux kernel.
• Further power optimizations will
be made with 4.14 GA Linux
kernel.
• Measurements have been taken
on the EVK, but require re-work
if the customer wishes to
reproduce the results.
• All power measurements have
been taken on typical process
silicon at room temperature
26C.
• Details on the use cases can be
found in the preliminary
application note posted on the
i.MX 8M Mini Alpha SharePoint.
0
500
1000
1500
2000
2500
DSM(11mW)
Idle(152m
W)
AudioPlaybac
k(345m
W)
Audio +Video
Playback
(1W)
GPU(MM07)(1.4W)
Coremark
(1.6W)
Stream(1.8W)
4-coreDhryst +VPU +Taiji
(2.3W)
NVCC_DRAM 1.762 37.507 91.315 256.824 308.923 64.542 434.085 323.289
VDD_GPU_VPU_DRAM 1.729 28.805 104.242 468.149 509.219 353.581 445.473 565.835
VDD_SOC 8.116 84.863 106.263 220.244 224.722 173.547 199.951 241.588
VDD_ARM 0 1.376 43.899 101.801 410.681 1036.201 765.193 1204.187
ME
AS
UR
ED
PO
WE
R (
MW
)
i.MX 8M Mini Power Consumption (Preliminary)
Page 26
COMPANY PUBLIC 25
i.MX 8M Family for Industrial and Media IoT
Family Namei.MX 8M Quad/QuadLite
i.MX 8M Dual
i.MX 8M Mini Quad/QuadLite
i.MX 8M Mini Dual/DualLite
i.MX 8M Mini Solo/SoloLite
i.MX 8M Nano Quad/QuadLite
i.MX 8M Nano Dual/DualLite
i.MX 8M Nano Solo/SoloLite
Sample / Production In Production – www.nxp.com/imx8m In Production – www.nxp.com/imx8mmini Announced – Sample Q2-19 / Production Q4-19
Main CPU 2x or 4x Cortex-A53 @ 1.5 GHz, 1MB L2 1x, 2x or 4x A53 @ 1.6-1.8 GHz, 512KB L2 1x, 2x or 4x A53 @ 1.2-1.5 GHz, 512KB L2
MCU Cortex-M4 266MHz, 256KB TCM Cortex-M4 400MHz, 256KB TCM Cortex-M7 600+MHz, 256KB TCM
DDR x16/x32 LPDDR4/DDR4/DDR3L x16/x32 LPDDR4/DDR4/DDR3L x16 LPDDR4/DDR4/DDR3L
GPUGC7000Lite (4 shaders)
OpenGL ES 2.0/3.0/3.1, Vulkan, OpenCL 1.2
GC NanoUltra 3D (1 shader) + GC320 2DOpenGL ES 2.0
GC7000UL (2 shaders) (*)OpenGL ES 2.0/3.0/3.1, Vulkan, OpenCL 1.2
Security CAAM, RDC, TrustZone CAAM, RDC, TrustZone CAAM, RDC, TrustZone
AI/ML OpenCL CPU: 32 GOPS A53 OpenCL CPU, GPU: 32 GOPS
SRAM 128KiB + 32KiB 256KiB + 32KiB 512KiB + 32KiB
Display Features 4K HDR, DCSS, LCDIF LCDIF LCDIF
Display Interfaces 1x MIPI-DSI, HDMI 2.0a Tx (ARC) + eDP 1x MIPI-DSI 1x MIPI-DSI
HDR HDR10, HLG, Dolby Vision - -
Camera Interface 2x MIPI-CSI (4-lanes each) 1x MIPI-CSI (4-lanes) 1x MIPI-CSI (4-lanes)
Video Decode 4Kp60 HEVC H.265, VP9, H.264 (*), legacy 1080p60 HEVC H.265, VP9, H.264, VP8 (*) -
Video Encode - 1080p60 H.264, VP8 (*) -
Connectivity
Audio Interface 20 I2S TDM (32b @ 384kHz), S/PDIF Tx+Rx20 I2S TDM (32b @ 384kHz), S/PDIF Tx+Rx, 8ch PDM
DMIC input
12 I2S TDM (32b @ 384kHz), S/PDIF Tx+Rx, 8ch
PDM DMIC input, ASRC
Expansion IO 2x USB3.0 Type C, 2x PCIe Gen2 2x USB 2.0, 1x PCIe Gen2 1x USB2.0
Network/Storage 1x GbE, 2x SDIO/eMMC, Raw NAND, QSPI 1x GbE, 3x SDIO/eMMC, Raw NAND, QSPI 1x GbE, 3x SDIO/eMMC, Raw NAND, QSPI
Process Techology TSMC 28HPC Samsung 14LPC FinFET Samsung 14LPC FinFET
Qualification Commercial, Industrial Commercial, Industrial Commercial, Industrial
Packages 17x17mm, 0.65p 14x14mm 0.5p (no microvias)*pin-compatible with i.MX 8M Nano
14x14mm 0.5p (no microvias)*pin-compatible with i.MX 8M Mini
Blue text = difference
* Not available on “Lite” version
4Q’19
Pin Compatible
External Module Options for Wi-Fi/BT
Page 27
COMPANY PUBLIC 26
Smart IoT Home Blocks
High Speed Interfaces
Smart Performance
Peripheral InterfacesLPDDR4,
DDR4 or
DDR3L
Audio
DACs
S/PDIF
MIPI DSI
PCIe
or SDIO
RGMII
SDIO
USB
SD Card
Wi-Fi
BT Combo
GbE PHY
USB TypeC
USB 2.0
Analog Audio
Up to 3000MTS
x16 or x32
4xCortex-A53
@1.8GHz
VPU
+
GPU
+
Audio
Interface
s
+
Camera
High
Speed
Interfaces
MIPI CSI
i.MX 8M Mini
Page 28
COMPANY PUBLIC 27
Making It Easy for You
Page 29
COMPANY PUBLIC 28
i.MX 8M Mini & 8M Nano Evaluation KitIn Production
Kit Contents
- i.MX 8M Mini/Nano Board (two
board stack comprising a compute
module plus base board). Compute
module not orderable separately.
- MIPI-DSI to HDMI cable.
- USB 3.0 to Type C cable.
- USB Type C power supply.
Compute Module: Overview
• NXP i.MX 8M Mini/Nano Quad
• Murata Wi-Fi Type 1PJ (QCA9377)
or 1MW (CYW43455)
• PMIC Rohm BD71847AMWV
• 2GB LPDDR4 or DDR4
• 16GB eMMC 5.1 or NAND
• 64MB QSPI Flash
• 8-layer PCB, No HDI
• Size 2”x2”
OS Support
• Linux, Android and FreeRTOS
BSPs from NXP
• Others: 3rd parties
Base Board: Overview
• Display Connector: 1x mini-
SAS MIPI-DSI
• Camera connector:1x mini-
SAS MIPI-CSI
• Audio DAC (WM8524)
• Microphone/headphone jacks
• 1x full-size SD/MMC card slot
• 10/100/1000 Ethernet port
(AR8031 Gigabit PHY)
• 1x USB 3.0 Type C
• Connectivity expansion
connector:
• M.2 connector (PCIe)
• Audio expansion connector:
• FPC 0.5mm pitch 60pin ZIF
• UART, I2S
• General purpose expansion
connector (RPI-like):
• UART, I2S
• NXP PN7150RPI board
supported via this interface
• Size 4.2”x4.2”
OV5640 5Mpix MIPI CSI Board
miniSAS based
MINISASTOCSI
Optional Add-ons
Base Kit: Compute Module + Base Board
i.MX 8M Mini Board
Part Numbers:
1. Config 1: 8MMINILPD4-EVK LPDDR4 / eMMC / Type1PJ (QCA9377) $399 (Now)
2. Config 2: 8MMINID4-EVK DDR4 / NAND / Type1MW (CYW43455) $399 (Availability in Q219)
3. Config 3: 8MNANOD4-EVK DDR4 / eMMC / Type1MW (CYW43455) $399 (Availability in Q419)
MIPI-DSI to HDMI
Connectivity• Murata Type 1PJ (QCA9377)
• Dual band 2.4G/5G
• 802.11a/b/g/n/ac (1x1)
• BT/BLE 4.1
• Murata Type 1MW (CYW43455)
• Dual band 2.5G/5G
• 802.11a/b/g/n/ac (1x1)
• BT/BLE 5.0
2 FLAVORS
OLED MIPI DSI Board
miniSAS based
5.5”, 1920 x 1080
MX8-DSI-OLED1
Page 30
COMPANY PUBLIC 29
Alexa Voice ServicesGoogle Voice Assist
NXP i.MX 7D
2-Mic Dev Kit for Amazon AVS
Voice Solutions for China
NXP i.MX 8M
Largest Voice Solution
provider for Mandarin
Enabling Audio & Voice Processing at the Edge 1,2,3
“Voice control front end to grow at 29% CAGR (‘17-’22) in the Smart Home”
1. Source: Google Android Things developer website2. Amazon AVS developer website 3. Source: ABI Research
NXP Enabled
Smart Speakers
Page 31
COMPANY PUBLIC 30
Kit Hardware Software
SOM + Baseboard: Pico-Pi i.MX8M Yocto 4.14 Linux
Mic Array: 2 Mic Voice HAT Alexa Device SDK & Wake Word
Antennae, Speakers + cables Audio Front End: SW based
Kit focus1) Easy out of box hardware set up
2) Detailed User Guide
3) Software set up simplified by
scripts to automatically install
3rd party libraries at runtime
NXP i.MX8M Development Kit for Amazon Alexa Voice
Service
Page 32
COMPANY PUBLIC 31
Alexa Multi-Room Music
Page 33
COMPANY PUBLIC 32
What is Alexa Multi-Room Music? (MRM)
Alexa, play jazz
music everywhere!
Page 34
COMPANY PUBLIC 33
Why is MRM Important?
Customers love whole-home music experiences – they may expect your product
to have MRM when they purchase it, and could experience frustration if it
doesn’t.
Page 35
COMPANY PUBLIC 34
Alexa Multi-Room Music
Interoperate with Echo Family
of Devices and other MRM-
enabled Alexa Devices
Direct interaction
Cloud-Controlled Speakers
Proprietary Multi-Room Music
Connect your cloud to Alexa using
Connected Speaker API skill
Implicit and explicit targeting
For Speakers, Soundbars, AVRs and Home Theater Systems
Options for Multi-Room Music (MRM)
Page 36
COMPANY PUBLIC 35
Typical Device Interaction
AVS Client
Alexa, play Despacito!
Media Player
MSP
Page 37
COMPANY PUBLIC 36
Alexa Targeted Music Playback
Media Player
AVS ClientMSP
AVS Client
Alexa, play Dubstep in the living room
Speaker name defined as
“Living Room” in Alexa App
Page 38
COMPANY PUBLIC 37
Alexa Multi-Room Music
Media Player
AVS Client
MRM (Receiver)
Media Player
AVS Client
MRM (Receiver)
Media Player
AVS Client
Encrypted, encoded unicast from sender to receiver
MSP
AVS Client
Alexa, play Trap
Music downstairs!
MRM (Sender)
MRM (Receiver)Media
Player
All speakers defined as part of group “Downstairs”
Page 39
COMPANY PUBLIC 38
Cloud-connected Speakers
3rd Party Sync
ReceiverMedia Player
Your Client
Your Sync
MSP
Your Cloud
AVS Client
3rd Party Sync
Receiver
Alexa, play my EDM
playlist in the back yard!
Your speakers defined as group “Back Yard”
Page 40
COMPANY PUBLIC 39
Linux OS
ARM v7+ Processor +450 DMIPS
+40 MB RAM and +30 MB Flash
DAC sync’d to host processor clock
Low drift oscillator
Platform Considerations for MRM
Page 41
COMPANY PUBLIC 40
MRM KPI – Group 2 Fail
Page 42
COMPANY PUBLIC 41
MRM KPI – Group 1 Pass Room Level
Page 43
COMPANY PUBLIC 42
Polk Audio added MRM with an OTA update
and customers were delighted!
Spotlight – Polk Audio Command Bar
Page 44
COMPANY PUBLIC 43
Next Steps – MRM Links and Resources
• Contact solution providers with MRM experience – alexa.design/SP
• Experiment with MRM qualified development kits – alexa.design/devkits
• Ask your Amazon Point of Content for access to MRM resources
• Learn more at alexa.design/speakers and alexa.design/MRM
• New to AVS? Build your first prototype at alexa.design/avstutorials
Page 45
COMPANY PUBLIC 44
Google Coral Dev Kit TPU + i.MX 8M
• Object Recognition
• Facial Recognition
• Gamming & Learning
• Full System on Board: SOC + ML + Connectivity all on
the board, running a derivative of Debian Linus (Mendel), so
you can run your favorite Linux tools with it.
• Direct ML Model Support: TFLite models can be
compiled and run on the Coral Dev. Board.
• Scale from Prototype to Production: The SOM can
be removed from the baseboard, ordered in bulk and
integrated into production hardware.
Page 46
NXP and the NXP logo are trademarks of NXP B.V. All other product or service names are the property of their respective owners. © 2019 NXP B.V.