Junzheng A1 Embedded NVR Function Board Overview

1. Product Overview

1.1 Application Scope of Junzheng A1 Function Board

The Junzheng A1 is an embedded NVR function board that uses standard DC12V power supply, supports dual SATA hardware interfaces, and is widely applicable to smart NVRs, video decoders, soft routers, and related products.

1.2 Product Overview

The Junzheng A1 features a 1.4GHz dual-core CPU, supporting 4K@90fps decoding, 4K native display, dual Gigabit Ethernet ports, and dual SATA interfaces. In response to the continuously upgrading intelligence requirements of the security industry, the A1 is equipped with 1.4T@int8/5.6@int4 AI computing power, meeting the needs for multi-channel video source synchronization for video structuring, detection, and recognition. Additionally, the A1 integrates key peripherals such as DDR, RTC, and Audio, significantly optimizing product implementation costs.

1.3 Product Features

Onboard dual SATA interfaces, dual Gigabit Ethernet ports, HDMI output, and VGA output. Supports inux+QT system customization, provides system call interface API reference code, perfectly supporting customer upper application APP development and SDK.

2. Product Performance

2.1 Main Chip Block Diagram

2.2 Main Chip Parameters

2.2.1 CPU

XBurst®2 up to 1.2GHz, dual-core, dual-issue, high-performance and low-power implementation of MIPS32 ISA R5

MIPS32 ISA R5 plus Junzheng SIMD512 ISA

Dual-issue, superscalar, superpipelined with simultaneous multithreading (SMT)

— Two hardware threads for each physical core

— Fetch four instructions per cycle

— Dual-issue instructions per thread per cycle

32K L1 D-cache + 32K L1 I-cache, 128~1024K L2 cache

High-performance floating-point unit and SIMD engine: FSE

– Two hardware threads FPU and SIMD units

– Dual-issue FPU and single-issue SIMD units

– 32×512-bit register set per hardware thread, 256-bit load/store to/from SIMD units

– Compliant with IEEE-754 2008 programmable Memory Management Unit (MMU)

– 1st level mini-TLBs (MTLBs)-8×2 entry instruction TLB, 16×2 entry data TLB

– 2nd level TLB: 32×2 entry VTLB, 256×2 entry 4-way set associative XBurst®2 processor system supports little-endian only

2.2.2 MCU

600MHz RISC-V co-processor

32bit, in-order, 5-stage pipeline core

32K L1-cache and 32K L1 D-cache

RV32IM instruction set architecture

2.2.3 AI Engine

Built-in neural network accelerator

Typical performance: 16Tops@int2, 4Tops@int4, 1Tops@int8

Supports int2N-bit (N=1~8) width; shared 1MB memory pool

Magic AI algorithm development platform available

2.2.4 Artificial Intelligence Preprocess (AIP)

Supports multifunctional graphics processing, aip_p (affine/perspective transformation processing), aip_f (scaling processing), aip_t (format conversion processing)

Supports configuration of multiple storage types, data can be accessed from DDR, ORAM

aip_p, aip_f, aip_t can run multithreaded, supports thread protection

2.2.5 Video Decoder & Post Processing

Supports H.264/H.265 video decoder

Supports maximum resolution up to 4K (4096×4096)

Maximum performance: 4096×2160@60fps; supports time-division multiplexing

Pass-through via video decoder to save bandwidth; scaling, cropping

Color conversion, input format: NV12, output formats: RGB888, BGR888, R16G16B16, B16G16R16

Supports dual-channel simultaneous output with different scaling ratios

– First channel supports both reduction and enlargement, input resolution (VGA-2160P), output resolution (VGA-2160P), maximum 3x enlargement

– Second channel only supports reduction, input resolution (VGA-2160P), output resolution (VGA-1080P)

2.2.6 JPEG Codec

Supports encoder and decoder

Supports maximum resolution: 4K (4096×4096)

Encoder maximum performance: 4K@30fps, decoder maximum performance: 4K@30fps

2.2.7 High Definition Multimedia Interface (HDMI)

Video formats:

– All CEA-861-E video formats up to 1080p at 60 Hz and up to 720p/1080i at 120 Hz

– Optional HDMI 1.4b video formats: (depending on configuration)

All CEA-861-E video formats up to 1080p at 120 Hz

HDMI 1.4b 4K x 2K video formats

HDMI 1.4b 3D video mode, up to 340 MHz (TMDS clock)

– Optional HDMI 2.0 video formats: (depending on configuration)

All CEA-861-F video formats

Dynamic Range and Mastering InfoFrame (DRM, packet header 0x87)

Colorimetry:

– 24/30/36/48 bit RGB 4:4:4

– 24/30/36/48 bit YCbCr 4:4:4

– 16/20/24 bit YCbCr 4:2:2

– 24/30/36/48 bit YCbCr 4:2:0

– xvYCC601

– xvYCC709

– Optional HDMI 1.4b colorimetry:

SYCC601, Adobe RGB

Adobe YCC601

Optional color space converter (CSC):

– RGB (4:4:4) to/from YCbCr (4:4:4 or 4:2:2)

Optional HDMI 1.4b supported InfoFrames:

– Audio InfoFrame packet extension to support LFE playback level information

– AVI infoFrame packet extension to support YCC quantization range (Limited Range, Full Range)

– AVI infoFrame packet extension to support content types (graphics, photos, movies, games)

– NTSC VBI infoFrame packet extension to support SCTE 127 payload containing VBI data

Audio formats:

– Uncompressed audio format: IEC60958 L-PCM audio samples

– Compressed audio formats: IEC61937 compressed nonlinear PCM (AC-3, MPEG-1/-2 audio, DTS®)

– MPEG-2/-4 AAC, ATRAC, WMA, MAT)

– HBR audio formats: such as Dolby®True-HD and DTS-HD Master Audio

– Multi-stream audio (L-PCM or IEC61937 compressed nonlinear PCM)

Up to 192 kHz IEC60958 audio sampling rate. Applicable to IEC61937 compressed audio

– HDMI 2.0a: up to 1536 kHz

– HDMI 1.4b: up to 768 kHz

Pixel clock from 13.5 MHz to 600 MHz

Optional removal of pixel duplicate clock from HDMI Tx interface for compatibility with third-party HDMI Tx PHY

Flexible synchronization enable settings for each clock domain functional power-down mode

AMBA APB 3.0 register access

I2C DDC, EDID block read mode

SCDC I2C DDC access

TMDS Scrambler supports 2160p@60Hz with RGB/YCbCr 4:4:4 or YCbCr 4:2:2

YCbCr 4:2:0 support enables 2160p@60Hz at lower HDMI link speeds

Integrated CEC hardware engine

Advanced PHY testability

Synopsys and external PHY interface

Configurable number of rising edge-triggered flip-flops connected for data synchronization

Single-channel DVI 1.0 backward compatibility

2.2.8 Video Display Engine (VDE)

Supports two video layers, GUI layer and mouse layer overlay

Supports HDMI up to 4K@60fps video output, while supporting output formats of RGB, YUV444, YUV422, YUV420

Supports VGA up to 2K@60fps video output, while supporting RGB output format

Supports up to 2K@60fps TFT video output, while supporting RGB output format

Supports simultaneous HDMI and VGA outputs

Supports V0 video layer output enlargement, final output enlargement

V0 video layer supports up to 36 video streams

V1 video layer supports up to 2 video streams

Supported input formats for video layers: NV12/21, NV16/61

Supported input formats for GUI layer: ARGB8888, ARGB1555, compressed images

Supported input formats for mouse layer: ARGB8888, ARGB1555

2.2.9 Image Process Unit (IPU)

Supports single image transfer, filling, flipping, mirroring, clipping, scaling, format conversion

Supports dual-image overlay, color keying, flipping, mirroring, clipping, scaling, format conversion

Supports image compression

Supports input/output formats: NV12, NV21, NV16, NV61, RGB444, RGB555, RGB565, RGB888, ARGB1555, ARGB4444, ARGB8565, ARGB8888

Supports overlay using M1 images

2.2.10 Serial Advanced Technology Attachment (SATA)

SATA 1.5Gb/s, SATA 3Gb/s speeds

Supports two ports; complies with the following specifications:

– Serial ATA 3.3 – AHCI Revision 1.3.1

– AMBA 2.0 and AMBA 3.0 AXI from ARM

Highly configurable PHY interface

User-defined PHY status and control ports

Recovery clock system RX data buffer

Data alignment circuitry

Active LED support

Power management features, including automatic partial sleep transition

BIST loopback mode

Internal DMA engine for each port

Disable RX and TX data clock in power-down mode

2.2.11 Gigabit Ethernet Media Access Controller (GMAC)

Full-duplex RGMII support (1Gbps/100Mbps/10Mbps)

Full-duplex RMII support (100Mbps/10Mbps)

IEEE 802.3az Section 78 (Energy Efficient Ethernet (EEE) features), standard support

1Gbps/100Mbps/10Mbps operation using 4-bit RGMII interface working at 125/25/2.5mHz

System side 64-bit data transfer interface

Little-endian support for sending and receiving data path

Enhanced DMA architecture for better performance and throughput

Supports TCP checksum offloading for received packets (IPv4 and IPv6)

MDIO Master interface (optional) for managing multiple PHY devices and their configurations

Supports TCP segmentation offloading (TSO) feature

2.2.12 Video Output

Video output

– Input format: NV12/NV21 8bit

– Output format: YUV422 8bit

– Output interfaces: 1. BT656 8bit parallel signal; 2. BT1120 8bit parallel signal

– Only supports progressive scan, interlaced scan not supported

– Supports color bar output test

2.2.13 Audio System

Integrated audio codec

– Built-in audio codec

– Single-channel recording, dual-channel playback

– Input supports single-ended or differential, output supports single-ended with echo cancellation

– Supports I2S interface

– Sampling rate supports 8K, 12K, 16K, 24K, 32K, 44.1K, 48K, 96K; AIC/I2S/CODEC

– Supports I2S/MSB-Justified mode serial input and output

– Sampling rate configurable: 8k, 12K, 16k, 24k, 32k, 44.1k, 48k, 96k

– Supports 16, 20, and 24-bit audio sample data sizes; 16-bit packed transmission data is supported

– Supports DMA transfer mode

– Supports interrupt mode to control CPU read/write data

– Supports dual-channel transceiving data, supports half-duplex and full-duplex

– Loopback left channel playback data for echo cancellation

– Supports internal codec and external codec, supports AIC as master/slave external codec

– Supports 6-wire mode and 4-wire mode in external codec

2.2.14 DDR Controller

Typical frequency of 800MHz

DDR PHY can connect one or two KGD

DDR maximum capacity: 4GByte

Memory type: DDR3

2.2.15 Memory Interface

On-chip integrated 2Gbit DDR3

2.2.16 System Functions

Clock generation and power management

– On-chip 2MHZ~30MHZ oscillator circuit, recommended 24MHZ oscillator circuit

– A four-chip phase-locked loop (PLL) with programmable multiplier

– CCLK, HHCLK, H2CLK, H1CLK, PCLK, H0CLK, DDR_CLK frequencies can be changed by setting registers for software

– Functional unit clock gating; timer/counter unit (TCU)

– Includes two TCUs (TCU0, TCU1)

– Includes 0/1/2/3/4/5/6/7 channels (A1 uses channels 0~1)

– General mode: counter counts on clock posedge or negedge or doubling

– Gate mode: counter counts when gate is 0

– Quadrature mode: counter counts due to quadrature input

– Direction mode: counter adds or subtracts due to signal input

– Pos mode: counter starts counting from 0 due to posedge or negedge

– Capture mode: counter counts one cycle and outputs high level time and cycle time

– Internal clock supports 1/4/16/64/256/1024 prescaling

– Can only support exclk and input_clk

– It cannot work in sleep mode

– Does not support exclk 1 prescaling

PWM

– Includes 16 channels, counter clock up to 500MHz. Each channel can be enabled independently

– Two modes for updating PWM configuration: normal update mode and DMA update mode

OS Timer Controller

– 64-bit counter and 32-bit compare register

– Supports generating interrupts when the counter matches the compare register

– Two clock sources: RTCLK (real-time clock), HCLK (system bus clock), selectable 1, 4, 16, 64, 256, and 1024 clock prescaling

– Has counter full match interrupt – counts via external clock (EXCLK), can be 1, 4, 16 prescaled

– Read via synchronous system clock

Interrupt Controller

– Total of 64 interrupt sources

– Each interrupt source can be enabled independently

– Priority mechanism indicating the highest priority interrupt

– All registers are accessible by CPU

– Unmasked interrupts can wake up the chip in sleep mode

Watchdog Timer

– Generates WDT reset

– A 16-bit data register and a 16-bit counter

– WDT has a half-interrupt

– Counter clock uses software-selected input clock

Only RTCLK can serve as the clock for the counter

Clock prescale ratio can be set by software to 1, 4, 16, 64, 256, 1024

Programmable Direct Memory Access Controller

– Supports up to 32 independent DMA channels

– Descriptor or no-descriptor transfer mode compatible with previous positive SoC

– Transfer data unit: 1 byte, 2 bytes, 4 bytes, 16 bytes, 32 bytes, 64 bytes, 128 bytes

– Number of transfer data units: 1 ~ 224 – 1

– Independent source and destination port widths: 8 bits, 16 bits, 32 bits

– Fixed three channel group priorities: 0~3 highest; 4~11 medium; 12~31 lowest

– Bridge from AHB2 to APB

– Data transfer related peripherals: aic, uart0~2, ssi0~1, i2c0~1, des, etc.; OTP from interface

– Total of 2048 bits. And used as 1024 bits to ensure security; Power-On Reset (POR)

– Provides reliable reset functionality for general applications

– Monitors 1.8V power for IO and 0.9V for core

– Typical 1.35V threshold for 1.8V power

– Typical 0.6V threshold for 0.9V power

2.2.17 Peripherals

General I/O ports

– Input/output/function ports configurable

– Low/high, rising/falling edge triggered. Each interrupt source can be masked independently

– Four interrupts, each corresponding to a group, to INTC

– Each port is connected to an internal pull-up and pull-down resistor. Pull-up/pull-down resistors can be disabled

I2C Controller (SMB0, SMB1)

– Two-wire SMB serial interface – consisting of serial data line (SDA) and serial clock (SCL)

– Two speeds: standard mode (100kb/s), fast mode (400kb/s)

– Device clock same as pclk (APB clock) – programmable SCL generator

– Master or slave SMB operation

– 7-bit addressing/10-bit addressing, regardless of configuration as master or slave

– Send broadcast calls or start bytes

– Level send and receive 64 deep FIFO

– Interrupt operation

– The number of devices you can connect to the same SMB bus is limited only by the maximum total bus capacitance of 400pF

– APB interface; SPI Flash Controller (SFC)

– SPI protocol supports Standard, Dual, Quad SPI

– Only send or only receive operations

– MSB first for command and data transfer, LSB first for address transfer

– 64 entries x 32-bit wide data FIFO

– One device select

– Configurable receive sample point

– Configurable timing parameters: tSLCH, tCHSH, and tSHSL

– Supports configurable flash address range

– Transfer formats: standard SPI, dual output/dual input SPI, quad output/quad input SPI, dual I/O SPI, quad I/O SPI, full dual I/O SPI, full quad I/O SPI

– Two data transfer modes: slave mode and DMA mode

– Configurable 6 software process stages

Standard synchronous serial interface (SSI0, SSI1)

– Supports 3 protocols: National’s Micro-wire, TI’s SSP, and Motorola’s SPI

– Full-duplex or only send or only receive operations

– Programmable transfer order: MSB first or LSB first

– 128 entries deep x 32-bit wide sending and receiving data FIFO

– Configurable normal transfer mode or interval transfer mode

– Programmable clock phase and polarity for Motorola SSI format

– Back-to-back character transfer/receive mode

– Loopback mode for testing

– Two slave device select signals, supporting up to 2 slave devices

UART (UART0, UART1, UART2)

– Full-duplex operation

– 5, 6, 7, or 8-bit characters, optional no parity or even parity or odd parity, and 1, 1½, or 2 stop bits

– 64×8-bit sending FIFO and 64×11-bit receiving FIFO

– Independently controlled sending, receiving (data ready or timeout), line status interrupts

– Internal diagnostic capability including loopback control and interrupts, parity, overflow, and framing errors

– Separate DMA requests for sending and receiving data in FIFO mode

– Supports modem flow control via software or hardware

– Slow infrared asynchronous interface compliant with IrDA specification

– Only UART0 supports hardware flow control

MMC/SD/SDIO Controller (MSC0, MSC1)

– Supports SD memory and SD input/output (SDIO) digital interface protocols, compliant with SD HCI specifications

– Supports eMMC Command Queuing Engine (CQE), compatible with CQ HCI specifications

– Supports host version 4

– Supports eMMC specification 5.1

– Supports SD specification 4.1

– Supports SD I/O specification 2.0

– Maximum data rate of 100MBps

– Supports data widths: 1-bit (only dat0 used), 4-bit (dat0 and dat3 used)

– Built-in programmable divider for MMC/SD bus

– Multi-DMA modes: SDMA, ADMA2, ADMA3

– Multi-SD functionality support, including multiple I/Os and combined I/O and memory

– IRQ support enables card interrupts for MMC/SD controller

– Single or multiple block access to the card, including erase operations

– Streaming access for eMMC cards

– Supports SDIO read wait, interrupt detection during 1-bit or 4-bit access

– Supports command completion signal and interrupts to CPU

– Command completion signal disable feature

– Maximum block length of 4KB (4096 bytes)

USB 2.0 OTG Interface

– Port0 supports Battery Charge (BC), OTG (master-slave)

– Port1 supports OTG (master-slave)

– Port2 supports HOST mode only

– Complies with USB 2.0 standard for high-speed (480Mbps) functionality and USB 2.0 On-The-Go supplement

– Supports Session Request Protocol (SRP) and Host Negotiation Protocol (HNP).

– UTMI+ Level 3 transceiver interface

– 16 endpoints

– Supports UTMI+ Level 3 interface (Revision 1.0, February 25, 2004). 16-bit data is supported

– Supports high-speed (HS) mode only

– Supports control, interrupt, ISO, and bulk transfers.

Digital True Random Number Generator (DTRNG)

– Pure digital logic circuit

– True random number

– Interrupt mode and non-interrupt mode

2.2.18 Bootrom

22kB Boot ROM memory

2.3 Board Parameters

Memory: 256MB

Flash: 16MB

Size: 132x70mm

3. Product Peripherals

3.1 Appearance Diagram

Front

Junzheng A1 Embedded NVR Function Board Overview

Back

Junzheng A1 Embedded NVR Function Board Overview

Interfaces

Junzheng A1 Embedded NVR Function Board Overview

3.2 Board Interfaces

3.2.1 HDMI Interface x1

3.2.2 VGA Interface x1

3.2.3 3.5mm Audio Interface x1

3.2.4 Gigabit RJ45 Interface x2

3.2.5 USB2.0 Interface x2

3.2.6 Power Indicator/Working Indicator x2

3.2.7 DC Power Supply Interface x1

3.2.8 Reset Button x1

3.2.9 SATA Interface x2

3.2.10 D-Type Hard Disk Power Supply Interface x2

3.2.11 TTL Debug Serial Port x1

3.2.12 RTC Battery x1

SigmaStar Low Power Series Chip AI Performance Comparison

SigmaStar Low Power Series Chip AI Performance Comparison (II)

sigmastar& Junzheng & Lianying Solutions

Junzheng T40 NOR Fast AI Camera

Junzheng A1 Full HD Network Video Decoder

Junzheng A1 33-channel (dual network port) streaming device

Junzheng T31ZX+GC4653 4MP PTZ Fixed Focus Camera

sigmastar 120 frames high frame rate smart network camera WIFI module

SC8836Q 2MP HDMI HD Camera Module

SSC8836Q 2MP HDMI HD Camera Module/Customized Solution

sigmastar SSD268G 12MP low-light smart HD network camera module

sigmastar SSC377D 5MP black light full-color smart HD network camera module

RK & Allwinner & Hisilicon Decoding Solutions

RV1103/1106/1109/1126 & RK3562/3568/3576/3588 AI Performance Parameter Selection Guide

Rockchip RK3399 Android Smart Mainboard

Rockchip RK3288 Android 8.1 10.1 inch Dual Screen Smart Interactive Digital Signage

Rockchip RK3288 Wall-mounted WeChat Face Payment Module

Hisilicon Hi3516AV100 Core Board Development Board

Hisilicon Hi3516EV300 Development Board/Core Board

Hisilicon & RK AI Perspective Solutions

Hi3516/3519/3559/3403 AI Performance Parameter Selection Guide

Hisilicon Hi3520 Single SDI Input HDMI Output Encoding Board

Yizhi SV830C Dual Eye (RGB+NIR) Face Recognition

RK3566 Commercial Display Smart Universal Standard Mainboard Solution

Hisilicon Hi3520 Single SDI Input HDMI Output Encoding Board

For more information, please open the link or add the following WeChat, thank you for your attention!

What is Yiniu Network? Let you understand Yiniu Network in 1 minute!

Junzheng A1 Embedded NVR Function Board Overview

WeChat ID

Junzheng A1 Embedded NVR Function Board Overview

Long press the QR code to add WeChat

Join the AI technology communication group

Leave a Comment