NCC S1 Neural Network Computing Card
Based on the AI-specific APiM framework, a modular deep neural network learning accelerator without any externalcaching can be used for high-performance edge computing, as a vision-based deep learning computing and AI algorithmacceleration. NCC S1 is small in shape, extremely low in power consumption and best peak performance. Equipped withcomplete and easy-to-use model training tools, network training model instances, and professional hardware platform, itcan be quickly applied in the artificial intelligence industry.
5.6Tops Best Peak Performance
Based on the AI-embedded Neural Network Processor (NPU), the NCC S1 possesses 28,000 parallel neural computing coresand supports on-chip parallel and in-situ calculations. Its peak up to 5.6Tops, dozens of times higher than othersolutions on the market. It can afford complex high-density calculations for high-performance edge computing field.
AI Processing Framework APiM
Based on AI-specific MPE matrix engine and APiM (AI processing in Memory) framework, it deals with AI in arevolutionary way. Without any instructions, bus and external DDR cache, plenty of data can be directly input or outputto the silicon chip by upgrading the network preloading once, which greatly lifts the processing speed of AI andreduces the processing energy consumption.
9.3Tops/W High Energy Efficiency
The NPU of NCC S1 neural network computing card uses the 28nm process technology. The power is only 300mW whenthroughput is 2.8 Tops, while the energy efficiency is up to 9.3 Tops/W. It maintains strong computing ability whileowning extremely low energy consumption, endowed with great advantages in the edge computing field of terminalequipment.
High-performance Hardware Platform
NCC S1 neural network computing card can be equipped with ROC-RK3399-PC open source main board. On condition that it isstocked with high-performance RK3399 six-core processor and abundant hardware interface, it can rapidly integratehardware platform for edge computing, set up product prototype, and thus accelerate AI product project process.
Supporting Model Training Tools
Provide the complete and easy-to-use model training tool PLAI (People Learn AI) which based on PyTorch, it can bedeveloped on Windows 10 and Ubuntu 16.04 systems to add custom network models more easily and quickly, which greatlyreduces the technical difficulties to applying AI and makes AI technology accessible to more people.
Provide Network Training Model
Support the following three network training model examples such as GNet1, GNet18 and GNetfc with more networkinstances continuing to emerge subsequently, making it possible to easily test a large number of deep learningapplications on the device.
|NPU||Lightspeeur SPR2801S (28nm process, unique MPE and APiM architecture)|
|Low Power||2.8 TOPs@300mW|
|Platform||Applicable ROC-RK3399-PC platform|
|Framework||Support Pytorch, Caffe framework, follow-up support TensorFlow|
|Tools||PLAI model training tool(Support for GG1, GNet18 and GNetfc network models based on VGG-16)|
Support Ubuntu, Windows operating system
ROC-RK3399-PC adopts Rockchip high-performance core configuration, unique multiple power supply mode and unique plate design.It can be connected to the expansion board, making the performance stronger and superior. When combined withthe metal casing, it becomes a pocket portable personal computer.
With"server-level"dual-core Cortex-72+ quad-core Cortex-A53 architecture, the frequency is up to 2.0 GHz, and 4GB LPDDR4 dual-channel64-bit RAM high-performance memory is configured to comprehensively improve the performance of mainboard.
Uniqueplate design, golden ratio, only 120 x 72 x 11.9 mm in size. When combined with the metal casing, it becomesa pocket portable personal computer.
The expansion board onboard balanced charging circuit to charge the battery, and it can also be powered by POE+. It has M.2M-Key interface to expand SSD and M.2 E-Key interface to provide SDIO 3.0 and USB 2.0 signals. ROC-RK3399-PCmain board combined with the expansion board can greatly improve the performance.
Supportsxserver, wayland display framework and multiple operating systems, such as Android, Ubuntu, Debian9, Linux+QT,etc. It has onboard SPI flash and supports boot with TF card, EMMC, SSD, USB flash disk, making system startupmore convenient.
ROC-RK3399-PC can be powered by POE+ (802.3 AT, output power 30W) or dual-cell battery. It has a Type-C PD 2.0 power controlchip and supports wide voltage (5V~15V) input. A variety of power supply modes can meet the choices of usersin different scenarios.
It has rich interfaces, such as MIPI/eDP screen interface, dual-channel MIPI CSI(supportingdual-channel cameras), Type-C x 2, USB 2.0 Host x 3, HDMI 2.0, Gigabit Ethernet (RJ45), GPIO, etc.
Rockchip RK3399 (28 nm HKMG manufacture procedure)
Six-core ARM® 64-bit processor. Main frequency is up to 2.0 GHz
Based on the big.LITTLE core architecture, dual-core Cortex-A72 (big core) + quad-core Cortex-A53 (little core)
ARM Mali-T860 MP4 quad core GPU
Support OpenGL ES1.1/2.0/3.0/3.1, OpenVG1.1, OpenCL, DX11
Support AFBC (frame buffer compression)）
Support 4K VP9 and 4K 10bits H265/H264 video decoding, up to 60 fps
1080P multi-format video decoding (WMV, MPEG-1/2/4, VP8)
1080P video encoding, support H.264, VP8 formats
Video post processor, de-interlacing, de-noising, edge/detail/color optimization
RK808-D PMU chip
4GB LPDDR4 dual-channel 64-bit RAM
Onboard SPI flash (16M Byte)
16 GB/32 GB/128 GB high-speed eMMC (optional)
Support MicroSD (TF) card, USB flash disk expansion
Support expansion board M.2 M-Key to expand SSD
Gigabit Ethernet (RJ45 interface)
Through the expansion board M.2 E-Key interface (providing SDIO 3.0 signal)
HDMI2.0 supports 4K 60Hz display, support HDCP 1.4/2.2
DisplayPort 1.2 ( up to 4K 60 Hz)）
Support eDP 1.3.
Support MIPI-DSI (dual-channel)
Support dual-screen display (Type-C + HDMI), support 4K + 2K output
1 x HDMI audio frequency output, DP audio frequency output
Dual-channel MIPI CSI interface (supporting dual-channel camera at the same time)
USB 2.0 Host × 3、Type-C × 2
Debug serial port × 1, for development and debugging
Recovery × 1、Power × 1
One channel infrared receiving head, support infrared remote control function
RTC real time clock x 1, onboard battery socket
1) POE+ (802.3 AT, output power 30 W)
2) Type-C PD 2.0 power control chip, support wide voltage input (5 V-15 V)
3) Dual battery power supply (7.4 V lithium battery)
Support 4K@60fps output (not supporting simultaneous output)
TypeC0 interface: Support DisplayPort 1.2 + Power Delivery 2.0+ USB3.0 OTG + USB2.0 Host,
Board power and DP output can be achieved simultaneously using a Type-C line
TypeC1 interface: Support DisplayPort 1.2 + USB3.0 HOST+ USB2.0 HOST, support external power supply
1) POE+ (802.3 AT, output power 30 W)
2) M.2 M-Key (extending SSD) and M.2 E-Key interface (providing SDIO 3.0, USB 2.0 signals)
3) Onboard battery (dual 7.4V lithium battery)
4) Onboard balanced charging circuit to charge the onboard battery
Support Ubuntu18.04 (supporting OpenGL es3.0, OpenCL, TensorFlow Lite),
Debian9, Linux+QT, support xserver and wayland display frameworks
Android 8.1 (supporting Android NN API)）
Onboard SPI flash, support boot from TF card, EMMC, SSD, USB flash disk
120 x 72 x 11.9（ mm）