ROHM Develops Ultra-Low-Power On-Device Learning Edge AI Chip，Enabling Real-time Failure Prediction Without Cloud Server Required

时间： 2022-12-01 来源：ROHM News

技术问答

ROHM has developed an on-device learning AI chip (SoC with on-device learning AI accelerator) for edge computer endpoints in the IoT field. It utilizes artificial intelligence to predict failures (predictive failure detection) in electronic devices equipped with motors and sensors in real time with ultra-low power consumption.

Generally, AI chips perform learning and inferences to achieve artificial intelligence functions. As learning requires that a large amount of data gets captured, compiled into a database, and updated as needed. So, the AI chip that performs learning requires substantial computing power that necessarily consumes a large amount of power. Until now it has been difficult to develop AI chips that can learn in the field consuming low power for edge computers and endpoints to build an efficient IoT ecosystem.

Based on an ‘on-device learning algorithm’ developed by Professor Matsutani of Keio University, ROHM’s newly developed AI chip mainly consists of an AI accelerator (AI-dedicated hardware circuit) and ROHM’s high-efficiency 8-bit CPU ‘tinyMicon MatisseCORE™’. Combining the 20,000-gate ultra-compact AI accelerator with a high-performance CPU enables learning and inference with an ultra-low power consumption of just a few tens of mW (1000× smaller than conventional AI chips capable of learning). This allows real-time failure prediction in a wide range of applications, since ‘anomaly detection results (anomaly score)’ can be output numerically for unknown input data at the site where equipment is installed without involving a cloud server.

Going forward, ROHM plans to incorporate the AI accelerator used in this AI chip into various IC products for motors and sensors. Commercialization is scheduled to start in 2023, with mass production planned for 2024.

Professor Hiroki Matsutani, Dept. of Information and Computer Science, Keio University, Japan
“As IoT technologies such as 5G communication and digital twins advance, cloud computing will be required to evolve, but processing all the data on cloud servers is not always the best solution in terms of load, cost, and power consumption. With the ‘on-device learning’ we research and the ‘on-device learning algorithms’ we developed, we aim to achieve more efficient data processing on the edge side to build a better IoT ecosystem. Through this collaboration, ROHM has shown us the path to commercialization in a cost-effective manner by further advancing on-device learning circuit technology. I expect the prototype AI chip to be incorporated into ROHM's IC products in the near future.”

About tinyMicon MatisseCORE™

tinyMicon MatisseCORE™ (Matisse: Micro arithmetic unit for tiny size sequencer) is ROHM’s proprietary 8-bit CPU developed for the purpose of making analog ICs more intelligent for the IoT ecosystem. An instruction set optimized for embedded applications together with the latest compiler technology to deliver fast arithmetic processing in a smaller chip area and program code size. High-reliability applications are also supported, such as those requiring qualification under the ISO 26262 and ASIL-D vehicle functional safety standards, while the proprietary onboard ‘real-time debugging function’ prevents the debugging process from interfering with program operation, allowing debugging to be performed while the application is running.

Detail of ROHM’s AI Chip (SoC with On-Device Learning AI Accelerator)

The prototype AI chip (Prototype Part No. BD15035) is based on an on-device learning algorithm (three-layer neural network AI circuit) developed by Professor Matsutani of Keio University. ROHM downsized the AI circuit from 5 million gates to just 20,000 (0.4% the size) to reconfigure for commercialization as a proprietary AI accelerator (AxlCORE-ODL) controlled by ROHM’s high-efficiency 8-bit CPU tinyMicon MatisseCORE™ that enables AI learning and inference with an ultra-low power consumption of just a few tens of mW. This makes the numerical output of ‘anomaly detection results’ possible for unknown input data patterns (i.e. acceleration, current, brightness, voice) at the site where equipment is installed without involving a cloud server or requiring prior AI learning, allowing real-time failure prediction (detection of predictive failure signs) by onsite AI while keeping cloud server and communication costs low.

For evaluating the AI chip, ROHM offers an evaluation board equipped with Arduino-compatible terminals that can be fitted with an expansion sensor board for connecting to an MCU (Arduino). Wireless communication modules (Wi-Fi and Bluetooth®) along with 64kbit EEPROM memory are mounted on the board, and by connecting units such as sensors and attaching them to the target equipment, it will be possible to verify the effects of the AI chip from a display. This evaluation board will be loaned out from ROHM sales.

发送到邮箱 |
+1 赞 0
收藏
评论 0
| 转发至：

本文由玄子转载自ROHM News，原文标题为:ROHM Develops Ultra-Low-Power On-Device Learning Edge AI Chip，本站所有转载文章系出于传递更多信息之目的，且明确注明来源，不希望被转载的媒体或个人可与我们联系，我们将立即进行删除处理。

全部评论（0）

暂无评论

【IC】芯科最新推出全球最小的新型蓝牙SoC，CSP封装小至2.3x2.6毫米，有助于简化开发流程

芯科科技近日宣布推出xG27系列蓝牙片上系统（SoC），包括用于蓝牙连接的BG27和支持Zigbee及其他专有协议的MG27，该SoC是专为极小型物联网（IoT）设备设计的新型集成电路系列产品。

新产品发布时间 : 2023-03-16

【产品】ROHM新开发数十毫瓦超低功耗的设备端学习AI芯片，无需云服务器即可实时预测故障

ROHM开发出一款设备端学习AI芯片（配备设备端学习AI加速器的SoC），该产品利用AI（人工智能）技术，能以超低功耗实时预测内置电机和传感器等的电子设备的故障（故障迹象检测），非常适用于IoT领域的边缘计算设备和端点。

新产品发布时间 : 2022-09-28

Silicon Labs Ultra-low Power SiWx917 Wi-Fi 6 SoC: Ideal for Addressing Wi-Fi IoT Networking Challenges

Silicon Labs is announcing the ultra-low power SiWx917 Wi-Fi 6 and Bluetooth LE SoC. SiWx917 was developed to radically reduce the energy consumption in Wi-Fi IoT networking while delivering more computing, faster AI/ML, and robust security to solve the challenges of the future.

新产品发布时间 : 2022-11-08

Silicon Labs（芯科科技）产品选型指南

目录- 8-bit Microcontrollers 32-bit Microcontrollers High-performance analog products wireless ICs, SoCs and modules Sensors Interface chip Timing products Isolation products Audio products Video products Voice products Data products Development Tools

型号- C8051T604-GS,SI8420AD-A-IS,WF111,SI7007-A20-IM1,SI8035AA-B-IU,C8051T604-GM,C8051F813-GU,SI8642ET-IS,C8051F541-AM,SL16020DC,C8051F813-GM,C8051F587-AM,C8051F044,SI2434,SI4613,C8051F043,SI4614,C8051F046,C8051F045,C8051F587-AQ,SI2435,C8051F047,C8051F541-AQ,TS3001,TS3002,SI2439,TS3003,SI8712CC-B-IP,TS3004,C8051F060,C8051F531-C-IM,TS3006,C8051F062,SI8710AD-B-IS,C8051F061,SI8712CC-B-IS,C8051F064,C8051F063,EFM32LG230FX - QFN64,EZR32WG230F64R69G,EFM32WG880FX - QFP100,EFM32WG980FX - QFP100,SI8233BB-C-IS1,C8051F531-C-IT,SI51210,SI8635BD-B-IS,C8051F975-A-GM,SI8230BB-B-IS1,SI8232BB-B-IS,C8051F394-A-GM,SI4624,C8051T616-GM,SI4622,SI8635BT-IS,C8051F573-IM,SI4629,C8051F530-C-AM,SI8920BD-IS,C8051F530-C-AT,C8051F801-GM,C8051F711-GQ,C8051F585-IM,SI21802,C8051F585-IQ,EFM32LG390FX - BGA112,SI8605AC-B-IS1,C8051F370-A-GM,EFM8BB10F2G-A-QFN20,C8051F066,C8051F065,C8051F067,SI2457,SI3452B-B02-GM,SI88243ED-IS,C8051F801-GU,C8051F506-AQ,SI7005-B-GM1,SI2167-C,SI21682-C,SI7022-A20-IM,SI8642AB-B-IS,SI8650ED-B-IS,SI2167-B,SI8719BC-A-IP,TSM9118,SI8631AB-B-IS,TSM9117,SI8234AB-C-IS,TSM9119,SI8234AB-C-IM,EFM32WG390FX - BGA112,SI8620AB-B-IS,SI8719BC-A-IS,SI8631EC-B-IS1,C8051F565-AM,EZR32WG330F64R55G,SI8235AB-C-IS1,C8051F835-GS,SI1132,SI3500-A-GM,TS1102-100,C8051F506-AM,C8051F565-AQ,SL2309NZ,SI8065AA-B-IU,EZR32LG230F64R60G,C8051F551-AM,EZR32LG230F128R60G,EFM32LG980FX - QFP100,SI8640AB-B-IS1,SI8600AD-B-IS,SI8710BC-B-IP,SI8710BC-B-IS,C8051F007,EFM32LG290FX - BGA112,SI8712BD-B-IM,EFM32GG295FX - BGA120,C8051F38C-GQ,SI8244BB-C-IS1,C8051F000,C8051F002,C8051F38C-GM,C8051F001,SI8620EC-B-IS,TS1005,TS1004,C8051F006,TS1003,C8051F005,TS1002,TS1001,C8051F963-B-GM,C8051F504-AQ,EFM8SB10F4G-A-QFN20,C8051F823-GS,EZR32LG230F128R61G,C8051F020,EFM32TG822FX - QFP48,SI8650BC-B-IS1,C8051F388-GQ,EFM32G200FX - QFN32,C8051F537-C-AM,C8051F563-IM,C8051F019,EZR32WG330F256R55G,C8051F018,C8051F563-IQ,C8051F011,SI2401,SI4822,C8051F010,EFM32TG222FX - QFP48,EFM32TG210FX - QFN32,SI4820,C8051F012,C8051F015,SI4826,SI2404,SI4827,C8051F017,SI4824,C8051F016,C8051F504-AM,SI4825,EFM32WG232FX - QFP64,C8051T602-GS,C8051T327-GM,C8051T602-GM,BGM111,EZR32WG330F128R68G,C8051F863-C-GS,SI88244ED-IS,C8051F811-GS,C8051F589-AM,SI8421BD-B-IS,C8051F537-C-AT,EFM32GG395FX - BGA120,C8051F531-C-AT,C8051F854-C-GM,C8051F022,C8051F021,SI4831,C8051F023,SI2415,C8051F854-C-GU,SI4835,SI2417,SI4836,EFM32WG290FX - BGA112,SI2169-C,TS1101-100,SI8455BA-B-IU,C8051F040,C8051F042,C8051F041,EZR32WG330F128R69G,SI8232AB-B-IS1,C8051F344-GQ,SI8662EC-B-IS1,SI4840,SI8452AA-B-IS1,C8051F701-GQ,SI4844,C8051F502-IM,C8051F575-IM,SI88622ED-IS,C8051F502-IQ,CP2201,CP2200,EFM32LG942FX - QFP64,SI8641BB-B-IS,EFM32GG942FX - QFP64,SI8655BB-B-IS1,SI8400AB-B-IS,SI8030AA-B-IU,C8051F587-IM,C8051F206,C8051F587-IQ,EFM32TG840FX - QFN64,SI8460AA-B-IS1,C8051F500-IQ,EZR32WG230F64R61G,C8051F546-IM,C8051F500-IM,EZR32WG330F128R67G,SI8660AB-B-IS1,C8051F827-GS,C8051F988-GU,SI52142,SI52143,SI7050,SI52144,SI7055,SI7054,SI7053,C8051F567-AQ,C8051F590-AM,C8051F567-AM,C8051F988-GM,SI8450BA-B-IS1,SI7023-A20-IM1,SL28EB742,SI8231BD-B-IS,C8051F855-C-GU,EZR32WG230F128R63G,EZR32LG230F64R68G,SI8631BC-B-IS1,C8051F300-GM,SI52111,SI52112,SI8920AD-IS,C8051F855-C-GM,SL28EB740,EFM8SB20F16G-A-QFN24,SI8641BB-B-IS1,C8051F300-GS,EFM8UB10F16G-C-QFN28,C8051F920-GQ,C8051F815-GS,C8051F920-GM,C8051F220,EFM8UB10F16G-C-QFN20,TSM9938F,EZR32LG230F64R67G,C8051F221,C8051T320-GQ,SI8651EC-B-IS1,C8051F226,C8051F862-C-GS,TSM9938H,SI8661BB-B-IS1,EZR32WG230F64R63G,SI8661ED-B-IS,SI8220DD-A-IS,EZR32LG230F64R69G,SI8405AB-A-IS1,C8051F565-IM,EFM8SB10F2G-A-QFN20,C8051F395-A-GM,C8051F565-IQ,SI8641BD-B-IS,EFM32TG232FX - QFP64,C8051F231,SI8661BC-B-IS1,C8051F230,C8051T631-GM,SL2305NZ,SI8441AA-D-IS1,SI8630BD-B-IS,C8051T600-GS,C8051F236,SI8442BA-D-IS1,C8051F580-AQ,C8051T600-GM,C8051F803-GS,C8051F580-AM,TS1103-100,EZR32WG230F128R61G,C8051F996-GM,SI21822,SI8645BT-IS,C8051F302-GM,C8051F996-GU,SI8232AB-B-IS,SI8622BC-B-IS,C8051F302-GS,TS7003,C8051F708-GQ,TS7001,EFM32WG890FX - BGA112,EZR32LG230F64R61G,EZR32WG330F128R63G,SI21662-B,SI8600AC-B-IS,SI21662-C,EZR32WG230F128R60G,EFM32G222FX - QFP48,TSM9938T,C8051F523-C-IM,SI21812,EFM32GG990FX - BGA112,EFM32LG360FX - CSP81,SL23EP09NZ,SI8651BD-B-IS,TSM9938W,C8051F510-IM,C8051F556-IM,SI8422AD-B-IS,SI8232AD-B-IS,C8051F411,C8051F410,C8051F413,SI8405AA-A-IS1,C8051F412,C8051F381-GM,SI8040AA-B-IU,C8051F589-IM,C8051F965-B-GM,SI8461BA-B-IS1,SI8655BA-B-IS,SI8660BA-B-IS1,C8051F381-GQ,C8051T621-GM,C8051F346-GM,EZR32WG230F64R68G,EFM8BB10F8G-A-QSOP24,SI51214,C8051F346-GQ,SI51211,SI51218,SI8655BA-B-IU,SI8662BD-B-IS,SI8630BB-B-IS,EFM8UB20F32G-A-QFP32,SI51219,EZR32LG230F64R63G,EZR32WG330F128R60G,C8051F930-GM,SI8630BB-B-IS1,SI8660BB-B-IS1,C8051F502-AM,C8051F825-GS,EFM8UB11F16G-C-QSOP24,C8051F544-IM,EZR32WG330F128R61G,C8051F34A-GQ,EFM32GG332FX - QFP64,SI7006-A20-IM,C8051F502-AQ,SI2165-D,C8051F544-IQ,C8051F986-GU,C8051F34A-GM,SI21832,WF121,EZR32WG230F64R67G,SI8620BT-IS,C8051F569-AM,SI8642BC-B-IS1,C8051F930-GQ,C8051F986-GM,EFM8UB20F32G-A-QFP48,EZR32LG330F64R55G,EFM32LG380FX - QFP100,SI8630

选型指南 - SILICON LABS - 2019/1/14 PDF 英文下载

研讨会2024高算力SOC/MCU新技术研讨会

描述- 9月26日直播，带来SOC，DSP，MCU，AFE，IMU等高算力及周边产品，分享在机器人，多模态感知，AI，低空无人机，自动驾驶，视觉惯性里程计，微型逆变器，光伏储能，智慧医疗，数字健康，光模块等领域应用，点击了解报名

活动发布时间 : 2024-07-04

GreenWaves Technologies Partners with Open-Silicon to develop Industry’s First IoT Processor Based on PULP and RISC-V

Open-Silicon, a system-optimized ASIC solution provider, today announced it was selected by GreenWaves Technologies to develop GAP8, the industry’s first IoT processor. GAP8 is built on the open source Parallel Ultra Low Power (PULP) and RISC-V ISA projects. Open-Silicon is providing GreenWaves Technologies with the complete RTL-to-physical design custom SoC implementation that is required to transform this smart IoT concept into working silicon in volume production.

产品发布时间 : 2024-08-20

Silicon Labs（芯科科技） Wi-Fi 芯片和模块选型指南

目录- Wi-Fi SoC and Module Selector Guide Wi-Fi Lineup Wi-Fi Development Kits IoT Wi-Fi Technology Leader Wi-Fi Applications Company Profile

型号- SLEXP8022A,SIWX915,RS9116,WF200,SIWX917,RS9116X-DB-EVK1,RS9116X-SB-EVK1,RS9116X-SB-EVK2

选型指南 - SILICON LABS - 2023/10/3 PDF 英文下载

GreenWaves Technologies Selects Dolphin Integration’s low power virtual components for the industry’s first IoT processor

Dolphin Integration, the leading Silicon IP provider for low-power SoC, announced today that it has been selected by GreenWaves Technologies to provide a consistent set of Silicon IPs for achieving the lowest power in sleep and active modes of GAP8, the industry’s first IoT processor.

厂牌及品类发布时间 : 2024-08-14

【经验】如何在Gateway Soc (internal flash)中实现 Ota Server？

EFR32MG多协议SoC芯片是Silicon Labs专门针对 IOT 推出的ZigBee方案，高达19.5dbm的发射功率，-102.7dbm的接收灵敏度，在业界内拥有最佳的RF射频性能，具有最全的Zigbee网络协议，被广泛应用在智能家居，工业控制等行业。本文介绍在开发 Zigbee Gateway 时如何在Gateway Soc (internal flash)中实现 Ota Server

设计经验发布时间 : 2019-01-03

【应用】联盛德IoT Wi-Fi/蓝牙双模SoC芯片用于智能插座，集成32位XT804处理器，支持固件加密存储

智能家居成为了一个趋势，给传统插座增添WiFi、电量计量芯片等，可以实现各种不同的功能。推荐联盛德IoT Wi-Fi/蓝牙双模SoC芯片W801作为智能插座的主控及物联设备。

应用方案发布时间 : 2023-01-07

GAP-8: A RISC-V SoC for AI at the Edge of the IoT

Current ultra-low power smart sensing edge devices, operating for years on small batteries, are limited to low-bandwidth sensors, such as temperature or pressure. Enabling the next generation of edge devices to process data from richer sensors such as image, video, audio, or multi-axial motion/vibration has huge application potential.

应用方案发布时间 : 2024-08-10