We live in an era where technology is driving the cornerstone of the entire civilization. But despite all the brilliant inventions and technological advances, the world today is more inclined towards speed and agility than ever before. We have moved from the traditional wired dial-up Internet connection to the fourth-generation wireless network. The wide distribution of optical fiber makes it possible to connect to the Internet and access data at a fast speed. Similarly, when it comes to processors and GPUs, we have moved from a traditional 8-bit 8080 microprocessor chip containing only 6000 transistors to the most advanced Octa core processor with a clock speed of up to 1.7 GHz.
The development of artificial intelligence is becoming more and more abstract and more complex. From the simple yes/no judgment in the early stage, to the later accurate recognition, specific targets can be found in complex scenes, and then later, AI that can make active decisions like AlphaGo appears, and even intelligence such as AlphaGo Zero, Can rely entirely on self-study to achieve rapid growth.
After so long of development of artificial intelligence, earth-shaking changes have taken place in the type, complexity, and amount of information processed by the network. In terms of network types, from the early AlexNet and GoogleNet to various GANs (Generative Adversarial Networks) and various deep reinforcement learning networks, their respective network structures are different. Developers often adapt to the latest networks. Encountered some troubles.
The amount of information processed is also increasing exponentially. With the increasing demand for computing power, robots that have more restrictions on the size of the processing unit actually have obstacles to upgrade to the level of intelligence. This is why artificial intelligence chips continue to upgrade and iterate.
The ultimate goal of artificial intelligence is to simulate the human brain. The human brain has about 100 billion neurons and 1,000 trillion synapses, capable of processing complex vision, hearing, smell, taste, language ability, understanding ability, cognitive ability, and emotion. Control, complex human body mechanism control, complex psychological and physiological control, and the power consumption is only 10-20 watts.
Many people may ask why in the field of artificial intelligence, NVidia GPU has an unshakable dominance, why AMD's GPU and NVidia GPU have similar performance, but the popularity in the field of artificial intelligence is worlds apart .
In 2011, Wu Enda, who is in charge of Google Brain, learned to recognize cats within a week by training pictures with deep neural networks. He used 12 GPUs instead of 2000 CPUs. This is the first time in the world that a machine recognizes cats.
In 2016, the robot AlphaGo developed by Google's Deepmind team defeated the world champion professional nine-dan chess player Li Shishi 4-1 (AlphaGo's neural network training uses 50 GPUs, and the chess network uses 174 GPUs), which sparked the Go world. The uproar of the game, because Go has always been regarded as the pinnacle of human intelligence, which can be regarded as another major milestone in the history of artificial intelligence.
Google is not the only company designing chips for AI tasks on such devices. ARM, Qualcomm, Mediatek, and other companies have all made their own AI accelerators, and GPUs made by Nvidia dominate the training algorithm market.
However, Google's competitors do not control the entire AI stack. Customers can store their data in Google's cloud; use TPU to train their algorithms; and then use the new Edge TPU for on-device inference. Moreover, they are likely to use TensorFlow to create their machine learning software-TensorFlow is a coding framework created and operated by Google.
This vertical integration has obvious benefits. Google can ensure that all these different parts communicate with each other as efficiently and smoothly as possible, making it easier for customers to play games in the company's ecosystem.
At the Google I/O conference in May 2016, Google announced its self-designed TPU for the first time. At the Google I/O conference in 2017, Google announced the official launch of the second-generation TPU processor. At this year’s Google I/0 2018 conference, Google released a new generation of TPU processors-TPU 3.0. The performance of TPU 3.0 is 8 times higher than the current TPU 2.0, reaching 1 billion billion times.
The full name of TPU is Tensor Processing Unit, which is a neural network training processor developed by Google, which is mainly used for deep learning and AI operations. At the Next Cloud Conference in July, Google released the Edge TPU chip to attack the edge computing market. Although they are all TPUs, the version used for edge computing is different from the Cloud TPU for training machine learning. It is a microchip specially used to process the AI ​​prediction part. Edge TPU can run calculations on its own, and does not need to be connected to multiple powerful computers, so applications can work faster and more reliably. They can co-process AI work with standard chips or microcontrollers in sensors or gateway devices.
Explosion Proof Quick Plug Connector
BX3 series of explosion-proof socket connector is applicable to the petroleum, machinery, chemical, mining, drilling, wharf, construction, can provide the power for the electric motor, electric welding machine, pump, compressor, outdoor lighting equipment.
Product design,compact structure, easy installation, high temperature, shock, anti-aging, has good insulating properties and mechanical properties, products built with trapezoidal hole rubber seal and clamp device can be used for different cable diameters, not afraid of the rain, in oil exploration, drilling, well field standardized electrical lines have been widely applied. Products meet the national standard GB3836-201 "electrical apparatus fo explosive gas atmosphere", after the inspection of the national explosion-proof electrical quality inspection, certificate of proof, explosion-proof marks CNEx19.1354X, EXnA II CT4Gc.
The product consists of two parts, plugs and sockets, which plug is removable (YT), there are three forms of socket options: 1,fixed (GZ), 2, straight-through (YZ), 3, hanging (GYZ). Connection plug and socket bayonet quick connect, contact with the wire ends to tighten the screws. Shell is made of PA66 insulation corrosion material, silver-plated contacts, access to reliable, easy wiring, sealing and insulation, waterproof, dustproof, shockproof and so on.
Main Technical Data
Operating temperature: -20 ℃ 60 ℃
Rated voltage: 0 - 500V Frequency: 50Hz - 60Hz
Rated current: 1OA/16A/25A/32A/40A/60A/63A/1OOA/125A/150A/200A/250A/300A/400A/500A/600A/630A
Withstand voltage: 1800VAC
Insulation resistance: > 100MO
Contact resistance :lower than 0.5 MO
Temperature rise: lower than 65K
Mechanical life: 500 times
Contact Poles: 3P/3P+E/3P+N+E
Housing protection: IP67
Recommend to use cables with suitabe specifications:
Product Selection Parameters:
Explosion Proof Plug Socket Connectors
Ningbo Bond Industrial Electric Co., Ltd. , https://www.bondelectro.com