Intel recently announced that it will fully deploy AI next year, and will also launch a number of dedicated AI processor chips, including the AI ​​chip codenamed "Lake Crest" and the AI ​​chip codenamed "Knights Crest" with the Xeon processor. Learning applications. As for the Xeon Phi chip, new products will also be introduced.
GPU's highly parallel computing power has been receiving a lot of attention in AI applications. However, CPU manufacturers that have a slight disadvantage in parallel computing have begun to improve CPU computing power in recent years, such as processor manufacturers Intel in the new generation of CPU processors. In particular, the number of chip cores has been increased in succession. Recently, Intel has announced that it will fully deploy AI next year, and even launch a number of dedicated AI processor chips.
Intel will launch multiple dedicated AI chips next year
With the painful experience of the previous layout of the action chip too late to catch up, Intel chose to cut into the AI ​​market early. Intel recently officially presented their blueprints and architecture for future AI development at the Intel AI Day event in San Francisco. Intel CEO Brian Krzanich said in the meeting that he will fully develop AI in the future, and predict that next-generation processors to be launched next year will also fully support AI applications, hoping to provide enterprises with another GPU than GPU applications. New choice.
Intel plans to introduce an AI chip codenamed "Lake Crest" in the first half of next year to provide customer testing for optimizing deep learning algorithms. Next, the Xeon processor will be launched with another AI chip codenamed "Knights Crest", which combines AI application acceleration mechanisms to specialize in neural network applications. Both Lake Crest and Knights Crest use technology from Intel's deep learning engine, Nervana Systems, which was acquired in August this year. In the future, the company will also be able to use Intel AI chips by combining Xeon and Nervana's AI technology. The server is used to speed up deep learning application development.
As for Intel's ultra-multi-core co-processor Xeon Phi for supercomputers and HPC servers, new products will be launched next year. Intel said it will launch a Xeon Phi processor chip codenamed "Knight Mill" next year that can perform parallel operations using multi-core processors and has a large amount of memory that can be used to train various machine learning models using in-memory operations. Intel has also added a new chip design approach to Knight Mill that can be used in deep learning applications that require a very multi-layered neural network architecture for training. Intel even claims that Knight Mill is four times faster than the existing Xeon Phi processor codenamed "Knight Landing" when performing deep learning operations.
In order to strengthen the AI ​​market layout, Intel has also frequently moved frequently to buy AI new entrepreneur Nervana and computer vision startup Movidius to accelerate the development of AI-specific hardware. Recently, it has joined forces with Google to strengthen enterprise cloud applications. There is more cooperation on AI applications, such as Google will enhance the performance of TensorFlow on Intel processors. TensorFlow is the latest machine learning technology released by Google last year, with built-in support for deep learning. Google will enable TensorFlow to be used in all Intel CPU cores to improve parallel computing, integrate high-performance libraries such as the Math Kernel library (MKL) into TensorFlow, and optimize the memory configuration and data layer operations of different network topologies.
Why we recommend iPhone 6S Battery with TI IC (Texas Instruments ) rather than ZY/ZW?
1. TI IC is original method for iPhone Battery.
2. Charging more faster, less fever heat to save battery life.
3. Accurate battery power monitorin
4. Keep stable performance and controlable temperture
5. The battery with TI IC more safe, charge more faster, lasts longer life.
iPhone 6S 0 cycle Battery with TI IC,iPhone 6S Battery TI IC,iPhone 6S 0 cycle battery TI IC
Shenzhen Aokal Technology Co., Ltd. , https://www.aokal.com