In the data center services market, there is no doubt that Intel is the oldest, IDC believes that its market share has reached 99% of the terror. But in the latest and hottest segment of the market, artificial intelligence , Intel failed to claim the top spot. In this area, nowadays, it is just like NVIDIA, an image chip expert, whose products are very popular in deep learning neural networks and are widely used in artificial intelligence tasks such as image recognition, speech recognition, and natural language processing.
At the latest IDF16 developer conference, Intel announced its latest processor dedicated to artificial intelligence processing - the third-generation Xeon Phi , code-named "Knights Mill", and Sword, the largest manufacturer in the field, Nvidia. 
The Xeon Phi processor will be available in 2017. Intel claims to add many " floating-point " calculations to the Xeon Phi processor, which is important for machine learning algorithms. Intel Xeon processors are currently widely deployed in data centers and are also used for almost all deep learning computing tasks. However, some users also deployed auxiliary processors for artificial intelligence tasks. Most of these auxiliary processors are from Nvidia's GPU (Graphics Processing Unit). Intel said that the Xeon Phi processor lineup has more processor cores than standard Xeon processors.
Intel mentioned that Xeon Phi processors can run most of the data analysis software without the need to use an external processor that could slow down the analysis. This is a key advantage of Xeon Phi over other products. Intel executives also said that Xeon Phi can match more memory than GPU-based solutions.
Intel hopes this chip will gain a place in the fast-growing (but still niche) machine learning market. Intel stated that only 7% of all servers are used to process algorithms related to machine learning, and only 0.1% is at operating depth. Neural networks (a branch of machine learning, simulations of neurons and brain synapses to deal with unstructured data). Especially the most popular deep learning nowadays, through the latest release of the Xeon Phi processor, Intel hopes to be the next-to-last in this market.
Currently in the field of deep learning, GPUs have played a very important role. The reason why Nvidia's GPUs are so popular is because they can perform " parallel computing " - this technology can perform multiple operations at the same time. This makes it much faster than general-purpose processors when running deep learning neural networks. In the past, a large number of CPUs and supercomputers were required to perform the calculations, and now only a few GPU combinations are required to complete them. This greatly accelerates the development of deep learning and provides a basis for the further development of neural networks. People who are familiar with deep learning know that deep learning requires training . The so-called training is to find the best value among thousands of variables. This needs to be achieved through continuous attempts to achieve convergence, and the final value obtained is not an artificially determined figure, but a normal formula. Through this kind of pixel-level learning, and constantly summing up the laws, the computer can realize thinking like a human being.
Diane Bryant, Intel’s executive vice president and general manager of the data center group, mentioned in an interview that GPUs solutions will encounter problems if such computing needs are rapidly expanding. "The GPUs solution can't be expanded - this market is still in its infancy, so the current use of GPUs will complete the task, but in the future the GPU can't be further expanded. "
In addition, Intel also found Baidu to endorse their own books. Baidu announced that it will use the Xeon Phi chip to run its " Deep Speech " natural language processing service. Previously, Baidu has been using NVIDIA's GPUs to speed up its own deep learning model. Just last month, Nvidia CEO Jen-Hsun Huang chose to hold a small-scale exchange meeting with deep learning experts at Stanford University. The company announced TITAN X, the most powerful GPU card in the industry, and gave the first TITAN X graphics card to the current Baidu American Researcher Chief Scientist Andrew Ng . Now Baidu chose to announce the Xeon Phi chip to run its “Deep Speech†natural language processing service on IDF16. Baidu’s spokesperson declined to disclose information about whether Baidu’s technology will continue to be used in the future.
This shows that Intel is becoming more and more aggressive in this future war on AI (artificial intelligence). In some recently released benchmarks, Intel claims that the Xeon Phi processor is 2.3 times faster than Nvidia's GPUs.
However, Nvidia mentioned in the latest blog that Intel is using outdated software and hardware, and the results are not convincing. Nvidia claims that if Intel uses the latest technology, Nvidia will be 30% faster than Intel in training in machine learning models. Nvidia also mentioned that TITAN X GPUs based on the latest data using four Pascal architectures run at five times faster than four Xeon Phi processors .
"It is understandable that newcomers who have just entered the field may not realize that the software and hardware in this area are making progress. " Nvidia mentioned in its latest official blog.
For Intel’s fierce counterattack, Intel’s latest reply is: “ Fully understandable that Nvidia is concerned about Intel’s actions in this area. Intel’s performance data is based on the current publicly available solutions. We have confidence in the data. "Regardless of the benchmark results, Intel said that relying on the GPU cannot accomplish all the acceleration tasks. In addition, as part of Intel's move into artificial intelligence, Intel announced last week that it will acquire AI start-up company Nervana for $408 million. For follow-up related information, please keep an eye on our reports.
PS : This article was compiled by Lei Feng Network (search "Lei Feng Network" public number attention) , refused to reprint without permission!
Via Forbes Nvidia Blog, etc.