Not too long ago, Intel announced its artificial intelligence strategy in the United States. On November 30, Intel Global Vice President and President of China Yang Xu said in an interview with reporters: “Enhancing skills, doing ecology, and focusing on key AI applications are the three keys to the Intel AI strategy.â€
Deep learning chip is faster than GPU after three years
On November 30, the Intel Artificial Intelligence Forum was held in Beijing. Intel’s internal “arts†in the field of artificial intelligence were fully attended, including Naveen G.Rao, General Manager of Intel’s Artificial Intelligence Solutions, and Jason, General Manager of Intel’s Data Center Solutions Group. Wiliam Savage and General Manager of Software and Services Division Wiliam Savage.
According to Intel's published AI graph, Intel's AI "learning internal strength" dimension includes "function and experience". At present, everyone's attention to Intel AI is more focused on the "functionality" dimension. After accepting Saffron, Movidius, and Nervana, Intel has been able to increase its capabilities in AI "functions" such as machine learning/deep learning, inference systems, and machine vision.
90 days ago, Intel Corporation completed the acquisition of Nervana Systems. Nervana is an artificial intelligence ASIC chip supplier, the acquisition of Nervana, is seen as Intel's response to the AI ​​market against the AI ​​market. At present, most of the deep learning and machine learning in the market are handled by the GPU, and it is for this reason that the stock price of the NVIDIA company has risen once again. Has Nervana won to allow Intel to beat NVIDIA's GPU?
Nervana co-founder Naveen G.Rao is now the general manager of Intel's artificial intelligence solutions division. He believes that GPU is not born for AI, so it can't be better than Nervana. "Nvidia's products are mainly for image rendering, which is similar to deep learning, but not exactly the same. From the current choice, GPUs are the most suitable products that can be used to accelerate deep learning. But in fact, people can create More optimized and more suitable for deep learning products. From the architectural point of view, Intel's deep learning chip is ten times faster than GPU accelerated deep learning.â€
NaveenG.Rao said. At present, Intel has announced more details on how to integrate Nervana's technology into its existing product roadmap. Intel will test the first chip (codenamed Lake Crest) in the first half of 2017 and sell it to major customers in the second half of the year. In addition, Intel has added a new product (codenamed Knights Crest) to the artificial intelligence roadmap, which tightly integrates Nervana technology with Intel Xeon processors. From the information disclosed, Intel hopes that Nervana will shorten the time to train deep learning models to 1/100 of the GPU solution three years later.
Not too long ago, Intel announced its AI product portfolio in San Francisco. In addition to Nervana, Intel also included Xeon processors, Xeon Phi processors, and FPGAs. Intel has given such a product portfolio, because the data shows that in the data center, the current AI-related calculation accounted for about 7%, or a relatively small part, for different workloads still need different chips.
According to Bo Anna, executive vice president and general manager of the data center division at Intel, the deep learning performance of the next-generation Intel Xeon Phi processor (codenamed Knights Mill) can be increased by a factor of 4 compared to previous generation processors. Listed on the year. In addition, Intel has provided an initial version of the next-generation Xeon processor (codenamed Skylake) to a specific cloud service provider partner, which uses the Intel Advanced Vector Instruction Set AVX-512 integrated acceleration technology to significantly enhance the machine Learn the workload's reasoning performance.
Despite the roadmap, Nervana's technology integration with Intel and the mass production of chips still take time. The future course may still be fine-tuned with the market.
Promote AI ecological layout key applications
NaveenG.Rao worked at Qualcomm and later founded Nervana. Why choose to "marry" Intel instead of Qualcomm or other companies? He stated that he values ​​Intel’s capabilities in data centers and ecology. To get Nervana to release more energy, it is a better choice to join companies like Intel that have ecological capabilities and strong data center and chip manufacturing capabilities.
AC power supply is a kind of power supply that converts the input mains or DC input into a pure sine wave output after AC-DC-AC or DC-AC conversion. The ideal AC power supply is characterized by stable frequency, stable voltage, zero internal resistance and pure sine wave output voltage wave (without distortion). Mainly used for power grid simulation tests conducted by various electrical appliance manufacturers on electrical appliances according to the voltage/frequency requirements of different countries, various AC motors, AC transformers, aircraft & mechanical equipment and other electronic devices that require pure, regulated and frequency-stabilized output.
Our AC power supply mainly includes three types
1. AC Power Supplies converting the input mains AC by AC-DC-AC, output voltage and frequency are stable and adjustable with pure sine wave output waveform, so called: Variable Frequency AC Power Supplies.
2. AC Power Supplies converting the DC output of the battery and other DC equipment by DC-AC, giving an AC sinusoidal output with stable output voltage and frequency. so called: Inversion AC Power Supplies.
3. AC Power Supplies converting the input mains AC by AC-DC-AC, giving a variable voltage AC output at constant 400Hz output frequency ultra-high output frequency stability. so called: Intermediate Frequency AC Power Supplies.
4. According to the difference in the number of output phases, the AC Power Supplies can be divided into single-phase output AC power supplies and three-phase output AC power supplies.
Different from Variac and AC voltage regulators, our AC power supplies support the setting of output voltage and frequency and has superior high precision, high stability, and high efficiency stable frequency AC output, not only for AC power conversion but also for AC high precision test purpose.
Through the friendly operation panel, you can read the output data such as output voltage, output current, output power, power factor, etc., providing accurate data records for your test, and can add RS485 interfaces as standard, following the MODBUS-RTU international communication protocol, which can realize remote control and operating status monitoring of the power supplies.
Our AC Power Supplies have comprehensive protection functions, such as: over voltage, over current, over temperature and short circuit protections, which can protect the AC power supplies and DUT from damages.
AC Power Sources, AC Sources, AC - AC Power Supplies, DC - AC Power Supplies, AC Switching Power Supplies
Yangzhou IdealTek Electronics Co., Ltd. , https://www.idealtekpower.com