ICT giant Huawei on Friday demonstrated its growing leadership in innovation in Artificial Intelligence (AI) with the launch of the world’s most powerful AI processor Ascend 910 and all-scenario AI computing framework MindSpore.

Huawei’s Rotating Chairman Eric Xu described the launch as a key milestone and a new beginning. More importantly, he stressed that Huawei will work closely with partners to make “AI more pervasive and accessible, and help bring the benefits of digital technology to every person, home, and organisation.”

Huawei defines AI as a new general purpose technology, like railroads and electricity in the 19th century, and cars, computers, and the Internet in the 20th century. The company believes that AI will be used in almost every sector of the economy.

However at present, according to Xu, AI is still in its early stages of development, and there are a number of gaps to close before AI can become a true general purpose technology. “Huawei’s AI strategy is designed to bridge these gaps and speed up adoption on a global scale,” Xu added.

At the launch held in Shenzhen, China and attended by international media including the Daily FT, Huawei Chief recapped the global giant’s recent strides in AI and what is in the offing. 

He outlined ten major changes the industry needs to make to ensure AI becomes more pervasive and accessible, including changes related to technology, talent, and ecosystem development. 

“Huawei hopes that all industry players will work together to drive these changes and close the gaps between the stellar achievements already made in AI and its otherwise lukewarm adoption,” Xu said adding “At Huawei, we’ve been working hard on several of these changes.”

The 10 challenges are:

 

• Provide stronger computing power to increase the speed of complex model training from days and months to minutes – even seconds.
• Provide more affordable and abundant computing power. Right now, computing power is both costly and scarce, which limits AI development.
• Offer an all-scenario AI portfolio, meeting the different needs of businesses while ensuring that user privacy is well protected. This portfolio will allow AI to be deployed in any scenario, not just public cloud.
• Invest in basic AI algorithms. Algorithms of the future should be data-efficient, meaning they can deliver the same results with less data. They should also be energy-efficient, producing the same results with less computing power and less energy.
• Use MindSpore and ModelArts to help automate AI development, reducing reliance on human effort.
• Continue to improve model algorithms to produce industrial-grade AI that performs well in the real world, not just in tests.
• Develop a real-time, closed-loop system for model updates, making sure that enterprise AI applications continue to operate in their most optimal state.
• Maximise the value of AI by driving synergy with other technologies like cloud, IoT, edge computing, blockchain, big data, and databases.
• With a one-stop development platform of the full-stack AI portfolio, help AI become a basic skill for all application developers and ICT workers. Today only highly-skilled experts can work with AI.
• Invest more in an open AI ecosystem and build the next generation of AI talent to meet the growing demand for people with AI capabilities.
  
   

On its part, Huawei is investing money, time and effort to steer these changes. Last year October Huawei unveiled its broader AI strategy which included invest in AI research; Build a full-stack AI portfolio; cultivate talent and an open ecosystem; strengthen existing portfolio and drive operational efficiency.

It was as part of this strategy that Huawei last week launched a full-stack, all-scenario AI portfolio which covers all deployment scenarios, including public cloud, private cloud, edge computing, IoT industry devices, and consumer devices. The portfolio is also full-stack: It includes chips, chip enablement, training and inference framework, and application enablement.

Xu said Huawei’s Ascend 310 processor has already seen wide adoption in a broad range of products and cloud services. The Mobile Data Center (MDC) solutions are based on the Ascend 310 processor, and they have been used by many leading global automakers in shuttle buses, new-energy vehicles, and autonomous driving. The Ascend 310-powered Atlas series acceleration card and server are now part of dozens of industry solutions (e.g., smart transportation and smart grid) developed by dozens of AI partners.

Ascend 310 also enables Huawei Cloud services like image analysis, optical character recognition, and intelligent video analysis. There are more than 50 APIs based on this processor. At present, the number of API calls per day has exceeded 100 million, and this figure is estimated to hit 300 million by the end of 2019.

Huawei has also progressed in ModelArts which provides model development services spanning the full pipeline, from data collection and model development to model training and deployment. 

At present, developers are using ModelArts to handle more than 4,000 training tasks per day, for a total of 32,000 training hours. Among these tasks, 85% are related to visual processing, 10% are for processing audio data, and 5% are related to machine learning. Currently, more than 30,000 developers use ModelArts.

Game changer to make AI more inclusive.

Encouraged by the progress thus far, Huawei believes the launch of Ascend 910, the world’s most powerful AI processor and all-scenario AI computing framework MindSpore as a game changer to make AI more inclusive.

Huawei said test results show that the Ascend 910 processor delivers on its performance goals with much lower power consumption than originally planned. For half-precision floating point (FP16) operations, Ascend 910 delivers 256 TeraFLOPS. For integer precision calculations (INT8), it delivers 512 TeraOPS. Despite its unrivalled performance, Ascend 910’s max power consumption is only 310W, much lower than its planned specs (350W).

Ascend 910 performs much better than we expected. It is used for AI model training. In a typical training session based on ResNet-50, the combination of Ascend 910 and MindSpore trains AI models about two times faster than other mainstream training cards using TensorFlow. Ascend 910 can train 1,802 images per second, while existing training cards can only train 965 images per second.

“Moving forward, we will continue investing in AI processors that meet the needs of a broad range of scenarios. In addition to the existing Ascend 310 processor, we plan to launch the Ascend 320 processor in 2021 to support edge computing,” Xu told the launch. 

Huawei’s MDC uses Ascend 310 to support research and development efforts for autonomous driving solutions. Soon, we will launch the Ascend 610 and Ascend 620, which will support large-scale commercial solutions in this field. “The Ascend 910 launched today focuses on AI training, and in the future we will also release the Ascend 920,” he added.

Referring to MindSpore, the full-scenario AI computing framework, Xu said AI computing frameworks are critical to making AI application development easier, making AI applications more pervasive and accessible, and ensuring privacy protection.

He said MindSpore delivers on what Huawei’s three development goals for AI framework - Easy development: Dramatically reduces training time and costs; Efficient execution: Uses the least amount of resources with the highest possible OPS/W and Adaptable to all scenarios: Including device, edge, and cloud applications.

“MindSpore marks significant progress towards these goals,” stressed Xu. It is adaptable to all scenarios – across all devices, edge, and cloud environments – and provides on-demand cooperation between them. Its ‘AI Algorithm As Code’ design concept allows developers to develop advanced AI applications with ease and train their models more quickly. Through framework innovation, as well as co-optimisation of MindSpore and Ascend processors, our solution can ensure stronger performance and more efficient execution. In addition to Ascend processors, MindSpore also supports GPUs, CPUs, and other types of processors.

Xu said none of the existing frameworks can support all scenarios. “At Huawei, our business covers devices, edge, and cloud solutions. Furthermore, these days privacy protection has become more important than ever, and support for all scenarios is essential for enabling secure, pervasive AI. This is a key component in our MindSpore framework. Resource budget environments can be big or small as needed – MindSpore supports them all,” he explained. 

MindSpore helps ensure user privacy because it only deals with gradient and model information that has already been processed. It doesn’t process the data itself, so private user data can be effectively protected even in cross-scenario environments. In addition, MindSpore has built-in model protection technology to ensure that models are secure and trustworthy.

MindSpore is built on a concept called ‘AI Algorithm As Code’. This design concept allows developers to develop advanced AI applications with ease and train their models more quickly. In a typical neural network for natural language processing, MindSpore has 20% fewer lines of core code than existing frameworks on the market, and it helps developers raise their efficiency by at least 50%.

Through framework innovation, as well as co-optimisation of MindSpore and Ascend processors, Huawei solution can help developers more effectively address complex AI computing challenges and the need for a diverse range of computing power for different applications. This results in stronger performance and more efficient execution. In addition to Ascend processors, MindSpore also supports GPUs, CPUs, and other types of processors.

Xu added that Huawei will announce more groundbreaking AI product at the upcoming Huawei Connect 2019 in September in Shanghai.

Huawei achieves key new milestones in quest to empower all via AI