Securing a portfolio of AI and ML optimized servers
Support for AI and ML project development environments across all industries

Cisco Korea announced on the 10th that it will help strengthen the capabilities of companies preparing for the increasing number of AI and machine learning projects by unveiling a new AI server tailored to AI and machine learning.

The adoption of AI and machine learning offers new solutions to many of the problems facing businesses today. However, it is also expected to have a significant impact on existing IT infrastructure and processes, posing new challenges. Cisco, therefore, provides IT infrastructure capable of accommodating the new traffic patterns and rapidly increasing throughput resulting from the adoption of AI and machine learning.
The new Cisco AI Server is designed to accelerate compute-intensive forms of deep learning and features NVIDIA GPUs to accelerate widely used machine learning software.

In recent years, data scientists and developers working in machine learning and deep learning environments require increasingly advanced computing capabilities, including IT architectures suited to handling massive amounts of data and tools for analyzing and utilizing that data for learning. To this end, Cisco collaborates with various solution partners to present optimized architecture modeling and provide guidance on actual introduction and deployment, thereby reducing customer deployment times and optimizing operations.

“Artificial intelligence and machine learning-based applications are expected to emerge as major trends in the enterprise sector in the coming years, supporting complex business issues while presenting new challenges for IT to overcome,” said Hwang Seung-hee, managing director of Cisco Korea’s data center and cloud division. “The new Cisco AI server will support the spread of AI across various industries with its powerful performance.”

He continued, “Customers in the financial sector who have early access to this AI server are using it to improve fraud prevention and algorithmic trading functions, while customers in the healthcare sector are using it to strengthen cognitive and diagnostic functions, improve medical image classification functions, and promote new drug development research.”

Additionally, with the addition of the new AI server (UCS C480 ML), Cisco offers a variety of computing options for each stage of AI and machine learning. This includes everything from data collection and analysis at the edge, to data processing and learning in the data center, to real-time inference at the AI core, all the way to customer delivery, supporting customers as follows:

First, benefits are provided for data scientists and developers. Thousands of customers worldwide are currently leveraging Cisco UCS for big data analytics, and this new AI server, tailored for AI and machine learning, accelerates the movement of data from the edge to the core, enabling customers to extract more information from their data and make faster, more informed decisions. Cisco is also accelerating the development of next-generation applications by providing data scientists and developers with tools and resources through its new DevNet AI Developer Center and DevNet Ecosystem Exchange.

And there's IT infrastructure support. Cisco UCS enables IT to easily add new technologies to the environment, and it also leverages Cisco Intersight, a cloud-based systems management tool, to provide convenient and accessible system management. This allows companies to build a policy-based, automated operating system for the computing infrastructure required for AI and ML software environments.

Finally, it is built on an ecosystem-based design. Cisco comprehensively supports containers and multi-cloud computing architectures that facilitate the deployment of open source software. Furthermore, Cisco is validating solutions and machine learning environments, including Anaconda, Kubeflow, Cloudera, and Hortonworks. UCS customers using Kubeflow on Kubernetes can easily deploy AI workloads to Google Kubernetes Engine, leveraging them as on-premises or cloud-based ML environments.

In addition, Cisco AI Server is expected to contribute to the development of deep learning computing systems through partnerships that significantly strengthen AI capabilities, such as supporting Kubeflow, a Google open source project, and integrating NVIDIA Tesla V100 Tensor Core GPUs and collaborating with the University of Wisconsin-Madison.

The Cisco UCS C480 ML M5 Rack Server builds on the Cisco UCS B-Series, C-Series, and HyperFlex systems portfolio and is expected to be available through Cisco partners in Q4, along with Cisco AI and ML enabled services spanning analytics, deep learning, and automation.