AI Everywhere – Scaling AI In The Cloud With Intel® Xeon®6

Today, the omnipresent AI that we’re starting to take for granted has become a critical tool for business. Market leaders across all sectors are using AI to enhance decision-making, automate routine workloads and optimize business processes.

All of this is made possible thanks to earlier investment in cloud infrastructure, enabling AI-powered services to be efficiently delivered to wherever they’re needed – a concept that’s become known as “AI everywhere”.

For example, AI-powered analytics algorithms can crunch through the vast mountains of data created whenever products are used or customers browse websites. They can even extract insights from messy, unstructured data such as voice recordings of customer service calls or video footage of manufacturing operations at work. This lets them improve customer experience, predict quality control issues, and reduce downtime caused by equipment failure.

However, all of this requires robust IT architecture and huge amounts of processing power. In recent years, chip makers have focused their R&D on creating silicon that can provide the scalability and performance necessary to deliver this performance.

Intel Unveils Its Xeon® 6 Processors

Recently, Intel has revealed the latest generation of its Xeon processors (please note that this article is sponsored by Intel), designed to meet the demand for increasingly powerful processors that can handle enterprise and industrial AI workloads.

Among the key features, it claims a 10x speed improvement for AI tasks, when compared to previous Xeon generations. This has been achieved by increasing the number of cores – the basic processing elements – in its chips to a staggering 144 in the 6700 model, and 288 in the 6900 series.

Each core is individually optimized for AI with the inclusion of Intel’s proprietary AI acceleration engines, known as Intel Advanced Matrix Extension (Intel AMX) and Advanced Vector Extension (Intel AVX).

It’s also improved the memory architecture, enabling access to up to 12 channels of super-fast DDR5 memory to ensure data can be transferred in and out at the high speed required for working with real-time datasets.

Two architectures were announced – the P Core, optimized for performance, and the E-core, optimized for energy efficiency.

It’s clear that Intel is positioning its Xeon® Scalable range as offering a unified solution across today’s leading business use cases – namely AI, high-performance computing, and cloud computing.

Speaking in a video accompanying the launch, Intel’s Justin Hotard says “We’re focused on enabling enterprises to unlock the value of their data and accelerate their deployments – we’re committed to bringing AI everywhere.”

The Sustainability Challenge

These days, concerns over sustainability issues around AI are raised with increasing frequency. This isn’t surprising considering that it’s been estimated that training just one AI model generates five times the amount of carbon emissions as the average car does in its lifetime!

Intel says it has addressed these concerns with the Xeon® 6 range by improving core density by 2.5x and performance-per-watt by 2.4x, compared to its fourth-generation processors still used by many organizations. This reduces the energy overheads and lowers overall carbon output, as well as reducing the cooling requirements, which also consume power. It estimates that this will result in reducing C02 emissions by 34,000 tons over a four-year period.

AI advocates argue that the power requirements of AI are offset by the fact that using it allows businesses to deliver a great many other efficiencies throughout their organization. As a simple example, a delivery company that uses AI to route its fleet of vehicles will create less emissions by optimizing the distance that vehicles will travel. And many big data center operators (Google and Microsoft, for example) are moving towards net-zero targets by switching to renewable energy supplies.

Nevertheless, the environmental impact of AI is an issue that needs to be addressed, and it’s good to see that manufacturers like Intel are treating sustainability improvements as a priority.

The Future Of AI In The Cloud

Looking ahead, its clear that the combination of enterprise AI and cloud computing will continue to be the primary drivers of business innovation for some time yet.

By adopting an AI-everywhere mindset, organizations can empower themselves to leverage emerging technologies like edge computing, computer vision, language processing and real-time analytics. This will mean more exciting products and services as well as improved efficiency throughout operations and processes.

Building these capabilities will require continuous advancements in infrastructure capabilities, while also creating a commitment to improving sustainability.

It’s no secret that Intel faces stiff competition, with the market for AI processors exploding in recent years. Nvidia, AMD and Apple have all put AI capabilities at the heart of their CPU and GPU offerings, and are all competing for a share of the fast-growing data center market.

And this is clearly the direction that Intel have decided to take, too – differentiating themselves through the high-performance capabilities and improved energy efficiency of the new Xeon® 6 Scalable range.

Xeon® 6 E models available now and P models due to hit the shelves in the next quarter. For more information, check out this link: https://intel.ly/3VmMfOY