Inside Intel’s transition into an enterprise tech company

Intel may be synonymous with PCs and microprocessors, but that’s an image the company has been trying to shed for several years now.

Today, Intel is making a bigger push into the enterprise, from datacentres and the cloud to the edge, where a lot more data processing is being done with the advent of the internet of things (IoT).

In highlighting Intel’s transformation journey, Santhosh Viswanathan, its managing director for Asia-Pacific and Japan, said the company has invested more than $40bn over the past five to six years to acquire companies and capabilities that will help organisations unleash the potential of data.

Viswanathan said these capabilities revolve around moving data faster across networks, as well as storing and processing the data, pointing to Intel’s portfolio of storage, networking and compute products, including visual processing units that support computer vision and artificial intelligence (AI) workloads at the edge.

“There are areas of computing all around us, especially in Singapore where you can see different contactless kiosks, terminal scanners, building monitoring systems or surveillance systems, all of which are becoming compute intensive.

“That’s not just about processing data but also trying to build intelligence around that so you don’t have to send the data all the way back to the datacentre to make sense of it. We see this as an opportunity to be way bigger than what we were primarily focused on,” he said.

Broadening its portfolio is only part of Intel’s game plan. It is also working with local companies such as Singapore’s KroniKare, serving as a solution architect to help budding tech startups bring their ideas to fruition.

In the case of KroniKare, Viswanathan said Intel worked with the company to build a thermal scanning device that detects wounds and uses AI to recommend treatments based on an image database. The idea was conceived out of the need to provide timely treatment of wounds before their conditions worsen to the point of requiring amputation. 

Using an Intel RealSense 3D camera, Viswanathan said the device could identity the type of wound within 30 seconds and suggest treatment options for nurses who would otherwise have to wait for a specialist doctor to prescribe the right treatment.

KroniKare has since deployed the same technology to detect faces and ascertain if people are wearing masks in public places using Intel’s OpenVino toolkit that facilitates deployment of deep learning models and computer vision workloads on Intel hardware.

“This device didn’t exist in the market before, so we had to work with the customer to design the product and engaged manufacturers in Taiwan or Malaysia to build a prototype,” Viswanathan said.

Just as Intel is repositioning itself to be an enterprise IT company, it faces competition in areas like edge computing where public cloud suppliers such as Alibaba and Google have developed processors that specialise in crunching AI workloads.

Google’s tensor processing unit, for instance, was developed specifically for neural network machine learning using its own TensorFlow software, while Alibaba has a field programmable gate arrays-based (FGPA-based) deep learning processor.

On the competition, Viswanathan noted that many cloud suppliers are working with Intel to build their cloud infrastructure and services, while the microprocessor market has always had multiple players.

“Intel is at its best when we’re focused on how we can continue the cadence of Moore’s Law and continue to bring in innovation year after year, so that we can serve our customers best,” he said.

That includes investments in cutting edge technologies such as silicon photonics and neuromorphic computing, whereby elements of a computer are modelled after systems in the human brain and nervous system.

The National University of Singapore recently developed an electronic skin system that could give robots and prosthetic devices a sense of touch in future. The system analyses data transmitted from a network of independent sensors using Intel’s Loihi neuromorphic chip.

“We’re in the early stages of building computers that are like the human brain, but it’s a very interesting area on what the future could look like,” Viswanathan said.

Whether Intel can successfully change market perceptions of it as a chipmaker will depend on how well it applies its famed Intel Inside branding to its enterprise IT portfolio.

“We’re not a consumer specific company alone, so the level of conversation and storytelling is very different from what the world has traditionally been hearing from us,” Viswanathan said. “Our teams across Asia are really working closely with our customers in the market to bring this story to life.”