Dell Technologies has introduced two solutions designed to help organizations embrace artificial intelligence more speedily and efficiently.
The new offerings, which will reach market over the next month, combine Dell EMC hardware with VMware software in pre-integrated reference architectures that businesses of any size can easily adopt and deploy, according to Ravi Pendekanti, senior vice president for Dell Technologies server solutions product management and marketing.
“We want to turn it so that all of the customers, developers, and applications can actually access AI and machine learning resources when they need to,” he says.
The first of the new solutions, called Dell EMC Ready Solutions for AI: GPU-as-a-Service, allows businesses to virtualize, pool, and then dynamically allocate GPU resources scattered across multiple, often severely underutilized, workstations.
“You want to be able to guarantee that you can keep uptime and maintain uptime for your GPU services, and that’s very difficult when you have a disaggregated distributed set of GPU resources,” says Paul Turner, vice president of vSphere product management at VMware.
Much as network-attached storage technology and server virtualization once solved similar problems for disk space and processing, he continues, the new Dell solution helps organizations reduce wasted GPU capacity.
“You get all of the benefits of sharing of resources, increased efficiency, and optimal utilization,” Turner says.
The second new offering, named Dell EMC Ready Solutions for Virtualized HPC (vHPC), performs a similar function for high-performance computing, allowing users to centrally provision HPC resources for AI applications in areas like computational chemistry, bioinformatics, and computer-aided engineering rapidly.
According to Dell, in fact, citing data from Forrester Consulting, Ready Solutions for vHPC can accelerate AI model development by up to 18 times and speed hardware configuration and integration up to 20%, resulting in a 111% return on investment.
Both new Dell EMC Ready Solutions utilize Dell PowerEdge servers running VMware Cloud Foundation. That platform now includes vSphere version 7, which itself includes technology that VMware acquired last August through its purchase of BitFusion, a maker of virtualization systems for hardware accelerated devices such as GPUs.
The PowerEdge servers included in both reference architectures are also equipped with Dell EMC OpenManage systems management software featuring new integration with VMware vCenter and vSphere Lifecycle Manager.
“It essentially means that any of our customers deploying the solutions we just talked about has a way to constantly monitor and manage their PowerEdge servers in a seamless fashion and take some proactive decisions that can optimize the performance and utilization of their PowerEdge servers,” Pendekanti says.
Customers who wish to pay for either new solution on an as-a-service basis can buy them via Dell Technologies on Demand, the consumption-based purchasing program that Dell introduced last November.
IDC expects global spending on artificial intelligence to climb at a 28.4% CAGR through 2023 to $97.9 billion. Both the GPU-as-a-service solution and the Virtualized HPC solution aim to provide partners new to AI a quicker route into that fast-growing market.
“We just want to make it easier for general partners to deliver not just an independent set of software and hardware, but actually something that is already tested, qualified to work together, making it easier for them to sell complete solutions to customers,” Turner says.
Partners can use the new offerings as a foundation for building custom or resellable AI solutions as well, Pendekanti adds. “We do expect some of our channel partners to go out and add their own IP on top of what we are providing.”
Dell’s latest solutions arrive, Pendekanti says, at a time when data is multiplying exponentially and interest in AI is ascending swiftly, yet only 14.6% of businesses have AI capabilities in widespread production, according to statistics cited by Forbes.
“So here is a perfect storm,” Pendekanti observes. “You have lots of data being created and a lot of applications that have the intent of using artificial intelligence, but less than 15% are in a mode to actually utilize this, which tells us that there is a huge untapped need for actually providing solutions that our customers can put to use quickly to ensure that they get real-time analytics to make decisions that could impact their own business requirements on a day-to-day basis.”
Interest in AI is as strong among SMBs as it is in the enterprise. Indeed, 88% of SMBs have a high level of interest in AI adoption, according to recent data from Zix and its AppRiver business unit. That figure rises all the way to 99% for larger SMBs with 150 to 250 employees.
