Dell AI Factory takes the spotlight at Dell Technologies World

Expanded partnerships, AI uses

The partner ecosystem continues to highlight all the ways Dell is hoping to empower its customers to embrace AI. Of course, we saw an expansion of the Dell AI Factory with Nvidia following its launch at Nvidia GTC. Qualcomm and Microsoft Azure were also highlighted in the keynote at Dell Technologies World.

The two companies that stood out the most were Hugging Face and Meta. It was fantastic seeing both on the keynote stage talking about the ways they're working with Dell. In what feels like an overwhelmingly cloudy AI world, the Dell Enterprise Hub on Hugging Face simplifies on-premises deployments of optimized open source models. For Meta, it's about taking arguably the most capable openly available Meta Llama 3 models and running them on Dell PowerEdge to support common use cases such as code generation, customer assistance, document creation and more.

Of course, Dell continues to expand its partnerships in the AI space. Seeing Nvidia on stage is all but expected at this point, but Meta and Hugging Face on stage highlights some of the investment areas Dell is making to cement itself as an AI leader. I expect that partner ecosystem to continue to expand over the rest of the year, especially as Dell leans in on the open source side.

We saw plenty of customers using Dell Technologies to support their AI initiatives. I was fortunate to chat with Kris Kristinsdottir, a quantitative research engineer at Northwestern Medicine, a reference customer Dell publicized in April.  

The healthcare services provider implemented Dell's AI infrastructure entirely in-house, across multiple hospital locations, to empower practitioners to improve workflows and disease diagnosis. One of the use cases is supporting the radiology department, where over 50 radiologists leverage a multimodal model that looks at chest x-rays and generates customized reports based off a system template. They report seeing 40% efficiency improvements and, most importantly, patients receiving results and diagnoses faster.

The fascinating part to me is how Kristinsdottir described the criticality of an on-premises infrastructure footprint. Between latency challenges and data volumes -- especially with unstructured data such as medical images -- it's about balancing time savings and accuracy.

There's been a push to reduce the time required to train a model, but in healthcare, it's more about being able to incorporate more data into model training to improve accuracy.

Apex, AI services and the elephant in the room

With any event, there are always certain products and partnerships that take a back seat. It's important to note that when that happens, it's not because it's less of a priority, but there's just too much else to cover that's new and attention grabbing. The two areas that I feel could have received more attention are the following:

1. Dell AI Factory on Apex. Dell recognizes that as-a-service offerings are attractive to customers, especially in the areas of AI. For organizations that are early on their AI journeys, the larger upfront investments to deploy and manage AI hardware on premises simply isn't feasible. Apex supports Dell AI Factory so organizations can pay for what they need and use, offering an ideal alignment of financial commitments and operational efficiency.
2. Dell's Professional Services for Generative AI. In my opinion, Dell massively underplayed the investments it has made in professional services. With so many organizations asking how to get started with AI, Dell has put its money where its mouth is by transforming its professional services organizations to strategize and identify the right use cases. It can support key areas of data management such as data preparation; align the right infrastructure offerings during proof of concept and implementation; help with AI-specific tasks such as model tuning, RAG and prompt engineering; and ensure a seamless transition into production with scalable operational management.

The other area that needed addressing was the elephant in the room with Broadcom's ownership of VMware. In January, Dell officially ended its distribution agreement with VMware.

Obviously, customers are angry and confused. What do they do? Where do they go? How long will it take?

While my colleagues covering the infrastructure and app/dev space likely have their own opinions on this, Redshift seems almost too perfect as the critical partner to answer the call. Of course, there's an excellent virtualization offload to containers angle, but the area being slept on is its AI chops. Red Hat OpenShift AI would quite nicely align to Dell AI Factory, including its openness and the ability to provide key AI software components not currently being delivered by Dell. That's a specific partnership that I'll be keeping a close eye on going forward.

While Dell has had its fits and starts in the AI world over the last several years, it always felt like there was an underlying hesitancy to fully lean in. After Dell Technologies World 2024, it's clear to me that they're leaning in and making the right investments internally and externally to fully embrace AI in the way that the company should. Now it's just a matter of the company not taking its foot off the accelerator.

Mike Leone is a principal analyst at TechTarget's Enterprise Strategy Group, where he covers data, analytics and AI.

Enterprise Strategy Group is a division of TechTarget. Its analysts have business relationships with technology vendors.