2024 has been a grand slam year for technology.
AI is being discussed in every board room and is much more than just an IT budget conversation. It’s still early innings, but the industry is unlocking new levels of productivity and capabilities every day. It’s unlike anything I’ve seen in my four decades in technology.
As we step up to the plate in 2025, I expect these five trends to lead the way:
1. AI will deliver meaningful ROI to the enterprise.
According to a recent McKinsey study, 65% of respondents are regularly using GenAI, nearly double from the previous ten months. Yet, many customers still don’t know where to start. In a recent survey of 1,600 Dell customers, we found nearly 80% are only at stage 0 (test and learn) or stage 1 (focused implementation with dedicated resources and roadmap to scale) with the ultimate goal being a fully integrated, enterprise-wide deployment with customized GenAI solutions.
For many, the test and learn phase is starting to pay off. A recent Morgan Stanley survey indicates that most companies are seeing ROI at or above expectation for their GenAI solutions, with ~40% exceeding expectations. Over the next year, enterprises will continue to see real ROI as they scale AI.
2025 is the year where enterprises will be hyper-focused on building and buying the platforms and tools they need to become the most competitive in their industry. While GenAI tools are still maturing, organizations who don’t figure out the right AI strategy and architecture will be at a disadvantage.
2. The AI PC will be the most important part of your distributed architecture.
In 2025, data will become increasingly distributed, which will cause a shift in where businesses apply AI to follow the data. This means moving beyond the data center and cloud to the edge and the PC. With AI PCs, data will be processed directly on the device, making it faster, more cost-efficient and more secure than relying on the cloud.
NPUs will be in nearly everything next year, meaning PCs will be capable of running AI workloads or apps locally on the device. Next year will bring a big PC refresh, with AI PCs, why wouldn’t you upgrade?
AI tools and applications on PCs will support all your daily tasks. Imagine a PC that evolves with you, anticipating your needs and boosting your productivity. A PC that can summarize your most important emails, action items or the five things you need to do on your first day back from vacation.
Between higher performing CPUs, GPUs and NPUs and more PC silicon options in the market than ever before, the choice and innovation will be the best it’s ever been.
3. Modern data center is driving an architectural revolution.
AI has experienced the fastest adoption curve of any technology in history. With an over 20% increase in productivity, we’ve never seen such a dramatic boost. By the end of 2026, AI adoption is expected to reach new heights with over half of data center demand coming from AI workloads¹.
The shift from training to inferencing is coming as inferencing is set to account for up to 90% of AI workloads by the end of the decade. Unlike training, inferencing is about maximizing where to run the workload relative to quality, cost, data, security and latency. With the shift, AI is moving towards disaggregated architectures where compute, storage and networks are lightning fast and can be scaled independently. Customers can reduce costs, breakdown silos, and prevent vendor lock-in.
Recent data highlights common pain points in the modern data stack including knowledge silos, latency issues, staffing challenges, and a lack of holistic data governance. These challenges impact organizations’ ability to deploy effective AI solutions, everything from shipping delays to producing inaccurate inventory insights. As a result, the current data center and IT operating model aren’t competitive enough.
4. Data center insights and efficiencies will power customer decisions.
The rising cost of energy combined with the energy demands and environmental impact of certain AI workloads continues to be tough to navigate. And at the same time, regulatory requirements are becoming more rigorous worldwide requiring new levels of disclosure.
For the data center, this includes focusing on energy-efficient hardware innovation, responsible retirement of older or obsolete equipment, and diversified energy sources to minimize impact and maximize returns. Liquid cooling innovation is a perfect example. In our liquid cooled compute solutions, you’ll see optimized cold plates and leak detection technology for unmatched reliability and efficiency.
It’s also essential to invest in workload management and monitoring tools to identify new efficiencies and optimize performance. Analyzing this data will help organizations right-size AI solutions for desired performance and to better serve their needs. A one-size-fits all approach won’t work.
5. AI agents will be everywhere — we will all be managers.
AI agents are software systems that can plan, make decisions and take action to achieve pre-defined goals. They will be everywhere next year and will have a big impact on how we use AI. By 2028 Gartner predicts that one-third of GenAI interactions will use AI agents and action models for task completion².
Next year, agents will evolve from passive applications requiring human input to multiagent systems that operate independently. Humans will soon have a whole team of agents, making everyone a manager. This will lead to mass customization as work experiences and intelligent applications become more tailored.
Current use case POCs in customer service, marketing and software development will expand into vertical solutions and even generalist AI agents. At Dell, we’re experimenting with agents to help teams sort through content and code artifacts and other knowledge bases and sharing our early adoption key learnings with our customers.
As a tech optimist, I’m confident AI will accelerate human progress and innovation. I’m ready to take the field – hope you’ll join me!
Interested in hearing other Dell leaders speak about their 2025 predictions? Click here to hear Global Chief Technology Officer and Chief AI Officer John Roese share what he expects to dominate the industry next year.
SPECIAL NOTE ON FORWARD-LOOKING STATEMENTS
Statements in this presentation that relate to future developments and events are forward-looking statements and are based on Dell Technologies’ current expectations. In some cases, you can identify these statements by such forward-looking words as “anticipate,” “believe,” “could,” “estimate,” “expect,” “intend,” “confidence,” “may,” “plan,” “potential,” “should,” “will” and “would,” or similar expressions. Actual developments and events in future periods may differ materially from those expressed or implied by these forward-looking statements because of a number of risks, uncertainties and other factors, including those discussed in Dell Technologies’ periodic reports filed with the Securities and Exchange Commission. Dell Technologies assumes no obligation to update its forward-looking statements.
1 https://investors.delltechnologies.com/static-files/a0d509af-2546-4875-8632-cfc38c6f7315#:~:text=In%20the%20last%2012%20months,center%20demand%20will%20be%20AI.
2 https://www.gartner.com/en/newsroom/press-releases/2024-03-11-gartner-predicts-one-third-of-interactions-with-genai-services-will-use-action-models-and-autonomous-agents-for-task-completion-by-2028
