As 2023 comes to a close, US tech behemoth Dell Technologies is already preparing for what’s next. In its annual media briefing last Dec. 1, it unveiled the technology trends which it said will influence both the global and the Asia Pacific and Japan (APJ) regions in the coming year.
In 2024, Dell projects that organizations across the globe will move from experimenting with generative artificial intelligence (gen AI) to actually integrating it into their operations, edge platforms will emerge to streamline edge network management, and zero trust architectures will become standardized.
John Roese, Global CTO of Dell Technologies, opened the discussion on the first trend by pointing out that worldwide gen AI implementation currently does not have a transformational impact on business. He anticipated, however, a radical change over the next year.
“By 2024, we are absolutely seeing a shift. We are seeing enterprises that have been experimenting with gen AI systems start to mature to point that they will no longer be thinking of these as a future technology. 2024 will be the year we will start to see them being put into practice,” Roese said.
As companies start to implement their gen AI projects, inferencing considerations will move to the forefront, Dell saod. Inferencing is the process where a Foundation Model (FM) or Large Language Model (LLM) takes the live data or the user’s queries then applies its training in order to accomplish its given task.
The company said enterprises will have to consider the security of their inference infrastructure, the ideal placement to reduce latency, and of course, inferencing cost.
Dell predicts that the cost of operating an AI model or inferencing will be unexpectedly high as every transaction, every prompt from a user, will have a price. In fact, Dell found that every company can have around 15 million transactions per month.
So as companies become aware of the operating expense demanded by gen AI on top of training and building the models, Dell foresees that they will become more selective about the gen AI projects they will pursue.
These unplanned costs, though, do not outweigh AI’s benefits. Peter Marrs, Dell’s president for APJ, estimated that AI spend in APJ alone will have a 41% CAGR until 2030.
Moreover, he argued that APJ companies are in a strong position to deploy AI projects effectively due to their general openness to the technology, willingness to experiment with it, and support from leadership.
Dell additionally predicts that the gen AI ecosystem and supply chain will improve in 2024, which may ease concerns about building and running gen AI models. There will be more FMs and tools to develop and customize gen AI apps, manage the data to train and run them, as well as secure them.
Significantly, while the shortage of GPUs will still exist next year, more types will be available to give companies choice.
Farther into the future, Roese also counts on quantum computing to help solve the issue of gen AI and AI’s massive demand for compute power.
He stated that AI will eventually be powered by hybrid quantum systems where the AI work is spread across a set of diverse compute architectures, including quantum processing units.
Related to gen AI, the third trend revolves around edge computing. Edge computing refers to process of running servers closer to where the data is produced on edge devices in order to reduce latency and improve user experience. Edge networks will become crucial in the AI era as AI products are brought to customers.
Dell believes that as edge devices proliferate on edge networks, the most efficient way to manage this mass of devices will be to build a multi-cloud edge platform. In this way, the edge becomes an extension of the multi-cloud infrastructure and management of these edge devices will be centralized.
Dell forecasts that these edge platforms will become more common in 2024 as companies all over the world are in the process of developing such offerings.
Edge networks, however, are not the only technologies that need organization. Marrs observed that because the multi-cloud has become mainstream in APJ, it has resulted in a problem he named “cloud chaos”.
Organizations are struggling to manage their multiple cloud platforms and are calling for ways to optimize their multi-cloud systems, the company said.
The fourth and last trend revolves around security. As edge networks grow and gen AI implementation gains serious momentum, new vulnerabilities crop up and bad actors have more opportunities to steal companies’ data.
To address this threat, zero trust architecture, generally known as a security model founded on the principles of preserving strict access controls and verifying users before trusting them, is seen as the solution.
Many companies have claimed to have adopted zero trust initiatives, but Dell finds that as of now, zero trust is simply a “buzz word”. There is no standard defining at what point an organization has achieved a zero trust architecture.
In 2024, though, Dell expects to see that technology, specifications, and even certifications will emerge to standardize what qualifies as zero trust.
Roese offered Dell’s project Fort Zero as an example. Coming in 2024, it is the first commercial full zero trust private cloud system in the industry.
“2024 will be year two of the AI era. AI is the center of the universe and what we’re going to now start to think about is how do we make AI real. That means we have to be able to run it at the edge, to solve the supply chain constraints, make sure we have the proper tool chains and models available to us, to make it secure, which is a tremendous opportunity for zero trust adoption,” Roese said.
“So that’s the overarching thread. It’s all about AI, but it’s not just about gen AI in a vacuum. Everything becomes a factor in making the AI revolution happen,” he declared.
