Tag: #genAI

Source: www.sdxcentral.com.

Predicting the future of anything is easier than you think. You look at the trends occurring now and figure out a path through those trends for a specific period.

Cloud computing is no different. You just have to ask, what are the issues that we’ll be dealing with?  What technology will we be focusing on?  How will this likely play out for enterprises?

To answer these questions, we need to focus on three major trends that are occurring in 2023 and will likely be a factor in 2024.  They are the following:

1. The move to ubiquitous and heterogeneous computing
2. The rise of generative artificial intelligence (genAI) supporting infrastructure
3. The continued need for cloud computing skills

The move to ubiquitous and heterogeneous computing

This is a significant trend that will likely continue through 2024 and beyond. The idea is that, while we’ve focused on moving to new and combined platforms, namely the public cloud, the trend is now to consider all platforms and to put the workloads and the data on platforms that make the most sense.

This began in the cloud world through the movement to multicloud deployments. The idea was that just using a single cloud provider limited the capabilities or cloud services you can access. Thus, adding more public clouds to the mix of systems you can leverage for any business purpose means you have more capabilities, including the higher likelihood that you’ll be leveraging best-of-breed solutions.

However, the movement to ubiquitous computing is more than just multicloud. Indeed, it’s about leveraging anything that makes sense, from mobile computing platforms, such as your smartphone or smartwatch, to traditional computing platforms, to the ones you usually find in enterprise data centers, to all kinds of cloud platforms — including those offered by major cloud providers, or “micro clouds,” that may only provide industry-specific services (e.g., industry clouds) supporting specific industries such as finance, healthcare, manufacturing, etc.

There are a few significant factors driving this trend. First, the hardware cost, including storage and computing, has fallen significantly in the last five years. This means that, in many cases, it can be more cost-effective. Second, cloud prices can get quite high, exacerbated by enterprises that aren’t optimizing these platforms. This can easily become a self-inflicted wound that makes the cost of cloud computing much too high if not addressed.

These trends will materialize in a few critical enterprise behaviors in 2024, including repatriating some cloud applications and data back to more traditional data centers. They will, however, lose the benefits of a public cloud, such as colocation with several valuable services like AI, serverless computing, state-of-the-art security, managed container orchestration, etc.

Also, there will be a focus on non-cloud and non-data center platforms, such as mobile computing and edge computing. The reduced hardware prices and greater availability will drive this and the rise of ubiquitous high-speed networking, such as 5G. This phenomenon makes access more cost effective and valuable. Thus, we’ll be able to leverage these resources wherever they exist. Or, better put, we will see a rise in ubiquitous access and resources within and outside of public cloud platforms.

The increase of genAI supporting infrastructure

This trend is a reaction to the rise of genAI, both in the cloud and not on the cloud. As enterprises see the value in using this technology and move from proof of concept to production systems, they will spend 2024 building the infrastructure to support genAI systems — both net-new and newly AI-enabled applications — both on the public clouds and not on the public clouds, including data centers, edge computing and mobile computing.

Much needs to be done to support this rise, though, including sizing systems for genAI. This means typically supporting training data with more database storage and keeping the storage of both structured and unstructured data. In many instances, enterprises will look to combine that data before it becomes training data, and that will likely work in the public clouds, given the ease of provisioning and scaling of cloud-based platforms.

Also, additional and sometimes different computing needs to be found. GPUs (graphics processing units) are more critical in genAI; thus, most are found and allocated. However, more traditional CPUs also play a role, and using specialized cloud services may be the easiest path to genAI for many enterprises.

As a result, cloud providers will see explosive growth around this trend in 2024. Traditional computing and storage platforms will also see changes, as many enterprises may opt to keep their productive AI systems on-premises for security and cost-efficiency reasons. At the same time, it will be a rising tide that raises all ships in 2024.

The continued demand for cloud computing skills

This is an old issue that we’ll see become more significant in 2024. This is due to the previous trends we mentioned and the expansion of the need for qualified people to design and build these systems.

Indeed, if you consider the way cloud has grown exponentially over the past several years, it’s not only been due to the capability of the cloud services; it’s also been the result of going too fast and not having the skillsets in place to make sure that the cloud deployment is cost-optimized to return the maximum value to the business. How do we address this?

Colleges and universities can help develop these skills in their graduates. Also, enterprises need to develop internal training programs to support the sharpening of their current employees’ skills.

In 2024, enterprises will likely become cleverer in this regard, such as focusing on skills-first hiring and supporting programs that attract specific groups of people to the profession — like those returning to the workplace after raising a family, veterans returning to civilian work, and individuals without a four-year college degree. They may also look to support internal upskilling through monetary incentives.

So, will all of this likely happen in 2024? We’ll have to stay tuned to find out.