Here's another thing AI can do: Spark a boom in edge infrastructure spending

Spending on edge computing is growing fast and because it's 2024, analyst firm IDC believes AI is a big reason for the boost.

"Edge computing will play a pivotal role in the deployment of AI applications," wrote IDC analyst Dave McCarthy. "To meet the scalability and performance requirements, organizations will need to adopt the distributed approach to architecture that edge computing provides. OEMs, ISVs, and services are taking advantage of this market opportunity to enable AI in edge locations."

That need for distributed scale means that by 2027 IDC predicts global spending on edge compute will approach $350 billion – well beyond its expected $232 billion spend in 2024, which itself represents 15.4 percent growth over 2023.

If those numbers seem high, that's because IDC defines edge as encompassing 19 industries, over 500 enterprise use cases, plus hardware, software, and services.

Edge computing matters to AI because the big GPU clusters used to train big models are built for that job – not for powering AI applications. Latency means it's nearly always better to put infrastructure closer to users, and as AI applications don't need the massive power of a GPU cluster, edge infrastructure has an important role to play.

In anticipation of AI spreading from the datacenter to the edge, Cloudflare last year began deploying large quantities of GPUs across its content delivery network (CDN) to serve machine learning workloads and even large language models to end users. Meanwhile, rival CDN provider Akamai this week revealed it was working with Neural Magic to accelerate AI workloads running on its CPU-based infrastructure.

Some of the most common edge AI use cases – encompassing about 22 percent of global edge spending, according to IDC – range from multi-access edge computing (MEC), which encompasses distributed compute capabilities, to content delivery networks, and virtualized networking functionality. IDC expects this segment to see the largest investment over the next few years with a compound annual growth rate (CAGR) of 19.1 percent.

• Can AI shorten PC replacement cycles? Dell seems to think so
• European Union lawmakers line up to defend world's first AI Act
• Leaked docs hint Google may use SiFive RISC-V cores in next-gen TPUs
• From quantum AI to photonics, what OpenAI's latest hire tells us about its future

Over the next two years, IDC analyst Marcus Torchia expects that investments in edge compute will favor MEC offerings. However, longer term, the analyst's report indicates increased investments in emerging edge use cases ranging from augmented reality, AI-enabled logistics and supply chains, autonomous mining, construction, and retail product recommendation services.

These developments are expected to drive investments in edge compute over the next five years, with IDC forecasting a CAGR in the low-to-mid teens across all 19 industries tracked by the report.

In terms of geography, analysts expect most of it to be centered in North America, Western Europe, and China. However, analysts note that the Chinese, African, and Middle Eastern markets are poised to grow the fastest over the next five years. ®