AI is already transforming the way employees work, as they leverage technologies to streamline workflows, understand business data and accelerate mundane but time-consuming tasks.
Yet this is only the beginning. As more AI workloads shift to running on the endpoint rather than the cloud, there’s potential for even greater efficiencies.
In its early stages, this shift is being driven in applications where latency, performance and security are core concerns, or in industries and functions where data privacy and compliance are central.
For example, many software engineers use generative AI (genAI) to speed up their coding workflows or are spending more working hours testing and fine-tuning AI models.
Running these AI workloads in the cloud isn’t as efficient as running them locally on compute engines built into the devices they actually use to code.
Similarly, in healthcare or financial services, professionals are looking to provide rapid diagnosis, fraud protection or personalized plans and forecasts in the field. Again, it’s more practical to run these workloads closer to the data without moving sensitive information to the cloud.
In time, local AI capabilities will become mainstream, as more organizations fine-tune large language models for their specific industry or corporate requirements and deliver them to users to support decision making.
Large, resource intensive industrial models are already being reduced through quantization and distillation for local inferencing on AI PCs.
The next wave of the technology will see AI agents providing more sophisticated, personalized assistance, translating complex objectives into autonomously managed tasks.
Giant leap
These agents will be more proactive, learning from experience and adapting to meet individual needs. The result could be a giant leap in productivity, particularly if these agents run on employee PCs, where they can be closely integrated across a range of applications.
“If you think about the evolution of AI workloads, they are going rapidly from a lot of experimentation, training and inferencing on the cloud or small pilots into, ‘okay, how do I scale up into full enterprise production environments”, explains Gaston Sandoval, Corporate VP, PC Marketing for AMD. “When you start to think about that, you’ve got to start thinking about where’s your data? Can I really increase accuracy for the outputs I’m getting from the AI inferencing capabilities?”
What’s more, Sandoval points out, there are cost implications. As businesses run inferencing workloads at high frequencies, the costs of doing so in the cloud can spiral. Running these workloads locally could mean substantial savings over the long term.
AI PCs based on AMD Ryzen™ AI PRO processors can drive this shift. High-performance neural processing units (NPUs) handle inferencing tasks with higher power efficiency, while freeing central processing unit (CPU) and graphics processing unit (GPU) resources. [1]
Powerful GPUs can run more demanding fine-tuning tasks, with high-bandwidth access to large, unified pools of system RAM.
Meanwhile, a commitment to open frameworks, such as Microsoft’s Windows ML, enables businesses to take advantage of a wider range of AI applications, and developers to code once and support a range of different hardware, from workstations to lightweight Copilot+ PCs.
“Workloads are rapidly becoming more specialized,” adds Sandoval, “and when you start to see that happen, then stepping back and understanding the end-to-end infrastructure you need in an enterprise becomes super-important.”
He sees the current PC refresh cycle as an “an opportunity to optimize from cloud to client” so that businesses can re-engineer their AI workflows based on proximity to data resources, user requirements and business process needs. The productivity gains are there for the taking. AI PCs give businesses the capabilities they need.
[1] Ryzen™ AI is defined as the combination of a dedicated AI engine, AMD Radeon™ graphics engine, and Ryzen processor cores that enable AI capabilities. OEM and ISV enablement is required, and certain AI features may not yet be optimized for Ryzen AI processors. Ryzen AI is compatible with: (a) AMD Ryzen 7040 and 8040 Series processors and Ryzen PRO 7040/8040 Series processors except Ryzen 5 7540U, Ryzen 5 8540U, Ryzen 3 7440U, and Ryzen 3 8440U processors; (b) AMD Ryzen AI 300 Series processors and AMD Ryzen AI PRO 300 Series processors; (c) all AMD Ryzen 8000G Series desktop processors except the Ryzen 5 8500G/GE and Ryzen 3 8300G/GE; (d) AMD Ryzen 200 Series processors and Ryzen PRO 200 Series processors except Ryzen 5 220 and Ryzen 3 210; and (e) AMD Ryzen AI Max Series processors and Ryzen AI PRO Max Series processors. Please check with your system manufacturer for feature availability prior to purchase. GD-220e.