Building Smarter Solutions: Exploring In-House AI Development and Infrastructure

Building Smarter Solutions: Exploring In-House AI Development and Infrastructure

Join HP, AMD, and industry experts as we dive into designing scalable AI infrastructures, leveraging open-source tools and frameworks to reduce costs, and navigating the challenges of data collection, storage, and processing.

Building Smarter Solutions: Exploring In-House AI Development and Infrastructure

Page 1

To deliver on the pressing need for AI-supported applications, the Federal government needs more performance than what traditional PCs can deliver. For complex workloads, including AI-powered operations, workstations are the answer.

“Everything in the Federal space today falls into the realm of data science,” including emerging AI-driven applications, said James Olguin, AI and Data Science Business Development Manager at HP. In this environment, agencies increasingly need to keep computing on-premises — to maintain security, lower latency, and trim costs.

Workstations check all those boxes, enabling power users to meet missions faster, more securely, and cost-effectively.

Z Workstations bring artificial intelligence to life “Soon, AI is going to be everywhere. Everyone wants these cool, smart agents that people can talk to,” Olguin said. But questions linger around how to bring AI to life. “We can use these algorithms, but how do we put this together and deploy it?”

With sophisticated chips, AI-optimized workstations offer a way to deliver AI effectively. “We now integrate a neural processing unit, or NPU1, with up to 50 TOPS of AI inference performance,” said David Diederichs, Product Marketing Manager at AMD. Inference refers to running data through an AI model to make a prediction and is key to getting effective outcomes.

With Independent Software Vendor (ISV) certifications, an AI-ready workstation can support many applications that utilize AI within their workflows. “You can take advantage of these built-in features to reliably run these applications and integrate AI into your workflow,” Diederichs said.

How Z by HP Workstations powered by AMD meet key Federal needs

Page 2

Hardware-enforced security built-in with every Z Workstation The drive toward AI raises serious new concerns around cybersecurity. For example, AI models need to get fine-tuned on vast amounts of data, and many agencies are reluctant to trust that data to the cloud out of concerns around security and privacy.

For those that handle sensitive information like health or financial information, “no one wants their data up in someone else’s server,” Olguin said.

Until recently, that training data “essentially had to exist in the cloud,” he said. With the performance of the latest generation of Z Workstations, the equation changes. “Now you’re doing that work on-premises. It’s secure. It never leaves the area.”

Unlock new workflows previously not possible In addition to fine-tuning the models, IT professionals must develop applications that deliver AI-powered experiences to the end user. Latency impedes these efforts.

When developers build applications on workstations, “they are not reliant on an internet connection. You can do the development right on this machine,” Diederichs said. That virtually eliminates the latency, driving new levels of efficiency.

When it comes to cost, iterating in the cloud gets expensive fast. With a workstation, “you own the device,” Olguin said. “You can iterate a thousand times, and it does not cost you more.”

Z by HP and AMD are supporting complex workflows in the government Z by HP AI Workstations with AMD processors2 are engineered for complex workflows.

The HP Z Book Ultra is a sleek, battery-efficient mobile workstation that enables users to unleash groundbreaking performance to take on complex AI workflows. In addition to liberating Federal data, users can simultaneously render and visualize graphics-intensive projects while working with local LLMs.

The HP Z2 Mini G1A likewise delivers game-changing capability in a compact form factor. Power users can tackle massive new workflows on a mini workstation.

Both these devices are advanced by the AMD Ryzen™️ AI Max PRO Series processor. With up to 16 desktop-class CPU cores, discrete-like integrated graphics, and 128GB unified memory, AMD makes it possible to tackle workflows in ways that were previously unobtainable on a laptop.

With the power of Z by HP and AMD, “you get high-level performance in a more compact form factor,” Diederichs said. “Together, we are redefining what’s possible in compact workstations.”

Speak with a government expert to learn more.

1 Features and software that require a NPU may require software purchase, subscription or enablement by a software or platform provider, and third party software may have specific configuration or compatibility requirements. Potential NPU inferencing performance varies by use, configuration, and other factors.

2 Multi-core is designed to improve performance of certain software products. Not all customers or software applications will necessarily benefit from use of this technology. Performance and clock frequency will vary depending on application workload and your hardware and software configurations. AMD’s numbering is not a measurement of clock speed.

©️ Copyright 2025 HP Development Company, L.P. The information contained herein is subject to change without notice. The only warranties for HP products and services are set forth in the express warranty statements accompanying such products and services. Nothing herein should be construed as constituting an additional warranty. HP shall not be liable for technical or editorial errors or omissions contained herein.


Item Type: pdf