The security tool performance problem: How to make sense of the impact

New AI-powered PCs and devices promise to accelerate employee productivity by harnessing neural processing units (NPUs) to run AI workloads locally, enhance threat protection, reduce latency, and enable new user experiences.

However, this transformation of the endpoint landscape introduces a new challenge: balancing performance, security, and employee experience. AI workloads demand significant compute resources, and important security tools can further strain system performance if not optimized for modern architectures.

Your organization faces a critical question: How do you ensure robust security without sacrificing the performance gains that AI PCs promise? The answer lies in data-driven benchmarking.

Why AI makes your security tools matter more than ever

Security tools have always been resource-intensive, but the stakes are higher in the AI era. Traditional endpoint protection solutions often rely heavily on CPU cycles, impacting responsiveness and battery life. AI PCs increase this complexity by distributing workloads across devices’ CPU, GPU, and NPU. If security tools fail to leverage NPUs efficiently, they can become a bottleneck – draining battery life and degrading user experience.

As SHI’s Field Chief Technology Officer (CTO) Dave Gruver notes, “Security tools are brutal on battery life. They’re also brutal on user experience. If I can move some of this workload over to a more efficient processing unit, the user’s happier.”

What impacts do security tools and battery life have on device performance? Field CTO Dave Gruver, Sr. Director – End User Compute Adam Reiser, and Sr. Account Executive Jim Joseph discuss.

Introducing BenchSmart: Purpose-built for assessing AI workloads

To address these challenges, we’ve developed BenchSmart, a benchmarking tool designed specifically for AI devices and AI-driven use cases. Unlike traditional benchmarks that provide a single composite score, BenchSmart delivers granular, actionable insights into how devices perform under real-world conditions.

Here are just a few ways BenchSmart stands out as a premiere benchmarking tool for the AI era:

• AI-specific metrics: Measures tokens per second, CPU/GPU/NPU utilization, and workload distribution.
• Battery performance analysis: Evaluates runtime under AI and security load, heat dissipation, and throttling behavior.
• Security tool impact: Quantifies how different security solutions affect performance and battery life, especially when optimized for NPUs.

This approach helps you make data-driven decisions about endpoint configurations, ensuring security and performance coexist without compromise.

Did you know: Your power source impacts device performance

One of the most eye-opening findings from early testing is the performance gap between plugged-in devices and those running on their battery. Some devices maintain consistent performance regardless of power source, while others experience significant degradation when unplugged.

For mobile workers such as executives, sales teams, and field staff, battery life often outweighs raw compute power. If security tools exacerbate battery drain, the business impact can be severe.

BenchSmart helps identify which device and security tool combinations deliver the best balance of performance and endurance, ensuring AI capabilities remain usable throughout the workday.

With AI PCs, security tools can avoid performance bottlenecks

Security solutions are essential, but can often be a drag on performance. AI-powered devices allow security vendors to offload certain tasks to their NPUs, which consume less power than CPUs or GPUs. When security tools are optimized for this architecture, the benefits are twofold:

1. Improved battery life: Lower power draw extends runtime.
2. Enhanced user experience: Reduced CPU contention means fewer slowdowns and interruptions.

BenchSmart measures these differences, helping you select security tools that maximize efficiency without compromising protection.

Why it matters for IT leaders

Without validatable data, you’re left guessing with AI-powered devices will best handle your AI workloads and endpoint security tools. With BenchSmart, you unlock more than performance measurements. You gain actionable outcomes. And by pairing BenchSmart with our Next-Gen Device Lab, you can:

• Test images and security stacks specific to your organization.
• Identify configuration issues that impact performance.
• Optimize security workloads across CPU, GPU, and NPU for maximum efficiency.

This collaborative approach ensures you don’t just collect data – you use it to fine-tune your endpoint security strategy.

Every year, enterprises invest millions in endpoint refresh cycles. With AI PCs and advanced security tools in the mix, these decisions carry even greater weight. BenchSmart provides the evidence IT leaders need to:

• Justify hardware investments with measurable ROI.
• Align device selection with business priorities (e.g., mobility, security, AI adoption).
• Avoid costly missteps caused by assumptions or vendor bias.

NEXT STEPS

Make the most of your next device refresh with real-world data and testing. Contact our experts to experience BenchSmart for yourself.