At the moment, everyone is talking about and anticipating the rise of Artificial Intelligence (AI). The phrase “AI” encompasses multiple capabilities and concepts, but generally, AI refers to the shift in technology where machines are capable of taking on tasks that could typically be performed by a person. Computer systems can now perform decision-making tasks, such as speech recognition, quickly and efficiently, and it's changing the way we work.
AI is becoming a creative disrupter
Chatbots like ChatGPT have accelerated this movement towards AI — we went from being able to teach the computer to perform simple functions for us to having the computer perform the intensive research for us. Now a creative disrupter, AI has grown so much in development that it has changed the game regarding the level of autonomous work needed by a human.
There is excitement building about what our next generation of technology will look like as AI advances. Many technology publications and reviewers predict that AI features will soon be embedded into future laptops and PCs, further simplifying and automating human-led tasks throughout our working days. With AI still evolving as a newer concept, however, there is little standardization in software stacks and not much consistency when it comes to measuring the performance of multiple inference engines.
Procyon AI Benchmarks designed with industry leaders and experts to compare performance
Comparing a device’s performance is where Procyon remains relevant and a key aspect of the movement to integrate AI into our everyday lives. The Procyon AI Inference Benchmark for Android has been available to license for some time and focused primarily on the smartphone market. AI has typically, until this point, been associated with smartphones — consumers have enjoyed using AI technology, such as image classification, for some time. Many of us are familiar with our smartphone’s capacity to sort and segment photos and automatically link similar images together or create a photo album of your pet. Now, the revolution is coming to PC’s, and our latest AI benchmark is focused solely on assessing this area of hardware.
Our recently launched UL Procyon AI Inference Benchmark for Windows addresses the demand for better AI capabilities. Developed for industry vendors who create AI-specific hardware, Procyon benchmarks provide neutral scoring. This neutral scoring allows the new benchmark to be a standardized tool for assessing all AI inference performance. Software development kits (SDKs) used to measure the AI interference performance include Microsoft® Windows ML, Qualcomm ® SNPE, Intel® OpenVINO™ and NVIDIA® TensorRT™.
A benchmark that can provide insights for the next generation of technology
AI has created possibilities of automation that we couldn’t imagine before. Those developing this technology are now racing to find ways to harness this automation, reducing the burden of manual tasks and finding new ways to make AI accessible in the everyday work we complete.
The Procyon AI Inference Benchmark for Windows runs performance tests, comparing different accelerators like CPUs, GPUs and dedicated accelerators and providing data surrounding inference performance of integer and floating-point accuracy. These comparisons can help you decide which inference engine will give you optimal performance. For example, our data indicate that GPUs are faster in AI performance than CPUs. There can be performance benefits in using lower precision models compared to Floating Point models. This kind of performance data could provide critical insights for the next generation of personal and professional devices.
Currently, the benchmark supports AI workloads such as image classification, image segmentation, object detection and super-resolution. The development for this benchmark is ongoing and will evolve to continue to make this benchmark relevant to new AI use cases.
New features could enhance productivity for employees
Ensuring your employees' fleet of computers is optimized for them to perform their work tasks efficiently and quickly is essential. As these new AI features become a reality within applications, the automation of certain tasks may pave the way for workers to be more creative, and productive. Procurement teams and IT leads will want to make sure their teams’ laptops are primed to utilize these new AI features. Understanding how AI inference engines perform on hardware in a Windows environment by regularly testing with an impartial benchmark can help make informed purchasing decisions on whether to keep or retire PCs based on how they are running.
Procyon Benchmarking Suite
The UL Procyon benchmark suite offers flexible licensing, letting you choose the benchmarks that best meet your needs. You can buy just one benchmark or add more in any combination.
Find out more about UL Procyon benchmarks.
