ASUS NUC 15 Pro Slim Full System Mini PC
ASUS www.asus.com/ $1,254.02 Price as of 08/23/2026 View on Amazon → ⓘ We may earn a commission when you buy through our links. This helps support MacSources and keeps our content free. Thank you!The ASUS NUC 15 Pro delivers strong productivity performance and practical local AI capabilities in an incredibly compact package, but its Core 7 configuration has important limitations for demanding AI and graphics workloads.
Over the past couple of years, AI has sort of taken the world by storm. It’s been fascinating to see the field grow, and using AI tools has without a doubt changed how I analyze data. In order to get the most out of using AI tools, though, you need a machine that can handle the processing power required by these increasingly demanding workloads.
That raises an interesting question when looking at a computer as small as the ASUS NUC 15 Pro. Can a compact mini PC provide enough performance for modern productivity while also handling AI workloads locally? ASUS specifically positions the NUC 15 Pro as a system designed for AI-driven productivity, edge computing, data visualization, digital signage, and multi-display business environments.
With its compact size and focus on productivity and AI, the ASUS NUC 15 Pro presents an interesting proposition for anyone who wants a capable desktop without dedicating a large amount of space to a traditional PC. The question is whether this particular configuration can live up to ASUS’s claims while still providing the everyday performance users expect from a modern desktop.
About the ASUS NUC 15 Pro
The ASUS NUC 15 Pro is a compact desktop computer designed to deliver capable everyday computing performance in a remarkably small footprint. Despite its compact size, the NUC 15 Pro offers a good selection of connectivity options, expandable memory and storage, and enough processing power for productivity, web browsing, media consumption, and a variety of everyday computing tasks.
The configuration I am reviewing is equipped with an Intel Core 7 240H processor, featuring 10 cores and 16 threads, a 2.5GHz base frequency, and a maximum turbo frequency of up to 5.2GHz. The processor is paired with 16GB of DDR5 memory running at 5600 MT/s, Intel Graphics, and a 1TB PCIe 4.0 NVMe SSD. This combination provides a solid foundation for everyday productivity while offering enough storage for applications, documents, media, and other files.
The NUC 15 Pro also provides an impressive selection of connectivity options for a system of its size. The review unit features two Thunderbolt 4 ports, HDMI 2.1, USB Type-A and Type-C ports, and 2.5Gb Ethernet. It also supports Wi-Fi 7 and Bluetooth 5.4, giving users plenty of options for connecting displays, peripherals, storage devices, and networking equipment without requiring a large desktop tower or extensive collection of adapters.
Another advantage of the NUC 15 Pro is its upgradeability. The system includes two SO-DIMM memory slots and support for additional M.2 NVMe storage, allowing users to expand the system beyond its original configuration. ASUS also supports VESA mounting, making it possible to attach the NUC 15 Pro to the back of a compatible monitor and create an even cleaner, more space-efficient workstation.
With its compact design, Intel Core 7 240H processor, 16GB of DDR5 memory, Intel Graphics, and 1TB PCIe 4.0 SSD, the ASUS NUC 15 Pro offers an interesting combination of performance, connectivity, and expandability. In this review, I will put this particular configuration through a variety of real-world and benchmark tests to see how well it performs as an everyday desktop computer.
Features & Specifications
Processor: Intel Core 7 240H Processor Base Frequency: 2.5GHz Processor Maximum Turbo Frequency: Up to 5.2GHz CPU Cores: 10 CPU Threads: 16 Configurable TDP: 25W Graphics: Intel Graphics Memory: 16GB DDR5 RAM Memory Speed: 5600 MT/s Maximum Supported Memory: Up to 48GB DDR5-5600 Storage: 1TB M.2 2280 NVMe PCIe 4.0 SSD Additional Storage: Supports a second M.2 2242 PCIe 4.0 NVMe SSD Wireless: Intel Wi-Fi 7 Bluetooth: Bluetooth 5.4 Ethernet: 2.5Gbps RJ45 Thunderbolt: 2x Thunderbolt 4 Video Outputs: 2x HDMI 2.1 Front USB: 1x USB 3.2 Gen 2×2 Type-C, 2x USB 3.2 Gen 2 Type-A Rear USB: 1x USB 3.2 Gen 2 Type-A, 1x USB 2.0 Type-A Display Support: Supports multiple displays through HDMI and Thunderbolt 4 Audio: Realtek ALC3251 Expansion: 2x SO-DIMM slots and two M.2 storage slots Mounting: VESA mounting support Form Factor: Ultra-compact mini PC Operating System: Configuration dependent, with Windows 11 support Security: Kensington lock and firmware TPM supportPricing & Availability
The list price of the ASUS NUC 15 Pro starts at $429.00. The configuration I tested has a list price of $1,499.00. It can be ordered from asus.com or Amazon ($1,254.02).
User Experience
Getting the ASUS NUC 15 Pro set up was very easy. It comes with a power supply and cable. It does not come with HDMI or Thunderbolt cables even though it has those ports available for video connectivity. Other mini PCs I’ve tested have usually included at least a short HDMI cable. So, this was a little bit of a surprise to me.
On a whim, I decided to connect the ASUS NUC to a Thunderbolt dock so that it would connect to all my peripherals. Now, I haven’t had the best luck with everything working perfectly with mini PCs, but this time, my keyboard, mouse, and monitor were all recognized and functioning without issue.
The startup of the ASUS NUC was a typical Windows 11 Pro experience. It did, however, take a little over an hour for updates to complete. Once I was logged into my Microsoft account and was on the desktop, I checked to see how many processes were running in Task Manager. Upon startup with nothing else running, there were 208 processes running. And, without installing any new apps, there were 76 pre-installed.
Many of those apps were Microsoft related. What I initially saw was multiple entries for Microsoft 365 and OneNote, including what appeared to be different language versions. After installing the available Windows and application updates and restarting the NUC a couple of times, these additional entries disappeared. While this initially looked unusual, it appears to have been related to the system completing its initial software setup rather than 19 separate installations of Microsoft 365 and OneNote.
Benchmark Performance
To get a better understanding of what the ASUS NUC 15 Pro is capable of, I ran it through several different benchmark suites covering CPU performance, graphics, AI workloads, and storage. These tests are useful because no single benchmark can tell the whole story. Looking at the results together gives us a much better idea of where this particular configuration performs well and where its limitations become apparent.
ASUS positions the NUC 15 Pro as a system designed for AI-driven productivity, edge computing, and business environments, with workloads including AI model development, data visualization, digital signage, and multi-display office setups. The benchmark results provide some interesting evidence in support of those claims, although there are also some important limitations to consider.
Geekbench 7
Geekbench 7 is designed to measure processor performance using a variety of real-world-inspired workloads. The single-core test is particularly useful for applications that rely heavily on one processor core, while the multi-core test demonstrates how well the processor handles workloads that can take advantage of multiple cores simultaneously. Geekbench’s GPU test provides an indication of how well the integrated graphics can handle compute workloads.
The ASUS NUC 15 Pro produced the following results:
Single-Core: 2,378 Multi-Core: 13,599 GPU OpenCL: 14,971The single-core result of 2,378 demonstrates that the Core 7 240H has plenty of processing capability for everyday applications. Tasks such as launching applications, browsing the web, working with office applications, and other workloads that depend heavily on individual CPU cores should feel responsive.
The multi-core score of 13,599 is more important when looking at workloads such as video encoding, compiling software, running multiple applications simultaneously, and other heavily threaded tasks. The 10-core, 16-thread processor is able to take advantage of its additional cores here, making the NUC considerably more capable than its tiny footprint might suggest.
The OpenCL score of 14,971 is a little more revealing. While the NUC’s integrated Intel Graphics is capable of handling GPU-accelerated workloads, this isn’t a system designed to compete with a desktop featuring a dedicated graphics card. The result suggests that the graphics hardware is better suited to productivity, display output, media acceleration, and lighter GPU workloads than demanding 3D rendering or gaming.
Geekbench AI
Geekbench AI focuses specifically on workloads associated with artificial intelligence and machine learning. This makes it particularly relevant to ASUS’s claim that the NUC 15 Pro is designed for AI-driven productivity.
The NUC produced the following results using the ONNX framework:
Single Precision: 3,764 Half Precision: 1,578 Quantized: 7,017The quantized score of 7,017 is the strongest of the three results. Quantization allows AI models to use lower-precision numerical representations, which can reduce memory requirements and improve performance on hardware designed to take advantage of these operations.
The results demonstrate that the NUC 15 Pro can run AI workloads locally, but they also need to be put into context. This particular configuration uses the Core 7 240H and Intel Graphics, rather than one of the Core Ultra configurations with a dedicated NPU. As a result, I would not interpret these results as evidence that this configuration is a particularly powerful local AI development machine.
For AI-assisted productivity applications, local inference, and experimentation with smaller AI workloads, however, the system is certainly capable. More demanding AI model development, particularly workloads involving larger models or intensive GPU acceleration, would benefit considerably from a system with a more powerful dedicated GPU or a configuration equipped with a more capable AI accelerator.
Real-World AI Workload
To see how the NUC 15 Pro handles AI workloads outside of synthetic benchmarks, I installed Ollama and ran Meta’s Llama 3.2 3B model locally. This provided a way to test AI inference without relying on a cloud service.
I started with a prompt asking the model to explain the differences between Wi-Fi 7 and Wi-Fi 6E in approximately 300 words. The model generated 389 tokens in 15.86 seconds, resulting in a generation rate of approximately 24.5 tokens per second.
I then ran a considerably longer prompt asking the model to produce a 500-word analysis of weather patterns in the Midwest United States. This generated 677 tokens in 29.33 seconds, or approximately 23.1 tokens per second.
The consistency between the two tests was encouraging. Despite the second test producing a substantially longer response, performance remained relatively close to the first test. This suggests that the NUC can maintain relatively consistent local inference performance with this model rather than slowing down significantly during longer responses.
I also monitored CPU and memory utilization through Windows Task Manager during both tests. CPU utilization increased to approximately 55-65% while the model was processing and generating responses, before dropping back to idle when the workload finished. Memory utilization increased from approximately 48% during the first test to 52% during the second. This indicates that the Core 7 240H has sufficient processing headroom to run a small local LLM without completely saturating the processor, although the system’s 16GB of RAM could become a limitation when attempting to run larger models while simultaneously using other applications.
The results demonstrate that local AI inference is practical on this configuration. At roughly 23-25 tokens per second, Llama 3.2 3B produced responses quickly enough for conversational use, brainstorming, summarization, and other lightweight AI-assisted productivity tasks. However, this should not be interpreted as evidence that the NUC is suitable for demanding local AI development. Llama 3.2 3B is a relatively small model, and larger models would place substantially greater demands on the system’s CPU and memory.
Overall, the test provides some support for ASUS’s claim that the NUC 15 Pro can be used for AI-driven productivity, but it also highlights an important distinction. This configuration is capable of running lightweight AI workloads locally, rather than serving as a high-performance AI workstation.
Cinebench
Cinebench is focused on CPU rendering performance and is particularly useful for determining how a processor handles sustained workloads. The single-thread test provides an indication of individual-core performance, while the multi-thread test shows how the processor performs when all available CPU resources are being utilized.
The NUC 15 Pro scored:
Single Thread: 452 Multi Thread: 2,850The single-thread result reinforces what we saw in Geekbench. The Core 7 240H is capable of providing responsive performance for everyday applications and workloads that don’t scale across multiple processor cores.
The multi-thread score is more important for heavier workloads. The NUC can make good use of its 10 cores and 16 threads, making it suitable for tasks such as content creation, software development, file compression, and other CPU-intensive workloads.
However, the difference between the single-thread and multi-thread scores also highlights an important characteristic of this machine. This is still a compact, relatively low-power system. It isn’t intended to replace a workstation-class desktop for sustained rendering or other extremely demanding workloads.
FurMark
FurMark is a graphics stress test that places a heavy workload on the GPU. Unlike the productivity-oriented OpenCL test in Geekbench, FurMark is particularly useful for seeing how the integrated graphics handles sustained 3D rendering.
The NUC 15 Pro produced:
Score: 1,276 Average Frame Rate: 21 FPS Resolution: 1920 × 1280 API: OpenGLThe 21 FPS result makes the limitations of the integrated Intel Graphics much clearer.
This isn’t a gaming-oriented system, and the FurMark result confirms that. At 1920 × 1280, the integrated graphics struggled with an intentionally demanding graphics workload. Users shouldn’t expect this configuration to provide the kind of 3D performance available from a dedicated GPU.
That doesn’t mean the graphics are inadequate for the intended purpose. For office applications, web browsing, video playback, multiple displays, digital signage, and other productivity-focused applications, dedicated gaming-class graphics aren’t necessary. The result simply establishes that graphics-intensive workloads are not the NUC’s strongest area.
CrystalDiskMark
CrystalDiskMark measures storage performance, including sequential and random read and write speeds. Storage performance can have a noticeable impact on application loading, Windows responsiveness, file transfers, and workloads that frequently read or write large amounts of data.
The 1TB SSD in this NUC produced particularly strong sequential results:
TestReadWriteSequential 1MiB Q8T17,112.95 MB/s6,441.72 MB/sSequential 1MiB Q1T14,549.39 MB/s3,448.43 MB/sRandom 4KiB Q32T1453.10 MB/s278.35 MB/sRandom 4KiB Q1T160.94 MB/s114.92 MB/sThe sequential read speed of more than 7,100 MB/s is particularly impressive. It shows that the included 1TB NVMe SSD is capable of taking advantage of the high bandwidth available through its PCIe interface.
Sequential write performance was also excellent at more than 6,400 MB/s.
The random 4K results are less spectacular, but that’s normal because random workloads are fundamentally different from large sequential file transfers. These smaller operations are more representative of the types of access patterns generated by an operating system and everyday applications.
Overall, the storage performance is one of the strongest aspects of this configuration. The fast SSD helps make the NUC feel responsive during normal use and gives it plenty of bandwidth for large file transfers and storage-intensive workloads.
What Do These Results Tell Us?
Taken together, the benchmark results paint a fairly consistent picture of the ASUS NUC 15 Pro.
The CPU is the strongest component of this configuration. The Core 7 240H provides good single-core performance while its 10 cores and 16 threads give the system considerably more multi-threaded capability than its tiny chassis would suggest. This makes the NUC well suited to general productivity, multitasking, software development, office applications, and other CPU-focused workloads.
The SSD is another standout feature. With sequential read speeds exceeding 7GB/s and sequential writes above 6GB/s, the 1TB drive provides excellent storage performance and helps prevent the system from feeling sluggish when loading applications or moving large files.
The integrated graphics are the biggest limitation. The Geekbench OpenCL score shows that the GPU is capable of useful compute workloads, but the FurMark result demonstrates that this isn’t a machine intended for demanding 3D graphics. That’s an important distinction because the NUC’s intended audience doesn’t necessarily require high-end graphics performance.
Does It Support ASUS’s Claims?
For productivity and business environments, yes.
The benchmark results support the NUC 15 Pro as a capable business and productivity machine. The Core 7 240H provides enough CPU performance for multitasking and demanding office workloads, while the fast SSD contributes to a responsive overall experience. The system’s compact design and extensive connectivity also make it well suited to space-constrained workstations and multi-display environments.
For edge computing, the results are also encouraging.
A compact system with a capable 10-core processor, fast NVMe storage, multiple display outputs, high-speed networking, and relatively modest power requirements can make sense for edge deployments, digital signage, kiosks, monitoring systems, and other applications where a traditional desktop would be unnecessarily large.
The AI claim requires more qualification.
The Geekbench AI results show that the system is capable of processing AI workloads, but I wouldn’t describe this particular configuration as a serious AI development workstation. The Core 7 240H model doesn’t have the dedicated NPU found in the Core Ultra variants, and the integrated Intel Graphics is not powerful enough to compete with a dedicated GPU for demanding AI workloads.
For AI-assisted productivity, smaller local AI workloads, experimentation, and applications that don’t require substantial GPU acceleration, the NUC should be capable. For training or running large AI models locally, however, a more powerful system would be a better choice.
The real-world LLM testing provides additional context. Running Llama 3.2 3B locally through Ollama produced approximately 23-25 tokens per second across two different workloads, demonstrating that the NUC can handle lightweight local AI inference without relying on cloud services. CPU utilization reached approximately 55-65% during the tests, while memory utilization remained around 48-52%.
These results make ASUS’s AI-driven productivity claim more credible, but they also establish the limitations of this particular configuration. The NUC is capable of running smaller AI models locally, but its 16GB of RAM and integrated graphics make it better suited to lightweight AI-assisted productivity than demanding AI development or large-model inference.
Ultimately, the benchmark results support ASUS’s positioning of the NUC 15 Pro as a compact productivity and business computer with some AI capability, rather than a high-performance AI workstation. For this particular Core 7 240H configuration, CPU performance, storage performance, connectivity, and size are its biggest strengths, while graphics-intensive and highly demanding AI workloads expose its limitations.
Conclusion
The ASUS NUC 15 Pro is an impressive example of just how much computing capability can be packed into a very small desktop. The Core 7 240H provides strong CPU performance for everyday productivity, multitasking, software development, and other CPU-focused workloads, while the fast PCIe 4.0 SSD delivers excellent storage performance. Its compact design is arguably its biggest advantage.
The NUC successfully ran Llama 3.2 3B locally through Ollama, while its Geekbench AI results also demonstrate that it can handle AI-related workloads. Because of this, I think it is positioned well as a compact productivity and business system that happens to have useful local AI capabilities. It can be used for AI-assisted productivity, smaller local models, edge applications, digital signage, and other workloads where processing data locally is useful. However, the Core 7 240H does not include the dedicated NPU found in some of the more AI-focused Core Ultra configurations, and its integrated graphics are not designed for the intensive GPU acceleration required by larger AI models or serious AI development. It simply isn’t the system I would recommend for demanding AI model development or large-model inference.
That distinction doesn’t diminish what the NUC 15 Pro accomplishes. In fact, it helps define where this system makes the most sense. If you need a small, powerful, highly connected desktop for productivity, business applications, edge computing, or multi-display environments, the NUC 15 Pro is a compelling option. If your primary reason for buying a computer is high-performance local AI, however, there are better-equipped systems available.
The biggest question is price. At a $1,499 list price for this particular configuration, the NUC 15 Pro is competing against larger systems that can offer more graphics performance for the money. Its value becomes much easier to justify if the compact form factor, connectivity, expandability, and deployment flexibility are important to you. If those characteristics aren’t priorities, the premium becomes harder to overlook.
Overall, the ASUS NUC 15 Pro is a capable and highly versatile mini PC that delivers considerably more performance than its tiny footprint suggests. It succeeds particularly well as a productivity and business computer, while its local AI capabilities make it more flexible for modern workloads. Just be aware that AI capability and AI workstation performance are two very different things, and this particular configuration falls firmly into the former category.
For more information, visit asus.com
Hence then, the article about asus nuc 15 pro review compact performance with local ai capabilities was published today ( ) and is available on MacSources ( Middle East ) The editorial team at PressBee has edited and verified it, and it may have been modified, fully republished, or quoted. You can read and follow the updates of this news or article from its original source.
Read More Details
Finally We wish PressBee provided you with enough information of ( ASUS NUC 15 Pro REVIEW Compact Performance With Local AI Capabilities )
Also on site :