Who is a mini PC for data science students suitable for?

Students working with Python notebooks

If you prepare data and use several tools at once, check the RAM installed and the requirements of your packages. The Python Packaging Authority points out that scientific software often has complex dependencies; scikit-learn also explains how to control parallel computations using the n_jobs parameter.

Students sharing a desk

If other people are working right next to you, the device's footprint, suitable ports and noise levels under load all matter. Notebookcheck found the Minisforum EliteMini UM890 Pro it tested to be quiet; this result does not apply to other models across the board.

Students taking GPU-based courses

If your coursework requires a particular GPU or software environment, check these requirements against the operating system and hardware before buying. PyTorch documents different options for CUDA, ROCm and the MPS backend on macOS.

What should you look for when buying?

First check the configuration you will actually receive and the software required for your course. Then consider noise levels and the ports you need for your workspace.

What matters What to check before buying Who should consider this?
RAM and upgrade options Check the RAM supplied and potential upgrades for the exact model separately. If I use lots of notebooks and analysis tools at the same time, the RAM actually installed and whether it can be upgraded matter to me. Around 31% of consumers prioritise RAM and upgrade options.
Noise during longer sessions Compare reviews of the exact model, ideally including how it performs under load. If other people are studying or working right next to me, I want to be able to judge the computer's noise levels as accurately as possible. Around 25% of consumers prioritise quiet operation.
Course software and GPU Check the requirements for the operating system, GPU and acceleration backend. If an assignment specifically requires a particular GPU or software environment, my computer needs to be compatible. Around 19% of consumers prioritise compatibility with course software and GPU tasks.
Ports and desk space Check the monitor, USB and network ports, and how they are positioned in relation to your desk. If my workspace is small or has fixed cabling, I want to connect my existing monitors and devices without unnecessary workarounds. Around 13% of consumers prioritise ports and desk space.
Complete setup Check what is included, which operating system it has and which components come with the version you have chosen. If I want to set up the device straight away, it matters whether RAM, an SSD and an operating system are already included. Around 12% of consumers prioritise a complete setup and manageable additional costs.

Common mini PC buying pitfalls

  • Assuming a barebone is the same as a complete PC

    A version sold as a barebone may not include RAM, an SSD or an operating system. Its specification cannot then be compared directly with a fully configured computer.

    Check the exact version and what is included before deciding which components you still need.

  • Assuming any GPU will work for a CUDA course

    PyTorch distinguishes between CUDA-capable NVIDIA GPUs, supported AMD ROCm configurations and the MPS backend on macOS. Simply having a GPU or GPU acceleration does not prove that the requirements of a particular course will be met.

    Check your course requirements against the hardware, operating system and backend you need.

  • Assuming a compact design means quiet operation

    Noise levels under load can vary between devices. Notebookcheck found the Minisforum EliteMini UM890 Pro it tested to be quiet, while the GEEKOM A8 it tested was audible under load.

    Use reviews of the exact model as a guide, and don't assume their findings apply to similarly named devices.

Key brands to consider

Minisforum

The EliteMini UM890 Pro is a specific model whose noise levels have been assessed in an independent review.

  • Notebookcheck described the cooling on the device it tested as quiet.
  • The review notes limited BIOS options and a storage conflict when OCuLink is used.

GEEKOM

The A8 models illustrate some of the different questions to consider around specifications and future memory upgrades.

  • GEEKOM says the A8 Max has upgradeable RAM.
  • Notebookcheck found an A8 with a Ryzen 7 8745HS and 32 GB RAM audible under load. This is not a noise review of the A8 Max.

Apple

The Mac mini running macOS is an option if the course software and tools you need are compatible with the platform.

  • Apple offers M6 configurations with different amounts of unified memory and SSD storage.
  • PyTorch documents MPS for macOS; MPS is not CUDA.

Lenovo

The ThinkCentre M75q Gen 5 Tiny is a business-focused option if easy maintenance access and security at your workstation matter to you.

  • Lenovo says this range offers tool-free access to the SSD and memory modules, as well as a Kensington slot.
  • These manufacturer specifications do not establish independent noise-test results or a particular level of data science performance.

HP

The Pro Mini 400 G9 is an example of a pre-configured business model with Windows 11 Pro.

  • For part number 937U0EA#ABD, HP specifies an Intel Core i5-14500T, 16 GB RAM and a 512 GB SSD.
  • The German product page marks this exact version as out of stock; this does not tell you anything about its noise levels or suitability for a particular dataset.

Suitable devices at Joybuy

This selection includes compact Windows mini PCs as well as Mac mini configurations. For each device, check that its specifications meet your course requirements and provide the ports you need.

SOYO Y5Pro mini PC with Intel Core i5 8250U/8350U, 8 GB RAM and 256 GB SSD

SOYO Y5Pro mini PC, Intel Core i5 8250U/8350U

A compact option if you need two HDMI outputs for your workspace.

8 GB DDR4 RAM and a 256 GB SATA SSD.

Two HDMI outputs, Type-C, four USB ports and 1 Gbit Ethernet.

The product name lists the 8250U/8350U processor variants; check which version is on offer.

View at Joybuy
Dell Pro Max Micro FCM2250 with Core Ultra 7, 16 GB RAM and 512 GB SSD

Dell Pro Max Micro FCM2250, Core Ultra 7

A micro PC configuration for a space-conscious workspace, if Windows 11 Pro suits your tools.

16 GB DDR5-6400 RAM and a 512 GB M.2 PCIe SSD.

Intel graphics, Gigabit LAN and Wi-Fi 7.

The product name specifies a 14-core processor with an NPU.

View at Joybuy
Apple Mac mini with M6, 16 GB unified memory and 256 GB SSD

Apple Mac mini, M6, 16 GB, 256 GB SSD

A macOS option if your course software is compatible with the platform.

12-core CPU, 12-core GPU, 16 GB unified memory and a 256 GB SSD.

PyTorch documents MPS for macOS. MPS is not CUDA and will not meet the requirements of a course that specifically requires CUDA.

View at Joybuy
Apple Mac mini with M6, 24 GB unified memory and 512 GB SSD

Apple Mac mini, M6, 24 GB, 512 GB SSD

A macOS option if your course software is compatible with the platform and you're looking for this memory configuration.

12-core CPU, 12-core GPU, 24 GB unified memory and a 512 GB SSD.

PyTorch documents MPS for macOS. MPS is not CUDA and will not meet the requirements of a course that specifically requires CUDA.

View at Joybuy
Apple Mac mini MHQL4_CTO_8 A3534 with M6, 24 GB unified memory and 1 TB SSD

Apple Mac mini MHQL4_CTO_8 A3534, M6, 24 GB, 1 TB SSD

A macOS configuration with a 1 TB SSD, if your course software is compatible with the platform.

12-core CPU, 12-core GPU and 24 GB unified memory.

PyTorch documents MPS for macOS. MPS is not CUDA and will not meet the requirements of a course that specifically requires CUDA.

View at Joybuy

Frequently asked questions about mini PCs for data science students

Do I need a dedicated GPU for every data science course?

No. Studying data science does not automatically mean you need a dedicated GPU. Check your course assignments: scikit-learn documents CPU parallelisation, while for courses that explicitly require CUDA, PyTorch specifies a CUDA-capable NVIDIA GPU. For macOS, PyTorch instead documents the MPS backend.

What should I check before setting up a barebone mini PC?

Check whether the specific version includes RAM, an SSD and an operating system. A barebone version listed at Joybuy is sold without these three components; this does not apply to other mini PCs or versions.

How can I tell if a mini PC will be too noisy in a shared office?

Compare reviews of the exact model, ideally including tests under load, rather than relying on the size of the case alone. Notebookcheck described the Minisforum EliteMini UM890 Pro it tested as quiet and the GEEKOM A8 it tested as audible under load; these findings apply to the devices tested.

What rights apply when buying online as a private customer in Germany?

For a standard online purchase of goods, you generally have a 14-day withdrawal period from the date you receive the goods, as explained by the Verbraucherzentrale. Your statutory rights if goods are faulty are separate from any voluntary manufacturer's guarantee; exceptions and the circumstances of the individual case are still relevant.

What should I check on a mini PC's product labelling?

For electrical and electronic equipment, the German Federal Network Agency lists correct CE marking, manufacturer details, a clear product identifier and German-language instructions among the points retailers must check before offering a product for sale. Your Europe explains the manufacturer's responsibility for meeting the applicable EU requirements. Check the labelling on the specific device and power supply rather than relying on a general product description.