Who is this buying guide for?

Student using local notebooks

You want to programme, analyse data and run several applications at once. Focus on the software requirements, memory and SSD. Rotterdam University of Applied Sciences recommends 16–32 GB RAM for Applied Data Science & AI, but the requirements for your own course should take priority.

Student between home and campus

You use the same laptop at a small desk and on the go, so you need to balance screen space with size, ports and portability. An external monitor or projector may also require a suitable port or adapter.

Student with local AI assignments

First check whether your assignments need to run with GPU acceleration locally. If they do, CUDA support and video memory become key factors in your choice.

What should you look for?

First, check your course's software, operating system and hardware requirements. Then compare the memory and storage configuration, screen size and ports of the exact model.

The points below are aimed at students choosing a laptop for data science work in a student room.

What to consider What should you check? Who should prioritise this?
Memory and storage Compare the RAM and SSD configuration with your course requirements, and check whether that model can be upgraded. Rotterdam University of Applied Sciences specifies at least 12 GB RAM for Applied Data Science & AI and recommends 16–32 GB. If I want to use local notebooks, datasets and other applications at the same time, I need to know whether the configuration can handle it. This matters to around 33% of consumers who prioritise local study work.
Screen space and portability Balance screen size with portability and desk or monitor use; check the ports and any adapters you may need. If I often take my laptop between home and campus, I need to balance a handy size with enough room for code and documentation. Around 27% of consumers make portability their top priority.
Local GPU and software Ask which software, operating system and computing environment your course uses. For local CUDA assignments, check the exact NVIDIA model and VRAM configuration. If my modules require local GPU work, I want to check the required support in advance. Local GPU processing is the top priority for around 20% of consumers.
Total cost of ownership Check what comes with the exact model, which software your course provides and what an additional service plan includes. If I want to understand the total purchase and running costs, I also check accessories and software. Around 20% of consumers put these costs first.

Common buying mistakes

  • Confusing an AI label with CUDA support

    An integrated GPU or an AI label does not prove that a laptop has the NVIDIA CUDA configuration required for a specific assignment. Not every course requires local GPU processing, either.

    Check your course software and assignments first. Then compare the GPU model and video memory of the exact configuration.

    If compulsory assignments require local CUDA processing, you may need a different suitable laptop or computing environment.

  • Treating one course's requirements as the standard

    Requirements vary by course: Rotterdam University of Applied Sciences and the University of Amsterdam do not set the same GPU and macOS requirements. So, one course's requirements do not automatically apply to every data science student.

    Use the published requirements for your own course and academic year as your starting point.

    Choosing a laptop based on another course's requirements could mean spending more than you need to or buying a laptop that does not suit your own practicals.

  • Focusing only on the laptop and forgetting accessories

    A projector or external monitor may need a suitable port or adapter. Rotterdam University of Applied Sciences notes that you may need an HDMI adapter.

    Before comparing the total cost, check the ports on the model you have chosen, the inputs on your screen and what is included in the box.

  • Confusing the manufacturer's warranty with your statutory rights

    According to ACM ConsuWijzer, your statutory guarantee is separate from any guarantee offered by the retailer or manufacturer. What you can reasonably expect from a product is relevant to your statutory protection.

    Check what an additional service plan actually covers. If a product does not meet the standard you could reasonably expect, contact the retailer.

    A paid warranty plan may cover something you already have or do not need; compare the terms before paying for it.

Brands and product ranges for reference

Lenovo

The ThinkPad E14 Gen 7 we reviewed, configuration 21T9001TMH, is a compact Windows option with a 14-inch screen.

  • Lenovo lists a Core 5 210H, 16 GB RAM and a 512 GB SSD for this configuration.
  • This configuration has integrated Intel graphics, so it is not a proven choice for local NVIDIA CUDA assignments.

ASUS

The Vivobook 16 M1607KA is a Windows option with a larger, anti-glare screen.

  • ASUS specifies a 16.0-inch screen, a Ryzen AI 7 350 and integrated Radeon graphics for the M1607KA.
  • Check the full configuration and its memory and storage specifications; details for the M1607GA do not automatically apply to the M1607KA.

Apple

The 13-inch MacBook Air with M5, 16 GB memory and a 512 GB SSD that we reviewed is a compact macOS option, provided your course supports the software you need.

  • Apple specifies a 13.6-inch display and two Thunderbolt 4 ports. PyTorch supports macOS.
  • Software and practical-class support varies by course; you may need an adapter to connect to an HDMI projector.

HP

The ProBook 4 G1a 16 we reviewed, configuration C5BX3EA#ABH, is a larger Windows laptop with 24 GB memory.

  • HP lists a 16-inch screen and a 512 GB SSD for this configuration.
  • This configuration has integrated Radeon 780M graphics, which is no proof that it is suitable for local CUDA assignments.

Joybuy product picks

This selection includes laptops with a dedicated NVIDIA GPU and models with a larger screen and integrated graphics. Before choosing, check whether your course requires local GPU processing and which configuration suits the way you work.

ASUS ROG Zephyrus G16 GU605CM-QR026W

ASUS ROG Zephyrus G16 GU605CM-QR026W

Worth considering if you need local GPU processing; check that the specifications meet your course requirements.

16-inch OLED screen, NVIDIA GeForce RTX 5060 Laptop GPU with 8 GB GDDR7, 16 GB RAM and a 1 TB SSD. The power adapter must be purchased separately.

Joybuy
ASUS ROG Strix G16 G614FR-S5252W

ASUS ROG Strix G16 G614FR-S5252W

This configuration may suit local GPU work if its GPU and software meet your course requirements.

16-inch screen, NVIDIA GeForce RTX 5070 Ti Laptop GPU with 12 GB GDDR7, 16 GB RAM and a 1 TB SSD. The charger must be purchased separately.

Joybuy
ASUS Vivobook S16 S3607QA-SH109W

ASUS Vivobook S16 S3607QA-SH109W

An option to compare if you want a 16-inch screen and integrated graphics; check that your course does not require a dedicated GPU.

16-inch WUXGA OLED screen, Snapdragon X processor, 16 GB LPDDR5X memory and a 512 GB SSD. The product description lists integrated Qualcomm graphics.

Joybuy
Auusda T156A laptop

Auusda T156A

A 15.6-inch screen may suit you if you often work at a desk; check that the size and full configuration are right for your studies.

15.6-inch FHD IPS screen, 16 GB DDR4 RAM, 512 GB SSD and 3050E processor. If you have local CUDA assignments, the GPU configuration is an important detail to check.

Joybuy

Frequently asked questions

Do I always need an NVIDIA graphics card for data science?

No, it depends on your course and assignments. For a new high-end laptop for Applied Data Science & AI in 2026/2027, Rotterdam University of Applied Sciences prefers NVIDIA with at least 8 GB VRAM; the University of Amsterdam says a powerful graphics card is not needed for its Bachelor's in Computer Science. Let the requirements of your own modules guide your choice.

Can I use a MacBook for Python and practical classes?

Yes, if your course supports the software and practical classes you need on macOS. PyTorch supports macOS and Rotterdam University of Applied Sciences accepts macOS for Applied Data Science & AI, but the University of Amsterdam does not support practical classes on MacBooks for its Bachelor's in Computer Science.

What should I check when connecting a monitor or projector?

Check the ports on the exact laptop and the inputs on your screen. Rotterdam University of Applied Sciences notes that you may need an HDMI adapter; Apple lists two Thunderbolt 4 ports for the 13-inch MacBook Air we reviewed.

How can I avoid unexpected software and service costs?

Ask your course which software packages you need and what it provides. Also check which power supply and accessories are included with the configuration you have chosen, and compare the terms of any paid service plan with your statutory rights.

How long is the cooling-off period and guarantee for an online purchase in the Netherlands?

According to ACM ConsuWijzer, when you buy a product online you generally have up to 14 days after delivery to change your mind, with some exceptions. You also have a statutory guarantee if the product does not perform as you could reasonably expect under normal use. There is no general fixed two-year period for this right, and the retailer is responsible for providing an appropriate solution.