I'm starting a degree in artificial intelligence and want a desktop that can support my coursework, university projects, personal programming and machine-learning experiments for several years. I'd also like to use it for gaming and possibly game development. I probably won't be able to upgrade the graphics card for around seven years, so I want to choose carefully.
My current options are an NVIDIA RTX 5070 with 12GB of VRAM or an AMD RX 9070 XT. I've heard that AMD is less suitable for AI because it does not support CUDA and has a weaker software ecosystem. Is that a serious limitation for someone who wants to become an AI engineer, or would the 9070 XT's additional performance and memory be more useful? I could also wait roughly a year for newer cards or for models such as the RTX 5070 Ti to become more affordable.
What would be the most sensible choice for a student, especially if I plan to run AI models locally as well as play games?
2 Answers
The right choice depends heavily on what you actually plan to run. For an entry into AI, you do not necessarily need a top-end graphics card, and many coursework projects can use smaller models, remote systems, or cloud services. However, NVIDIA generally has the smoother experience for machine learning because CUDA and related libraries are more widely supported. If local AI work is a major priority and the budget allows it, an NVIDIA card—possibly a 5070 Ti if its price improves—would be the safer software choice.
The 9070 XT may offer stronger gaming performance and more VRAM, but hardware specifications are only part of the decision for AI. Some frameworks can work with AMD through alternative software stacks, yet compatibility and setup are often less straightforward than with NVIDIA. Also, 12GB is not automatically a deal-breaker for learning; the models and workloads you intend to use matter much more. Waiting could make sense if you are not in a hurry, since newer cards or lower prices may give you better options.
Related Questions
Lenovo Thinkpad Stuck In Update Loop Install FilterDriverU2_Reload