Hi everyone,
I’m diving into machine learning and looking for a new laptop that can handle models and libraries like TensorFlow and PyTorch. I’ve heard great things about AMD Ryzen 5 processors, and I’m curious to know which models stand out for this kind of work.
Here are a few questions to get the discussion going:
Performance: What AMD Ryzen 5 laptop have you found to perform best with machine learning tasks? Are there specific models you recommend based on your experience?
RAM and GPU Considerations: For running ML models efficiently, what specs should I be looking for in terms of RAM and GPU? Any recommendations on combinations that work well?
Thermals and Cooling: How do different Ryzen 5 laptops handle heat during intensive computations? Are there any models known for good thermal management?
Budget-Friendly Options: If anyone has recommendations for budget-friendly Ryzen 5 laptops that still offer decent performance for machine learning, I’d love to hear them!
Real-World Experiences: If you’re using an AMD Ryzen 5 laptop for ML, what has your experience been like? Any challenges or standout features?
Thanks for your input! I’m looking forward to hearing your recommendations and experiences. Let’s help each other find the best gear for our machine learning projects!
1 Like
I don’t own such a nice laptop, so I can only generalize, but first of all, a laptop for business use that is designed for paperwork, etc., would be out of the question.
Among gaming laptops, there are stylish models and bulky models with multiple fans for cooling, but I think bulky models are better. The reason is that generative AI is more damaging to the PC than gaming, since it is always running.
The CPU is not affected as much. The more powerful and multi-threaded the better, but any current Ryzen 5 will meet your baseline.
what specs should I be looking for in terms of RAM and GPU?
Anyway, definitely a model with an nVidia GPU and a lot of VRAM. It’s hard with AMD’s or Intel’s because libraries rarely support them.
Prioritize VRAM capacity even if other specs are ignored.
If possible, it is preferable to use a model with a lot of main RAM from the beginning, or a model that can be expanded later. Even 64GB is often not enough.
In any case, the more the better.
I personally own a Ryzen 7 processor I got cheap for like $200 in a mini-pc, but I think the best for your use case would be an Intel 5,7,9 with at least a Nvidia RTX GEFORCE 3050, because it has CUDA capabilities and kind of the entry level for ML. Good luck and hope you find a great deal out there as prices keep dropping. 
1 Like
I will not recommend it, if you are totally begginer try using Google Colab and other cloud computing services. Instead of buying a laptop for only that.
1 Like
I agree with you that it would be better to use it with a cloud service rather than leaving it all to the laptop.
With Colab, I heard that even the free version has a 16GB VRAM GPU at some moments.
However, I think it would be better to have a dedicated PC (desktop or laptop) for light work (comfortable coding, conversion of models and datasets, and testing of operation).
If you try to do everything on a free cloud service, you will spend a lot of time struggling with the limitations of the cloud service. It would be great to have both a PC or workstation with specs that can do everything and a paid cloud service, but that’s not exactly an entry-level application.
For now, let’s not get a laptop for business use. Those are not even usable for conversion purposes. In terms of generative AI applications, they are for cloud services only. They have good CPUs, but they are desperately short on RAM and cannot be expanded.
1 Like