Best laptop for machine learning

The best laptop for machine learning does not need a powerful processor. Most applications are only multi-threaded on the GPU, so don't go broke on the CPU. Therefore, you should buy a processor with the highest clock speed, i.e. i7-6700K at 4 GHz.

Motherboard

Since we want to fit 4 graphics cards in our case, our choice is very limited. The best option would be the MSI Extreme Gaming X99A SLI Plus motherboard. ASUS X99 Deluxe II is also suitable.

If you use less than 4 cards, there are many more options. In motherboards, the main thing is stability. I was convinced of this when I assembled systems for cryptocurrency mining. Gigabyte has developed an excellent line of reliable motherboards. The X99 Ultra Gaming model is optimal and costs about 20 thousand.

Frame

Cooler Master Cosmos II is the ultimate EATX form factor case. Its stylish appearance will perfectly complement your system.

For those looking for a smaller case, the Cooler Master Maker 5T is the way to go.

I never saw the point in buying a cheap case. However, there are many decent budget options on the market, and listing them here does not make sense.

Your machine learning machine does not need a powerful processor. Most applications are only multi-threaded on the GPU, so don't go broke on the CPU.

Therefore, you should buy a processor with the highest clock speed, i.e. i7-6700K at 4 GHz. Yes, it may seem like overkill, but lately the price of it has only been falling, so why not?

Power Supply

The EVGA Modular 1600 Supernova G2 PSU is the best option for a 4-GPU build. It will cost you about 18 thousand.

Titan X consumes about 250 watts - now a thousand, and there is not so much left for everything else, and it is better to take the PSU with a margin.

If you're using fewer graphics cards, the 1300W model will suit you. The DiS alternative is the 750W model.