Friday, March 29, 2024

DIY AI PC Builds

How to Build the Best AI PC - The AI revolution has arrived. Is your PC ready? (2024)
https://www.pcworld.com/article/2226887/build-the-best-ai-pc.html
"The most essential component in a next-gen AI computer is the Graphics Processing Unit. Originally built for high-end gaming and design workflows, GPUs are vastly superior to CPUs at running math-intensive calculations in parallel, which is essential to fast AI performance with deep learning and large language models ..."

"NVIDIA and its GeForce RTX 40 Series GPUs in particular have led the way in developing AI applications.
...
A GeForce RTX 4090 Ti GPU will perform up to 1.7x faster than a prior generation RTX 3090 GPU in AI applications, so it's worth prioritizing the GPU upgrade over CPU performance if you're on a budget."


How to Build an AI PC (2024):
https://www.cclonline.com/article/2215/Blog/CCL-Specialist-Gaming-PCs/How-to-Build-an-AI-PC/
"With the advent of the Nvidia 40 series GPUs and Intel and AMD CPUs embedded with cutting-edge AI technology, users now have access to hardware that not only excels in traditional computing tasks, but also possesses the neural prowess to tackle complex AI workloads with ease."
"Some of both Intel and AMD's newest lineups of CPUs come fitted with a specialised block called a Neural Processing Unit, or NPU for short. "


Best PC for Machine Learning – An entry-level Guide (2022):

"CPU Recommendations:
  •     Performance – AMD Ryzen Threadripper 3960X: With 24 cores and 48 threads, this Threadripper comes with improved energy efficiency and exceptional cooling and computation.
  •     Value – Intel Core i7-12700K: At a combined 12 cores and 20 threads, you get fast work performance and computation speed.
  •     Budget – Intel Core i5-12600K: With a combined 10 cores and 16 threads, you get a great CPU at an amazing price."

"GPU Recommendations:
  •     Performance – GeForce RTX 3090 super: This absolute beast of a GPU is powered by Nvidia's Ampere (2nd gen) architecture and comes with high-end encoding and computing performance and 24GB of GDDR6X RAM. It will chew through anything you throw at it.
  •     Value – GeForce RTX 3060: Based on the same architecture as the RTX 3090, the RTX 3060 is an excellent card with RT and Tensor cores that'll give great training performance and efficiency.
  •     Budget – GeForce GTX 1060: The GTX 1060 is still a respectable GPU when it comes to most tasks, however, I wouldn't expect much when it comes to large parameter deep learning models or neural networks."

"RAM Recommendations
  •     Performance – Corsair Vengeance LPX 32GB (2 X 16GB) 3600MHz DDR4 C18
  •     Value – Crucial Ballistix 16GB (2 x 8GB) 3000MHz DDR4 CL15"

"Storage Recommendations

  •     Performance – Samsung 980 Pro 1TB NVMe Gen 4 SSD
  •     Value – Samsung 980 500GB NVMe Gen 3 SSD
  •     Budget – Crucial MX500 250GB SSDs"

"Motherboard Recommendations
  •     Performance – ASRock TRX40 Creator – Best motherboard for AMD Threadrippers
  •     Performance – Gigabyte Z690 AORUS MASTER – Best motherboard for Intel Alder Lake CPUs
  •     Value – Gigabyte Z690 UD – Best motherboard for budget Intel Alder Lake CPUs"



No comments: