Machine for nerfstudio | Information dissemination media for research and development TEGAKARI

A customer involved in information engineering research in education asked us about a workstation to use the NeRF framework "nerfstudio."We would like to make a proposal with a budget of 60 yen.

The installed GPU can be found on the official nerfstudio website.Currently supported CUDA architectures in the docker imageI would like the model listed in the table above, and the overall conditions are as follows.

・CPU: Core i9-13900K
・ Memory: 64GB
-Storage: 2TB SSD M.2
・Network: 2.5GbE x1
・GPU: GeForce RTX3080 or GeForce RTX4070Ti, model listed in the table on the official website
・Power supply: Compatible with 100V environment
・OS: Ubuntu 22.04
・Software used: Conda, Python 3.11, nerfstudio
・Budget: Approximately 60 yen

Based on the conditions you communicated with us, we proposed a configuration that incorporates the parts specified by the customer.

CPU	Intel Core i7-13900K (3.00GHz 8 cores + 2.20GHz 16 cores)
memory	64GB
storage	2TB M.2 SSD
video	NVIDIA GeForce RTX4070Ti 12GB
network	on board (2.5GbE x1) Wi-Fi x1
Housing + power supply	Middle tower type housing + 850W
OS	Ubuntu 22.04

This configuration is equipped with the Intel Core i2023-10K, the latest 13th generation Intel Core series as of October 9.This CPU is a high-end model with 13900 cores, the most in the series.

For the video card, the customer has selected the Geforce RTX4070Ti, which was assumed to be a candidate.In terms of price, this product is at the top of the range among the products you can choose within your budget.
*Geforce RTX4070Ti is a model that is also listed in the table on the official nerfstudio website.

Additionally, this configuration supports operation in a 100V environment.

The configuration of this case study is based on the conditions given by the customer.
We will flexibly propose machines according to your conditions, so please feel free to contact us even if you are considering different conditions than what is listed.

■ Keywords

・What is nerfstudio?
nerfstudio is a user-friendly framework for working with NeRF. It was launched as an open source project in October 2022 by students at the Kanazawa AI Laboratory at the Berkeley AI Institute at the University of California, Berkeley, and is currently being developed by the school's students and community developers.By making each component modular, you can more easily create, learn, and test NeRF.

reference:nerfstudio *Jumps to an external site

・What is NeRF?
NeRF is a method of estimating the 3D shape of an object and the radiance of its surface using deep learning.The shape is represented as a radiance field in a continuous 3D space instead of a point cloud, and estimated using a neural network from the image.High-quality images can be generated from the estimated radiance field at arbitrary viewpoints and lighting.Compared to conventional methods, photorealistic image generation is possible, and since 3D shapes can be represented continuously, it is suitable for use in VR/AR.

・What is Python?
Python is an object-oriented programming language copyrighted by the Python Software Foundation (PSF).Its programming syntax is simple, making it highly readable, and it also features a wide variety of components, such as libraries and frameworks, that are suitable for different purposes.A popular language for programming beginners to advanced users.

Reference: Python *Jumps to an external site

Reference: [Special article] Why is the programming language Python so popular? - Tools to accelerate your Python programming * Jump to our owned media "TEGAKARI"