Neu.ro announces plans to develop zero emission AI cloud solution

Neu.ro, a tech company that uses Machine Learning (ML) and Deep Learning (DL) to develop business solutions, has revealed plans to develop a zero-emissions AI cloud solution in unison with Finnish cloud infrastructure partner, atNorth.

The plan comes in response to companies constantly incorporating ML (machine learning) for running their business effectively, which consequently requires massive amount of energy to test, run and build models in production.

According to Neu.ro, atNorth is offering Tier 3, ISO 27001-certified data center, running Nvidia A100-powered DGX and HGX systems. The data center offers a capacity of 80MW, which functions entirely on geothermal and hydro energy.

Additionally, the location is near to arctic which offers a free cooling solution, eventually helping Neu.ro to create machine learning models using its solutions.

After researching the problem, Max Prasolov, the company’s co-founder, said that they found that computing and telecommunications are consuming around 9% of total energy globally.

As per their research, this number could be doubled in the next decade. Building a machine learning model will be a significant part of that, and they decide to collaborate with atNorth for reducing their carbon emissions footmark.

Prasolov stated that they’ve decided to move all of its experiments and operations to zero-emission cloud. With achieving zero emissions being the goal and their firm has realized that they were spending too much on energy and computing power to train their clients and understood that this was producing the most carbon footprint.

Currently, Neu.ro has a commercial solution, but is working on an open source version of the stack which will be launched prior to end of the year. The firm’s goal is to provide a cloud-based AI solution for smaller cloud vendors, including regional vendors across the world.

Source Credit - https://techcrunch.com/2021/11/22/neu-ro-announces-zero-emissions-model-building-solution/