Leafcloud Sustainable Cloud GPUs

By: Guy Pathak - August 25, 2021 ~ 2 min read

Leafcloud cutting carbon emissions with truly green cloud-based GPUs 

We at Leafcloud are excited to announce the addition of sustainable cloud GPUs to our range of services. GPUs allow for high-performance compute servers, and their use is growing rapidly due to the widespread adoption of AI, machine learning, deep learning. Unfortunately, regular cloud GPUs are extremely energy-intensive processes.

Introducing truly green NVidia Tesla v100 GPUs:

Top performance starting at € 1,47/hr

We’re offering top of the line GPUs for industrial-strength number crunching: 

  • TESLA V100: 1x, 2x or 4x GPUs per machine 
  • Supports all major frameworks like TensorFlow, PyTorch, scikit-learn, Spark ML, Torch, Huggingface, and Keras 
  • Flexible OpenStack compute platform – Build any workflow, and use it for inference 
  • Hosted at climate-positive Leaf sites where your workloads will actually help reduce carbon emissions 

“GPUs are the SUVs of cloud-related emissions.” 

Machine learning, big data analysis, and AI are important tools in uncovering patterns, unlocking potential, and fighting waste. They help us make systems more efficient, responsible, and sustainable. Unfortunately, running these processes requires a lot of GPUs, and GPUs use lots of energy. In fact, using a regular cloud GPU for a year is equivalent to driving about 10.000 km in a gasoline-powered car.  

To stick with the car analogy: “GPUs are the SUVs of cloud-related emissions.” 

Cut emissions by 2100kg CO2 annually per GPU 

Our Leaf site solution providing hot water to households in Amsterdam can save over 2100kg CO2 annually per GPU. This translates to a reduction of over 1690 tons of CO2 annually for each Leaf site, the equivalent of 368 cars being driven around nonstop for a year.

It’s past time to cut cloud-related emissions and choose the responsible and sustainable alternatives to US-based Big Tech. Visit our website and try our truly green cloud GPUs for yourself.