Cloud Credits for Science

Free cloud computing credits for research projects (AWS and/or Google Cloud Platform)

Thank you for all submitted proposal


Enzo Tagliazucchi
Cognitive Neuroscience Laboratory - University of Buenos Aires


And the winners are ...




Read more about the projects here



Germán Mato
Instituto Balseiro - CONICET

Mariano Dominguez

US $15,000 Cloud Credits for Science

If you are part of a public or private sector researchers that needs cloud infrastructure to carry out their research, you can apply to get up to $15,000 in credits for cloud computing.



  • Proposal with rationale behind request (up tp 2 pages)
  • CV of lead researcher
  • Approximate budget of resources required.
  • How many CPUs/hour or any other cloud resources needed.

December 31, 2019. The sooner, the better. We will consider all proposals submitted before that date.

The Prize

The first prize is up to US $15,000 in credits for running the researcher's workload in the cloud plus technical support. The winning team should sign a partnership agreement with Toyoko LLC so they can use the assigned infrastructure. If there are more proposals we may give another prize with a lower amount.

The Projects

Computational cognitive neuroscience of the human conscience

Enzo Tagliazucchi

Cognitive Neuroscience Laboratory - University of Buenos Aires

They will conduct high risk – high gain research projects involving the quantification of conscious states, and the landscape of their possible transformations in human subjects. This is achieved using a combination of behavioral, neuroimaging and bioinformatics data, together with computational modeling and the application of different machine learning algorithms. They will work on computational semantic and non-semantic analysis of free unconstrained language produced under different states of conscious awareness. They will perform NLP analyses of language to test if different markers linked to disorganized speech in psychosis can also manifest under the acute effects of LSD. In another work they will try to develop diagnosis and prognosis tools that can assist health care professionals in Alzheimer’s disease and other neurodegenerative conditions. Finally, the will try to predict natural language descriptions of experiences with psychoactive compounds using a dataset of three-dimensional structural data of G protein-coupled receptors with ligands from the Protein Data Bank.

Towards intelligent telescopes

Mariano Dominguez


They will focus on supervised learning and the applications of the state-of-artdeep learning approaches on optical image transformations. The aim is to reconstruct and transform astronomical images from a certain imaging system into deep learning-enhanced newimages that have e.g. improved resolution, less saturated sources, defects, masked pixels, and aberrations, statistically matching the images that would be expected from an astronomical pipeline imaging system. This specific research could lead us towards faster and simpler on the fly analysis of astronomical images using pre-trained networks that are the expected product of this proposal and could be easily put in inference chips that will conduct the reduction process in an intensive, near-real-time, high-volume computation.

Development of the new AI techniques applied to the quantification of the cardiac function and diagnosis of multiple pathologies

Germán Mato

Instituto Balseiro - CONICET

They will work on solving three different problems: 1- Analysis of cardiac nuclear magnetic resonance imaging (CMR) for the detection and automatic classification of the most relevant structures of the heart. 2- Detection of acute coronary disease in patients evaluated in the emergency department with precordial pain from electrocardiogram signals such as atrial fibrillation. 3- Quantify the longitudinal effort in an efficient way with a new methodology that will convert CMR axial images to an isotropic resolution by using Generative Adversarial Networks (GAN).

Some services available on AWS and Google Cloud Platform

Complete list of Google Cloud Platform products:

Complete list of AWS:


Who can participate?

If you are part of a public or private sector researchers that needs cloud infrastructure to carry out their research, you can apply to get up to US $15,000 in credits for cloud computing.

Can I keep the result of the research confidential?

No. You should publish the results of your research in an open access peer-review journal.

Which language should I write the proposal?

It could be in English or in Spanish. In either case, you should be able to publish the results of your work in an English language journal.

Who will evaluate the proposal?

Evaluation committee is composed of Sebastian Bassi, Virginia Gonzalez, Nicolas Palopoli and Gustavo Parisi.

Which rules and terms should I be aware of?

Amazon Web Services (AWS) and Google Cloud Platform (GCP) have their own "Terms of use" that you should read and agree. You will also be bound to an agreement with Toyoko LLC.

Should I pay something to access all these resources?

No, if you are selected you are given the access without the need to pay anything.

What if I have a small-scale project and I don't need all the credit?

Just state it in the budget how many resources you need. For example you may need just to host a small project. We may give limited prizes if we have the resources.

If I have more than one project, can I submit them all?

Yes, we will see all the proposals. Make a separate entry for each one.

I have a question that is not here, where should I ask?

Write to



Follow us

+1 510 545 4521