The OCRE ProjectThe Open Clouds for Research Environments project (OCRE) aims to accelerate cloud adoption in the European research community. It does so by bringing together cloud providers, Earth Observation (EO) organizations, and the research and education community. Cloud-based services can offer the European research community a wealth of powerful tools. Unfortunately, for many researchers, these are currently out of reach, with suitable services difficult to find and select. The OCRE project addresses this need through ready-to-use service agreements and €9.5 million in adoption funding. As a recipient of an OCRE IaaS+ framework agreement, CloudSigma is one of the providers that will facilitate easing the access to commercial cloud services for researchers and research institutions in 40 European countries.
The OCRE Cloud Framework and OCRE Cloud Catalogue support state-of-the-art research and explore new ways forward for procuring, funding, and using in-demand digital solutions while benefiting from volume discounts. You can find CloudSigma’s contact details and services here.
One of the research initiatives that was awarded a CloudSigma voucher through OCRE’s initial voucher program is TensorCell, an independent research group founded in 2017 by Paweł Gora. The research group develops algorithms for optimizing complex processes such as road traffic and cancer treatment. Despite the fact that such processes are from different domains, they (unexpectedly) share many interesting properties and can be tackled using similar techniques.
The research group uses mostly methods based on AI and multiagent simulations based on cellular automata, hence the name TensorCell. In addition, the researchers train machine learning models to approximate the outcomes of the simulations in order to accelerate computations even faster.
The TensorCell research group has received computational grants from different cloud providers in order to run their experiments aiming to develop AI and machine learning algorithms. This allowed them to gain a very good understanding of what they need from a cloud provider to make their work process as optimized as possible.
One of the issues the researchers were facing while working with bigger cloud providers, was the amount of time it took to get a response whenever they had doubts or issues. As a result, their experiments would be delayed, which ultimately slowed down their research findings. Since the computations for this specific research are extremely time-consuming, it was crucial to find a provider that would help the scientists speed up the process.
Another challenge for the TensorCell scientists was that they were not experts in cloud technology, and as such their knowledge about working with cloud infrastructure is mostly self-taught and intuitive. Thus, they found hyperscale cloud providers to be complicated to operate. Тhat’s because larger cloud providers offered too many settings, options, and procedures to set up in order to create virtual machines which further made the process cumbersome and slow.
“I have had experiences with many cloud providers, and for me, the interfaces of large cloud providers are very complex, I would say. So, there are really many options and I guess that hyperscalers are rather a solution for more advanced users probably.” -Paweł Gora, Founder @ TensorCell
Furthermore, in the case of research projects, the cost of the services is another important factor. As a research group, TensorCell does not necessarily have the budget that other cloud providers require for the use of cloud resources. Therefore, they were looking for a cost-efficient solution that could still fulfill their requirements for high performance and scalability.
When learning about the OCRE voucher program, the researchers from TensorCell jumped at the opportunity to receive cloud vouchers that would help them run their experiments. The amount of computations the scientists needed to perform to be able to run their experiments is really huge, so having free cloud resources at hand proved to be essential.
“I was contacted by Manuel Delgado Blasco who is an Earth Observation Expert Engineer working for the OCRE project. It was a pleasure for me to be offered cloud vouchers. After a series of evaluation interviews, we started using CloudSigma for running experiments with particle swarm optimization and genetic algorithms.” -Paweł Gora, Founder @ TensorCell
Having CloudSigma’s cloud vouchers at their disposal made TensorCell’s work much more efficient. As the researchers have had experience with several different cloud providers, they were immediately able to appreciate our user-friendly Web App and interface.
“I have worked with Google, AWS, and Azure before, and I can pinpoint one area that quickly differentiates CloudSigma. CloudSigma’s intuitive graphical user interface makes it extremely easy to just set up a virtual machine and start running experiments. Compared to the other providers I have used, CloudSigma’s user interface is more intuitive and the configuration of machines is much quicker.”Paweł Gora, Founder @ TensorCell
Initially, the researchers encountered some issues with logging in to the Web App as a default user. However, after using CloudSigma’s technical support Live Chat feature, they were able to resolve the problem in under one hour.
More than that, CloudSigma has useful documentation and information available on our website and upon request that would solve a lot of the issues customers might be having.
“After having one of my issues resolved, I realized that all the information I needed was also available on the website, but I didn’t check it at that time. Nevertheless, the online real-time technical support proved to be very useful to speed up finding a solution.” -Paweł Gora, Founder @ TensorCell
“CloudSigma’s technical support was extremely helpful for us. Asking a question is very simple- you just need to use the chat function on the website, and you get an immediate response. In the case of other providers I had to submit a new request, then receive a ticket, and after that had to wait for an answer. I would not be surprised if the quick and helpful technical support is the biggest reason that companies are coming to CloudSigma.”Paweł Gora, Founder @ TensorCellAs a research group, TensorCell was looking for a provider that would help them perform their experiments faster, and thanks to CloudSigma they were able to do just that. In addition, having cloud vouchers enabled them to achieve quick results, without the burden of having to pay the high cost of cloud resources.
The implementation procedure was extremely easy and straightforward for TensorCell. First, the team had to make sure that they had the OCRE voucher installed successfully, which was a very simple process.
“The only thing I had to do was register for the voucher, and the resources were available soon after the registration. I needed some help at first, but I quickly got my questions answered.” -Paweł Gora, Founder @ TensorCell
The next step was to migrate their system to CloudSigma’s platform. This was done easily by preparing a Docker file and then configuring it on a CloudSigma virtual machine.
“Migrating the Docker file to CloudSigma’s platform was straightforward. I remember that we had technical issues with some Python libraries at the beginning, but that turned out to be a problem with the Docker file. The whole process took not more than one hour.” -Paweł Gora, Founder @ TensorCell
In addition, the migration was facilitated as much as possible by CloudSigma’s friendly team of account managers and free 24/7 technical support. Our skilled support team was able to quickly solve some of the initial issues the team was facing, thus ensuring the implementation is smooth and efficient.
Ever since receiving OCRE vouchers for CloudSigma resources, the research team has been able to accelerate the pace of their experiments.
“Thanks to the availability of cloud resources, we were able to run our experiments much quicker. The number of computations that we need to conduct in order to generate a training set, or just acquire all the required results, for scientific publication, it’s really huge. So, we needed computational resources. Thanks to using CloudSigma’s virtual machines, we were able to run the required computations much faster. So, for sure it accelerated our experiments. It helped us get good results much earlier.”Paweł Gora, Founder @ TensorCell
In addition, the team was able to take advantage of a stable and high performing cloud platform.
“Another thing was that access to the computational power was very good. And I was also able to just give access to virtual machines to other team members from my research group. This would not have been possible if we used our own machines.” -Paweł Gora, Founder @ TensorCell
Having CloudSigma as a cloud provider enabled TensorCell to conduct their experiments with more agility and flexibility. As an added bonus, the team found the platform especially suitable for research projects such as theirs because the scientists were able to benefit from working with a simple and intuitive user interface and free 24/7 online technical support.
“Researchers don’t necessarily have to be experienced IT engineers. So, having a simple and intuitive web browser-based GUI, and user-friendly Web App, with technical support available to help you is a huge advantage for researchers who are not necessarily computer scientists by education.” -Paweł Gora, Founder @ TensorCell
In addition, using CloudSigma’s resources allowed the team to be much faster when setting up virtual machines. As a result, they were able to perform their experiments more efficiently, with quicker and improved outcomes.
“With CloudSigma, it was much easier for me to set up the parameters of a new machine. I was able to keep many options at their default value. Thus, it took me one minute or even less to create a virtual machine.”Paweł Gora, Founder @ TensorCell
То make it easier for clients and users who don’t have in-depth knowledge with cloud solutions, CloudSigma’s virtual machines are already pre-configured with default settings. If needed, each client can further customize their machines to their liking, but the initial setup is as user-friendly as possible.
As for the future, there is no doubt that TensorCell researchers will continue using CloudSigma. They are currently planning their new batch of experiments, and they will again need cloud resources. They are planning to apply for another CloudSigma voucher that would allow them to conduct their research with more flexibility.
Other future plans for TensorCell include their goal to establish a startup company, for which they have no doubt that they will use the services of CloudSigma.
- Using Asynchronous Primary-Secondary Replication by PostgreSQL for Auto-Clustering - January 27, 2022
- Exploring CloudSigma PaaS: How to Utilize Ruby PaaS Hosting Services? - January 10, 2022
- Improving Availability and Site Performance Using a Replica Set with MongoDB Auto-Clustering - December 22, 2021
- Excellent Technical Support and Intuitive User Interface Make CloudSigma the Perfect Cloud for Research Groups - December 8, 2021
- CloudSigma’s High-Performance and Cost-Effective Cloud Solution Enables a BIaaS Provider to Better Serve its Customers - November 26, 2021