Great support, good availability, and seamless integration capabilities
What is our primary use case?
I am primarily using Saturn Cloud as a student, primarily for training deep reinforcement learning (RL) agents. My projects usually involve Python 3.10, CUDA, and PyTorch and they typically require heavy computation power like GPUs and multi-core CPUs.
Saturn Cloud's environment has been able to sufficiently cater to these needs, providing a platform where I can run and manage these resource-intensive tasks easily. The support has been very helpful in clearing up questions, such as creating a custom image.
How has it helped my organization?
Before using Saturn Cloud, I was relying on my laptop for running my deep RL models. The training process was time-consuming and it was difficult to multitask. However, with Saturn Cloud, I can offload this heavy computing to the cloud. This not only accelerated my research process but also enabled me to multitask efficiently.
It didn't take long to see that Saturn Cloud could scale with my needs, providing more resources when required. This scalability is invaluable for a student researcher like myself where tasks can vary widely in computational requirements.
What is most valuable?
One of the features I appreciate the most about Saturn Cloud is its seamless integration with Jupyter notebooks. It provides an interface that I am familiar with and use extensively. It makes it easy for me to run, track, and debug my RL models.
The second feature I like is the ability to easily scale up and down the resources (like GPUs and CPUs), providing a flexible and cost-effective solution.
The third feature that is useful is the availability of pre-configured environments. It saves a lot of time and hassle, especially when working with complex setups involving packages like CUDA and PyTorch.
What needs improvement?
While Saturn Cloud offers a rich set of features, there are areas where I feel there could be improvements. For instance, the process of setting up custom environments could be more user-friendly. Currently, it is a bit technical and might be daunting for beginners.
Additionally, providing more detailed and beginner-friendly documentation, especially for advanced features, could greatly enhance the user experience. So far, I was relying on the support for setting up custom images and understanding why things don't work properly.
For how long have I used the solution?
I've been using the solution for one month.
What do I think about the stability of the solution?
What do I think about the scalability of the solution?
The solution is very scalable; machines can go up to extreme configurations.
How are customer service and support?
Support is very friendly, proficient, and helpful. They typically reply within hours.
How would you rate customer service and support?
Which solution did I use previously and why did I switch?
Before using Saturn Cloud, I was primarily running models on my laptop and occasionally using Colab. However, Colab was not feasible with the captchas and deleting the runtimes after a few minutes of being away from the keyboard.
As my models became more complex and required more computational power, I found that this was not scaling well with my needs. I switched to Saturn Cloud primarily for its superior computational power, scalability, and ease of use. In the beginning, the free 30 hours of computing convinced me to try it.
How was the initial setup?
The custom image was a bit more tricky and I required support. That said, it was solved within the same day.
What about the implementation team?
The initial setup was handled in-house.
What was our ROI?
I don't have an ROI; this is an academic project I am working on.
What's my experience with pricing, setup cost, and licensing?
In terms of pricing, they should make it as transparent as it is here.
Which other solutions did I evaluate?
I did not evaluate other options.
its easy to work with kind of platform
What do you like best about the product?
easy to learn the working steps, I find every required tool in one place which makes me like this platform even more. this platform is user-friendly.
What do you dislike about the product?
still to find one which i am uncomfortable with.
What problems is the product solving and how is that benefiting you?
its best platform to work on if you are working in machine learning field.
Great Resource for "upping" your research
What do you like best about the product?
User friendly-good information-lots of examples and visuals
What do you dislike about the product?
It can be a bit busy, but the information is helpful
What problems is the product solving and how is that benefiting you?
It is giving me another source of information that helps in my understanding.
Good JupyterLab experience
What do you like best about the product?
Good conda support with built-in mamba for faster package resolution. Setting up a new project is quick and easy, with various docker images available and support for SSH connections.
What do you dislike about the product?
The 30h per month isn't a lot considering other free alternatives, but none of them support stock JupyterLab so that is acceptable, although more hours would be welcome.
What problems is the product solving and how is that benefiting you?
Running computation locally is costly and requires a considerable amount of man-hours. Saturn Cloud lets us quickly iterate over our ML tasks without worrying about maintaining computing infrastructure on-premises.
Very helpful support, easy to use, transparent prices, and great compute resources
What do you like best about the product?
I love that you can build your own images, spin up pcs of your own configuration. Additionally, the platform is amazingly easy to use.
The support answers quickly - I had my problems solved within 1-2 hours always.
What do you dislike about the product?
I had to build my own image for python 3.10, but that was no problem at all, with the help of the support.
The default shut-down timer is at 1 hour, which I didnt see at first so watch out for that.
What problems is the product solving and how is that benefiting you?
Training Reinforcement Learning algorithms on GPUs for my Master's thesis. Solely using CPU is not feasible in my case and my own laptop is not able to do this computation.
Best service for fast scaling and training for teams!
What do you like best about the product?
The part I love the most - how easy it was to configure everything and start training. Intuitive UX and nice docs!
What do you dislike about the product?
Nothing yet. I have not met anything that I dislike
What problems is the product solving and how is that benefiting you?
Scaling of training and optimization processes
One of the best data science platforms
What do you like best about the product?
available to use always.
data never gets deleted after server shutdown.
ssh connection
What do you dislike about the product?
Nothing yet, it's a very friendly and easy to use environment
What problems is the product solving and how is that benefiting you?
limitations and electric consumption of local PC
The solution provides prebuilt images that allow us to quickly spin up a readymade Python environment
What is our primary use case?
We use Saturn Cloud to perform data analysis on large volumes of data. Saturn fetches and updates data, We can use it for machine learning training and prediction, and perform experimental work on various data using multiple machine learning techniques. In some cases, parallel computation is also required to perform the analysis as quickly as possible.
The environment has eight CPU cores and 64 GB RAM. In some cases, we are using GPU. The development environment includes Python, Scikit Learn, XGBoost, Jupyter Lab, and Jupyter Notebook.
How has it helped my organization?
Saturn Cloud provides prebuilt images that allow us to quickly spin up a readymade Python environment, speeding up the setup of the environment required for data science projects.
Saturn Cloud also provides interactive development environments, such as Jupyter Lab and Jupyter Notebook, enabling faster code editing. Saturn's Cloud environment offers high availability and reliability, improving the efficiency of our development work.
What is most valuable?
Saturn Cloud supports GPU as part of the environment, which is essential for many computational tasks in machine learning projects. It also allows us to edit the environment, including the image, before we start the cloud resources. This feature lets us quickly set up the environment without the hassle of moving the data and code to another cloud device.
The solution supports Jupyter Lab and Jupyter Notebook environments as direct options. We can simultaneously work with Saturn Cloud in both modes, enabling faster interaction and sophisticated development support.
What needs improvement?
Saturn Cloud should include prebuilt images for advanced data science packages like LightGBM in the next release. If possible, they should also provide a Kaggle image, which contains the most common Python packages used in machine learning.
I would like the option to check or uncheck the data science subpackages we need in the environment. Usage reporting should be more precise and quantify use in the number of minutes instead of just hours.
For how long have I used the solution?
I have used Saturn Cloud for more than a year.
What do I think about the stability of the solution?
Saturn Cloud provides reliable and available resources most of the time.
What do I think about the scalability of the solution?
The computation and memory can be easily scaled from the current requirement.
Which solution did I use previously and why did I switch?
Our previous solution lacked fast interactive environments like Jupyter Lab and Jupyter Notebook.
What was our ROI?
Which other solutions did I evaluate?
We also considered Paperspace Gradient and Kaggle.
What other advice do I have?
I rate Saturn Cloud 10 out of 10.
Compare to Google colab Saturn cloud is much worth with amazing computational power
What do you like best about the product?
Great thing about saturn cloud is , its amazing computational power for big data , machine learning .
I used both google colab Pro and Saturn cloud free hours and winner is Saturn cloud .
1) No need to reinstall on each kernal start
2) Very quick project running
3) High RAM access for big data analysis
4) GPU is powerful
What do you dislike about the product?
If saturn cloud provided , data folder sharing in free hours , it was amazing .
But even in free we can download data .
So thanks to saturn cloud for amazing service
What problems is the product solving and how is that benefiting you?
Saturn cloud providing high computing power with GPU access . This will help in solving and learning some advance Machine learning and Bioinformatics tasks free for students.
very good cloud experience
What do you like best about the product?
cloud environment experience and good computing power
What do you dislike about the product?
may be pricing more expensive, may lost some files from storage randomly
What problems is the product solving and how is that benefiting you?
delete file from my workspace randomly
