FAQ
Getting Started
How do I get access to the data?
First, register on the Nightingale OS Platform.
Before granting access, we’ll ask you to provide the following:
-
Contact information for a person at an academic institution who can verify your non-commercial use case for Nightingale OS
/datasets -
Certificate of completion for a suitable training program in human research subject protections, such as the Collaborative Institutional Training Initiative (CITI) Program’s Data or Specimens Only Research course
Do I have to pay to use Nightingale OS?
Nightingale cpu.xsmall instances are free. Larger instances will require you to deposit funds in a project. If you need someone else to fund your account, that’s ok. You can generate an invoice and email it to the appropriate person.
How much money is Nightingale making from computing?
As a nonprofit, we are providing computing resources at as close to cost as we can. For researchers who need additional help, we are working with our funders to provide scholarships that ensure open, equitable access to use Nightingale data. Please tell us if you need assistance.
Computing
Can I install Python packages?
Yes. You won’t have access to the internet, but you do have access to our PyPI mirror for pip installs. You won’t have access to conda at this time.
Are TensorFlow and PyTorch set up for me?
Yes, of course. 😉
Can I use a different programming language?
Not yet. We support Python because it satisfies the largest number of users. We can support diverse computing environments it the future.
Can I use my local editor?
No, only a browser-based IDE. We do plan to support additional browser-based editors in the future.
What instance sizes are available?
CPU-Only Instances
| Size | vCPUs | Memory | USD/Hr |
|---|---|---|---|
| cpu.1x | 1 | 8 GB | Free |
| cpu.4x | 4 | 30 GB | $0.25 |
| cpu.8x | 8 | 60 GB | $0.50 |
| cpu.15x | 15 | 125 GB | $1.00 |
| cpu.30x | 30 | 250 GB | $2.00 |
| cpu.60x | 60 | 500 GB | $4.00 |
GPU Instances
| Size | GPUs | GPU Memory | vCPUs | System Memory | USD/Hr | Est. Perf Index | Est. $/Perf Index |
|---|---|---|---|---|---|---|---|
| gpu.nvidia-t4.1x | 1 | 16 GB | 7.5 | 30 GB | $0.75 | 40 | 163 |
| gpu.nvidia-a10g.1x | 1 | 24 GB | 7.5 | 30 GB | $1.21 | 132 | 333 |
| gpu.nvidia-t4.4x | 4 | 64 GB | 45 | 150 GB | $3.91 | 160 | 163 |
| gpu.nvidia-a10g.4x | 4 | 96 GB | 45 | 150 GB | $5.67 | 528 | 333 |
The GPU performance index above is drawn from public benchmarks, which can vary significantly. A value of 100 represents performance roughly equivalent to a single NVIDIA V100 GPU. Your actual performance will vary based on the specifics of your model and the dataset.
Data Use
Can I download Nightingale data and work with in it my own environment?
No, our Terms of Service specify that Nightingale data must remain inside of the Nightingale environment. You can’t download and store it anywhere else.
Can I download my own results derived from the data?
Yes. It is your responsibility to ensure whatever you remove from the Nightingale environment is consistent with the Terms of Service. We monitor all network traffic. We will investigate all suspicious activity and aggressively pursue violations to the full extent of the law.
Can I upload my own data or pre-trained models?
Yes.
Can I access resources on the internet?
No. If you need something, please contact us.
Collaboration
How do I collaborate with other people?
Everything in Nightingale OS sits within a project. You can invite other registered users to join your projects.
Your instances are private, and your $HOME directory is too. But projects are made for collaboration. Each project gets a unique storage volume that you can access at $HOME/project. Every instance in a project gets the same project directory. You decide what to share with other project members (via the project directory) and what to keep private.
Can we use GitHub, GitLab, etc.?
No. You can use the shared storage in your projects to create shared Git repositories, but we do not offer access to public repos at this time.