Time flies. It's been more than a year, from the time we launched. Thanks to all our early adopters for trusting and supporting us. Without your love, feedback and patience we would have not come this far.
In the last few months, we have been working on some key features
- Managing instance life cycle through Python API
- Bring your own container
- Start up script
- Spot instances
- Weekly and Monthly prices
- Live invoice
- New website, docs and blog page
Jarvislabs.ai API
Now you can automate the entire instance life cycle events using python API.
- Launch a GPU/CPU powered instance
- Pause
- Resume
- Destroy
- Add/Update/Delete startup scripts
Example: Launch a GPU powered instance
instance = Instance.create(gpu_type='A100',
num_gpus=1,
hdd=20,
framework_id=0,
name='IamAI',
script_id=1)
You can find more details in the docs.
Bring your own container - BYOC
Now you can bring in any publicly available container, just share the name and choose the configuration like GPU type, quantity and storage. Using BYOC you can choose
- Framework
- Cuda libraries
Learn more about BYOC here.
Spot instances
Now you can rent any spare GPU from jarvislabs.ai for a discounted price and save upto 60% on your costs.
When the demand of reserved instance increases, the spot instances will be auto paused immediately. So it is recommended that you can use spot instances for quick experimentation or handle checkpoints in your code.
It is also super easy to launch a spot instance, just a toggle away.
| GPU | Spot price |
|---|---|
| A100 - 40GB | $0.99/hr |
| A6000 - 48GB | $0.79/hr |
| A5000 - 24GB | $0.59/hr |
| RTX6000 - 24GB | $0.39/hr |
| RTX5000 - 16GB | $0.19/hr |
Weekly and monthly prices
For sometime now, we have been offering discounts for users who use GPU powered instance for long duration like weeks and months. With this new release, we have automated it. Now you can easily avail weekly and monthly discounts in few clicks. You can check the discounted prices here.
Saving environment
This is a favourite feature for our users and us, which lets you use the same environment from where you left. But the implementation was causing delays while resuming and other challenges. In the last few months, we tried our best to solve it, but could not come with a solution. So we are removing it, and users can use startup scripts for installing any additional libraries on the start of an instance.
We love this feature so much, we will invest our time to build a more robust solution in the coming months.
New website, docs and blogs
Check out our new website, blogs and docs. Hope you will like it.
Thank you
I am so grateful to everyone of you for your continued support and feedback. We are getting started and the path will be determined by your needs. If you have any suggestions or feedback, we are all ears. Please reach to us.