Add info about ReckOn chip

biphasic commented 6 months ago

The website

The code for the live version of open-neuromorphic.org is hosted on the main branch at our GitHub repository. The site contains educational content on neuromorphic software libraries and hardware.

What to improve

The current entry for ReckOn is located here and is missing most info.

Following info should be added in the header:

[ ] applications. What applications can the chip be used for.
[ ] release_date. When was the chip released? That means made available for customers to buy/access for industry chips, or the first results published for research chips.
[ ] weight_bits. What are the possible bit resolutions of weights?
[ ] activation_bits. What are the possible bit resolutions of activations?
[ ] neurons. How many neurons fit on one chip? If it depends on the neuron model, provide an example for the most common neuron model. This field can take a (short) sentence if need be.
[ ] power. How much power does the chip typically burn.
[ ] synapses. How many synapses for a given neuron config? This field can take a (short) sentence if need be.
[ ] on_chip_learning. Does the chip support on chip learning?
[ ] release_date and release_year.
[ ] summary. Provide a summary that's roughly 300 characters long.

Following info should be added in the text body.

[ ] What makes the chip stand out in neuromorphic computing?
[ ] Practical applications of the chip
[ ] What's the workflow to develop and deploy models for the chip? What are the software toolchains involved?
[ ] Availability to purchase / access
[ ] Scalibility - Can I stack this chip easily?
[ ] Form factors. Are there boards available that can go on small drones? Robots? Data centers?

Following things would be nice to have:

[ ] Links to documents / data sheets that describe the chip.
[ ] Technology and fabrication details.
[ ] Future development roadmap.
[ ] List of publications that show results on the hardware, including date, title, authors and journal/conference.

Feel free to add any relevant information you think would be beneficial. If you have questions or need clarification on any topic, please open a PR and ask. We encourage both seasoned contributors and newcomers to participate. Our team is ready to assist you in this process, both on Github or on Discord. Please keep the discussions public at all times.

Workflow

Comment on this issue to express interest, so that our team can assign it to you. That'll let other people know that you're working on it.
Fork this repository
Make the changes in your fork and commit them on a new branch.
Follow the instructions to build the website locally and check that the new page renders nicely.
Create a pull request from the new branch in your fork into main in the original repository.

Writing style

Our aim is to create a central hub for educational content on neuromorphic computing. The content should be formally written, accessible to the public, and include references or images where appropriate. Feel free to include yourself in an authors statement at the bottom of the page. Please avoid overly general comments (as sometimes generated by ChatGPT).

SahilKonjarla commented 6 months ago

Hi, I am willing to work on this issue!

biphasic commented 5 months ago

done in 312112e56ea9b34c73c918faf02e1054480b6814

SahilKonjarla commented 5 months ago

Hi, Gregor I think you might have closed the wrong issue by accident. The issue you added was the ODIN chip while I am working on the ReckOn Chip, unless I am mistaken

biphasic commented 5 months ago

Hey @SahilKonjarla ! I added info for both chips in the commit above. Please open a PR on top of the latest version of https://github.com/open-neuromorphic/open-neuromorphic.github.io/blob/main/content/english/neuromorphic-computing/hardware/reckon-frenkel/index.md if you have additional info!

open-neuromorphic / open-neuromorphic.github.io