We have Shunya

CHAITANYA DESHPANDE
AIoT0
Published in
5 min readNov 9, 2020

You can now run AI on your computers even if you don’t have those “will necessarily be required specs” or those GPU’s on your machines and that too at a low cost.

Imagine if those embedded devices you use, could do more than just sitting there and analyzing a process, or just gathering data or just exchanging it? What if those devices could LEARN to think from the data they gather?

Interesting isn’t it? Well, that kind of feat has been possible, BUT only by a very few. And one of them from that very few who seem to have achieved this thing is…

You heard it right, it’s — ShunyaOS

Well firstly, I would like to say that computers will be going into everything. And for COMPUTERS going into THINGS, they would need an operating system. And ShunyaOS has achieved this phenomenal endeavour. But, what is even ShunyaOS that I am talking about?

Shunya is a AI integrated OS that’s been created with the motivation that the embedded devices would not only sit there and analyze data and/or exchange it, but they would also perform AI tasks. Shunya has the ability to perform Vision and Voice tasks. It has the capability to carry out AI tasks at the edge.

What does the last sentence mean? AI at the edge means you take an AI model and put it into an embedded device. This embedded device now has it’s own brain and it can perform AI tasks that you would want it to perform. This also means that the device has it’s own storage wherein you accommodate the AI and the data it gathers and learns from.

Shunya brings AI capabilities to the edge which makes the computation pretty fast and allows for developers to use it’s APIs to create applications that would otherwise have to be written from scratch.

Let’s take a deeper look into ShunyaOS

As the purpose of AI grows, so do the resources it demands. And hence the need for ARM ML chips and NPUs.

The world is moving towards putting computers into everything out there. And in time these computers are going to get small. And this would be meaning that we would be doing complex AI tasks on these smaller computers. To pull this feat off, we would be requiring something even smaller that could do the whole computing on the edge — I’m talking about the hardware.

Let me tell you what Shunya has done on the hardware, systems and applications side:-

  1. On the hardware side — Shunya has the World’s First NPU Enabled SOC — kirin 970 — which is exclusively available with Shunya. You can read why that’s something significant about having the chip.
  2. On the Systems’ side, Shunya has:- an optimized system, an optimized kernel, a lighter AI footprint and an extensive AI library.
  3. On the Application side — Shunya has a dedicated team for application development.

Also the most important aspect for Shunya was, since it would be running embedded devices, it needs to utilize all of the computational power of the embedded device very efficiently. The reason being — embedded devices have little memory unlike a traditional computing system that I am using to write this blog. What Shunya also does is, it uses less applications — only the bare minimum required to run the whole system and not only this but, also these applications are optimized so they consume less RAM and ROM.

What does Shunya offer?

Shunya OS team helps promising startups productise their ideas

An innovation trigger has been pulled and that’s started to see rise of AI, IoT startups centered around Edge AI. While this has opened the market for opportunities for these startups, there are serious challenges they need to address, before they start building products for their customers. And according to my observation, following are the challenges and solutions that Shunya addresses:-

  1. ShunyaOS is geared towards supporting other IoT/AI hardware product startups innovate faster.
  2. Fast POC and MVP development at a lower cost — this is where Shunya has an edge over other players in the market.
  3. Low-Code/No-Code platform.
  4. Most of the vendors offer application development but, that takes them quite an amount of time — 2 months approximately. But with Shunya, it would take very less time because it allows you to use it’s built-in APIs to develop any feature/application — 10 days.
  5. At a very small footprint you can run AI even without internet, on a small embedded device.

And that’s just a few things about Shunya. I would encourage you to take a look at it and get to know more about Shunya.

What if you could get an opportunity to work on something interesting like this?

You are the average of the five people you surround yourself with

If you want to do something of significance, something that would give you importance in your career, I would encourage you to meet the maker of Shunya. This isn’t all of it, I have more for you. You’ve just met the maker of Shunya, that doesn’t end there. If you want to know more about the maker and want a shot at Shunya then here it is for you.

Is that all about Shunya? Well NO! There’s one more thing…

Shunya can put around 100,000 faces onto a small embedded computer and if you were to put 100,000 faces onto a small device, how much space do you think would Shunya take?

Not in GB’s…

Not in Mb’s…

You guessed it right, Shunya would take some Kb’s ! That’s a feat Shunya has been able to pull.

AND?

How much space of your main memory do you think Shunya would take?

GBs?

A few hundred Mb’s?

No, Shunya would require mere 14Mb of your main memory.

And?

Well, I wouldn’t spill more beans and would leave some for you to discover about Shunya. So here’s how you can know and connect with Shunya.

https://share.hsforms.com/1pxEQDdYzR6Wc4dkClmugZA48dwms

  1. Shunya:- http://shunyaos.org/
  2. Brains behind Shunya:- http://iotiot.in/
  3. Facebook Page:- https://www.facebook.com/iotiot.in/
  4. Youtube Channel:-https://www.youtube.com/channel/UC8E7r1XT_90HoPB2jWr4q9w
  5. Instagram:- https://www.instagram.com/iotiot.in/?hl=en
  6. Follow on Twitter:- https://twitter.com/iotiot_in?lang=en

#ShunyaOS #Shunya AIoT #ai on edge #aiot #iotiot.in

--

--