Our CEO, Jama Mohamed, recently had a conversation with Differential Ventures about his philosophy on the future of hardware and why he started building Sahara Cloud after spending 5 years helping hardware startups build their embedded systems as a consultant.
Below is an excerpt of the interview. Read the full version here.
Hardware Innovation Needs to Catch Up
The software ecosystem has been seeing a lot of rapid growth and innovation, and the hardware ecosystem isn’t keeping up.
There are two sides to hardware: the chips themselves and the products you create with chips. For example, in order to support machine learning these days, new chips have to be created, and we’re seeing a lot of chips with very special purposes. In that sense, hardware is innovating to support software.
But, on the other hand, if we look at the process by which people make products with these chips, not much has not changed there in the last several decades. This is where Sahara Cloud focuses—trying to find ways to do physical things more efficiently.
Take COVID, for example, we saw people trying to do things in a more agile and remote way, but it’s been hard to do that in the hardware space because, traditionally, developing hardware has been a physical process. It’s hard to virtualize because there’s a lot of custom compute involved. We’re trying to make product development as a whole faster and more efficient, and hardware is the first place we’re starting.
Why we’re Building “AWS for Hardware”
Over the past few decades, we’ve seen that hardware has gone from an esoteric, complicated, analog thing to being more digital.
These days, you don’t need the high-level, physics background to understand electrical engineering and build devices. Electronics in general is moving more towards the software space, and, as a result, a lot of the functionality that used to require custom analog circuits is being packed into smaller chips with digital interfaces.
This means you can build very application-specific devices that collect, process, and send data to the cloud really easily, and that’s where IoT growth has come from this past decade. You can make billions of simplistic devices with new chip technology, send them out into the world, and then you can use them to collect data or interface with the real world. Once you realize that the IoT can be this simpler interface, you can do a lot of things for engineering—you move from hardcore engineering to a more hybrid or digital type of engineering.
Moving from Analog to Digital electronics
Most of the value that today’s devices create is the ability to capture data points and do significant processing on the edge. Datapoint capture used to be done purely in analog, but today’s chips are more integrated, so much of the analog complexity is being packaged up such that the engineers don’t have to deal with it.
For example, rather than building out a full power management circuit, integrated battery management chips are encapsulating most of the complexity under the hood and exposing simple digital interfaces. What engineers do really have to deal with is digital stuff, such as connectivity and data models. But regardless of whether you’re dealing with a digital or analog component, however, the real problem is that the process for making and testing those physical devices is still the same—you have to be there and physically do things.
Our goal is to move both types of engineering problems fully into the digital domain without sacrificing accuracy—it becomes much easier to solve these problems when the data is in the cloud and you have the option to access remote hardware. If you’re able to leverage a true “AWS for hardware” solution in the cloud, you can complete 90% of your product before turning to the 10% of development that you can’t digitize. For that reason, we’re starting with the digital components first at Sahara Cloud, but from there we do plan to expand to support traditional, analog models, as well.
Building the Future of Hardware
The single most exciting thing for me is changing the product development paradigm. We’re at this inflection point in which companies are changing their product development processes. It has created an opportunity to rethink how engineering should be done—not “how is engineering performed now” but “what is the most efficient way to perform engineering moving forward?”
If we can correctly define that for the future, not only will companies and corporations be able to make things more efficiently, but individuals will become empowered. In 5-10 years, a non-engineer individual should be able to have an idea and build it without an engineering background and without a lab.
If you want to try a new way to develop hardware products digitally, sign up here to be notified when our private beta becomes available!