FPGA Computing Platform for machine learning in data analytics

Why does SnowBell offer such great potential for companies with machine-learning applications?

SnowBell is the first FPGA software that enables users to develop their machine learning application once and then deploy it on any FPGA. Nowadays, many applications require considerable computation and calculations within milliseconds. A SnowBell-based system is ten times faster than standard big computers; it consumes less energy, is easy to manage, and takes up less space. Ultimately, it makes any application more cost-efficient. Many companies lack the skillsets or budgets to develop their own FPGA applications. Using SnowBell, businesses will be able to seamlessly integrate FPGAs into their systems and scale up their services and data processing capabilities much faster and with minimal investment costs.

A broad range of services and products require ultra-rapid solutions, including security in online credit-card payments, fraud detection, IoT (Internet of Things), predictive maintenance, autonomous assistance/driving, as well as flight search engines or online remarketing and sales recommendations. 

What were the challenges?

One of the main challenges was recruiting hardware engineers with the required skillset - the task is as rare as it is complex, resulting in a very small talent pool. Another major challenge is teaching companies about the FPGA technology and its potential business benefit for them. As FPGAs are an unconventional technology for most companies, it takes time for businesses to gain trust and for business values to materialize. SnowBell’s goal is to facilitate the use of the FPGA technology for companies, and to lower the investment barrier so that it becomes attractive and its long-term benefits can be seen.

What are your goals for 2020?

I intend to found my company in the first half of 2020. I am in contact with potential customers and am excited to launch some pilot projects in the business areas of online recommendation systems, fraud-detection systems, and predictive maintenance.

In the meantime, I am further exploring, discussing, and learning about machine-learning companies’ interfaces and how to “marry” the systems. This is important not only from a SnowBell product development perspective, but also for those companies’ future performances, adaptability, and competence in a rapidly growing and changing environment.

Besides, I am recruiting people to build a team of software engineers, business development specialists, and marketers. I am also seeking investors. I am happy to see that the response thus far from VC has been positive.