The underlying ideas of “deep learning,” which evolved in the 1950s, are required for self-driving automobiles and other AI-based applications. This necessitates the development of training models based on patterns found in the human brain. This, in turn, necessitates a vast amount of computational capacity, which TPUs (tensor processing units) or GPUs (graphics processing units) running for long periods of time may provide. Most AI developers, however, cannot afford this compute capacity, which they must rent from cloud computing platforms such as AWS or Azure. So, what are the options?
Gensyn, a British firm, has adopted one method. It’s taken the concept of distributed computing power from earlier initiatives like SETI@home and the COVID-19-focused Folding@home and applied it to AI developers’ quest for deep learning. As a consequence, a method for extracting high-performance computing power from a dispersed network of computers has been developed.
Eden Block, a web3 VC, led a $6.5 million seed round for Gensyn. Galaxy Digital, Maven 11, Coinfund, Hypersphere, Zee Prime, and the creators of certain blockchain technologies are also taking part in the round. This follows a $1.1 million pre-seed investment led by 7percent Ventures and Counterview Capital in 2021, with participation from Entrepreneur First and id4 Ventures.
“The rising need for hardware – and hefty margins – is why the typical names like AWS and Azure have struggled to capture such high market share,” Harry Grieve, co-founder of Gensyn, said in a statement. As a result, the market is both pricey and concentrated…. We created a better solution — one that is cheaper, has infinite scalability, and has no gatekeepers.” Gensyn plans to do this by launching a decentralized computing network for training AI models. This network use a blockchain to ensure that deep learning jobs are completed correctly, with payments made via a token. The unused computational power is subsequently monetized in a verifiable manner. It’s also a more ecologically friendly alternative, according to Gensyn, because this compute power would otherwise be wasted.
“Gensyn’s ambition of really democratizing computation using decentralized technology is possibly the most ambitious initiative we’ve come across,” said Lior Messika, managing partner of Eden Block. By dramatically decreasing the costs and friction involved with training neural networks at scale, the team hopes to favorably disrupt one of the world’s largest and fastest-growing businesses.” “Our assessment is that it’s up to 80% lower in the average pricing per unit or the sort of basic Nividia GPU, or 45 cents an hour, compared to roughly two bucks an hour for other cloud services,” Grieve said over the phone.
