Next Unicorns: Unlocking the power of photonic computing with Lightmatter CEO Nick Harris | E1787

Added: Aug 3, 2023

Lightmatter has developed two types of technology: one for computation using tiny optical components, and another for communication using optical interconnects. The company's products aim to address the underutilization of hardware in AI applications, where a large percentage of power is spent on moving data between chips rather than actual computation. Lightmatter's products offer higher energy efficiency and faster data transfer rates, allowing for more efficient use of computational resources. The customers for Lightmatter's products are cloud infrastructure providers, such as OpenAI and Google, who require heavy-duty hardware for running large AI models. While Lightmatter's compute products can be compared to Nvidia's offerings, its interconnect products are more widely applicable and can be used by various chip manufacturers, including AMD, Intel, and Amazon. The impact of photonic computing on language models and data processing is significant. The ability to move data between chips more efficiently and reduce energy consumption will enable the development of larger AI models and more compute-intensive applications. However, the demand for computational resources is growing rapidly, and it is unlikely that the industry will be able to catch up in terms of supply. This may lead to a future where supercomputers become the norm, and companies like Lightmatter play a crucial role in pushing the boundaries of computational power. The conversation also touches on other topics such as the potential timeline for achieving general AI, the role of quantum computing, and the recent breakthrough in room temperature superconductivity. While Harris believes that general AI could be achieved within the next 10-20 years, he emphasizes the need for both hardware and software advancements to support its development. Quantum computing is seen as a promising technology but still requires significant investment and theoretical advancements. The recent room temperature superconductivity discovery is met with cautious optimism, as further research and replication are needed to confirm its validity. Overall, the podcast highlights the growing importance of deep tech and the need for advancements in hardware and software to support the increasing demand for computational resources. Lightmatter's work in photonic computing represents a significant step forward in addressing the limitations of traditional computing and enabling the development of more powerful AI models and applications.