Share to FacebookShare to TwitterShare to Linkedin
Running AI models can be incredibly costly, with a single Nvidia AI chip priced at over \(30,000. The operational expenses for ChatGPT at OpenAI are estimated to be a staggering \)700,000 per day. These exorbitant costs often restrict AI development and deployment to large corporations with substantial financial resources. However, SqueezeBits, a young startup based in South Korea, aims to address this issue by reducing expenses and democratizing access to this cutting-edge technology. Recently, SqueezeBits secured backing from a prominent internet company in the country to support its mission.
In a recent funding round in January, SqueezeBits raised 2.5 billion won (approximately \(2 million) in pre-Series A funding, although the startup has not disclosed its valuation. Sources familiar with the matter suggest that this round values SqueezeBits at around \)15 million. Among the investors in this round are Kakao Ventures, the venture capital division of Kakao, a major player in South Korea’s internet industry.
Justin Shin, a senior associate at Kakao Ventures, highlighted the importance of democratizing AI applications, emphasizing the significance of broadening access to such technology. Kakao Ventures has also previously supported Rebellions, a company in South Korea developing more cost-effective AI chips compared to Nvidia’s offerings. This strategic investment approach aligns with the goal of fostering innovation and accessibility in the AI sector.
SqueezeBits had previously secured 1 billion won in seed funding in 2022 from D2 Startup Factory, backed by Naver, another internet giant in South Korea. Overall, the startup has accumulated 3.5 billion won in venture funding to date, reflecting growing interest and confidence in its approach to optimizing AI efficiency.
Hyungjun Kim, the cofounder and CEO of SqueezeBits, emphasized the need to streamline AI models by eliminating redundant parameters and data. By enhancing efficiency and reducing unnecessary computational costs, SqueezeBits aims to make AI models faster and more cost-effective. Kim’s expertise in electrical engineering and computer science from Postech underpins the startup’s innovative strategies for optimizing AI performance.
While SqueezeBits faces competition from other companies specializing in AI model optimization, such as OmniML and Xnor.ai, its unique approach sets it apart in the evolving landscape of AI technology. By focusing on enhancing AI efficiency and performance, SqueezeBits aims to bridge the gap between rapid algorithmic advancements and hardware limitations, offering tailored solutions to meet the evolving demands of the industry.
