Nvidia CEO Jensen Huang introduces the next-generation Rubin graphics processing unit and BlueField-4 data processing unit during the Nvidia CES 2026 Live event at Fontainebleau Las Vegas in Nevada, Jan. 5. Yonhap
Startup representatives researching biotechnology say they have transitioned from "biotech" to "techbio." While it sounds like a simple play on words, the shift highlights a critical reality for the industry: Biological research is now entirely dependent on heavy hardware.
Biological data operates on an unprecedented scale. Analyzing biological tissue measuring just 2 centimeters by 2 centimeters generates data in the petabytes (PB), equivalent to one quadrillion bytes. Consuming 1 PB is equivalent to watching a 1-gigabyte high-definition movie continuously for 13 years, or filling 1,024 1-terabyte hard drives.
The sheer volume creates a severe logistical problem; the CEO of one biotechnology startup has said that transferring several PB of data to a cloud center via the internet takes four days.
For urgent international collaborations, researchers face a frustrating reality. They must physically pack storage drives into their luggage and board airplanes. In the digital age, this physical transport — often called a "sneakernet" — remains faster than internet transmission for the life sciences field.
To resolve this bottleneck, biotechnology companies urgently need data processing units (DPUs).
A DPU is a specialized semiconductor designed to compress and rapidly transmit massive datasets. Central processing units and memory chips have traditionally handled these tasks, but the biological data load has far surpassed their capacity. Just as artificial intelligence (AI) developers rely on graphics processing units (GPUs) to train models, a biotech startup's success now depends on its ability to secure DPUs.
Consequently, biotechnology firms are utilizing venture capital to aggressively hoard DPUs and GPUs. Some startups are even meeting with semiconductor manufacturers to commission custom-designed chips. In a landmark move reflecting this trend, Illumina, the world’s largest genetic analysis equipment company, acquired semiconductor design firm DRAGEN for $100 million.
However, this necessary evolution into "techbio" is currently colliding with a global supply chain crisis. The survival of these companies now depends on acquiring specialized semiconductors in a market entirely monopolized by the AI boom. With governments and major tech firms aggressively securing chips for AI development, biotechnology companies are struggling to procure the hardware they desperately need. As biological datasets continue to grow exponentially, this semiconductor shortage could ultimately determine the fate of the techbio sector.
The writer heads the Startuplab section of the Hankook Ilbo, where he covers the IT and startup sectors. This column from the Hankook Ilbo, the sister publication of The Korea Times, is translated by a generative AI system and edited by The Korea Times.
Source: Korea Times News
