In Episode 5 of The Splashcast, Standard Fluids takes a deep dive into one of the most urgent conversations happening in AI, HPC, and data center technology: the rapid evolution of data center immersion cooling. Host Nolan Mackey is joined by two of the industry’s most experienced voices—Steve Pignato, chemical engineer and longtime innovator in fluorinated fluids, and Kelvin Cabrera, global engineered fluids specialist.
Together, they unpack the history, science, and future of immersion cooling—and why the next leap in thermal management will reshape global data infrastructure.
Immersion Cooling: A Technology Decades in the Making
While immersion cooling may feel like a modern innovation, Steve and Kelvin trace its roots back more than 50 years. IBM experimented with the first dielectric immersion bath in 1969, and Cray famously used single-phase cooling in the Cray-2 supercomputer.
The technology eventually paused when chip architectures shifted and heat loads dropped, but the rise of high-density computing, crypto, and now AI has brought it roaring back.
Today’s GPUs are generating up to 2 kW of heat each, making traditional air cooling (and even direct-to-chip systems) insufficient for what’s coming next. Data center immersion cooling isn’t a trend. It’s the next infrastructure cycle.
Single-Phase vs. Two-Phase: What’s the Difference?
The episode breaks down the engineering pros and cons of:
Single-Phase Cooling
- Uses hydrocarbons or engineered fluids
- Circulates fluid through a system
- Proven, but limited by conduction-based heat transfer
Two-Phase Cooling
- Uses a boiling/condensing cycle to remove heat
- Enables dramatically higher heat flux
- Reduces pumps and mechanical complexity
- Supports the extreme power densities of future AI chips
Steve points out that 2-phase is not new. It simply took decades for hardware and heat loads to catch up to what the chemistry could already do.
The Real PFAS Conversation: Science Over Sensationalism
One of the most valuable segments of the episode tackles the complicated topic of PFAS.
The team explains:
- PFAS is an umbrella term for thousands of chemicals
- Only a few (PFOS, PFOA) are persistent, bioaccumulative, and toxic
- Standard Fluids’ engineered fluids are not among them
- SF 649 fluid and the SF 7100 fluid series do not bioaccumulate,
- Do not partition into water,
- And break down naturally within days, not decades
This science-based clarity is essential as more companies seek Novec replacements and safe, sustainable alternatives for high-performance cooling.
Why Data Centers Are Paying Attention
As global compute demand accelerates—edge networks, AI clusters, hyperscale facilities—the industry is rapidly approaching a point where air cooling alone cannot scale.
Immersion cooling offers:
- Lower energy consumption
- Reduced water usage
- Simplified thermal architecture
- Higher rack densities
- Long-term sustainability advantages
Kelvin sums it up clearly: “Immersion cooling is set for universal adoption. It’s not a matter of if—it’s when.”
Where Standard Fluids Fits In
Standard Fluids is helping reopen the door to fluorinated fluid-based immersion cooling by providing:
- High-purity C6 ketones (SF 649 fluid)
- Hydrofluoroethers (SF 7000 fluid series)
- Extensive industry expertise
- Application support and testing guidance
- Clear science-driven communication around PFAS
- Trusted relationships with tank builders and system developers
As Nolan highlights, the conversation today isn’t just about chemistry—it’s about confidence, transparency, and partnering with customers to build the next generation of data center technology.
Final Takeaway
Episode 5 makes one thing undeniable:
Two-phase immersion cooling is the most powerful, scalable, and sustainable thermal solution for the AI era.
Between surging GPU heat loads, infrastructure limitations, and environmental pressures, the next decade of data center growth will rely heavily on the technologies discussed in this episode—and Standard Fluids is positioned at the forefront.