← All stories
● Covered by 1 source · 1 reportMedium impact1 neutral

Frore Systems Claims LiquidJet Can Reduce Nvidia Rubin GPU Temperatures by 10°C

🔄 Updated 1d ago
New to BrevFeed? We gather this story from every outlet covering it into one summary — ranked by real-world impact, not just the latest headline — so you never miss what matters. What is BrevFeed? →

Key points

  • Frore Systems' LiquidJet technology can reduce Rubin GPU temperatures by 10-12°C.
  • A 10°C temperature drop could boost token generation by 15%.
  • Improved cooling directly translates to higher sustained performance for AI hardware.
  • Modern AI accelerators like Nvidia Rubin can dissipate up to 2,400 W.

Cooling for AI Accelerators

Frore Systems, a manufacturer of cooling solutions, has presented findings on reducing temperatures in next-generation AI accelerators. The company's white paper suggests that improvements across the entire cooling stack, including GPU packaging, thermal interface materials (TIMs), coldplates, and coolant temperatures, can increase token generation per watt by over 30%.

LiquidJet Technology's Impact

Based on an analytical thermal model, Frore Systems projects its LiquidJet coldplate technology can lower Nvidia Rubin GPU junction temperatures by up to 12°C. This temperature reduction is estimated to translate into a 10% to 25% improvement in tokens per Watt, with a 10°C reduction specifically linked to a 15% increase in token generation.

The Challenge of High Temperatures

Modern AI accelerators, such as the upcoming Nvidia Rubin, can dissipate up to 2,400 W, leading to die temperatures that can exceed 95°C. While 95°C may not immediately impact silicon longevity, high temperatures significantly increase leakage current, which doubles approximately every 10°C. This also reduces transistor switching efficiency.

Performance Implications

Elevated temperatures necessitate higher voltages to maintain clock speeds, which often triggers Dynamic Voltage and Frequency Scaling (DVFS) to reduce clocks to stay within thermal limits. This ultimately leads to decreased performance and lower token output. Therefore, better cooling directly correlates with higher performance and token generation in AI data centers.

✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →

The daily brief

One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.

One email a day. Unsubscribe in one click, any time.

Today's brief

Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.

~7 min · 6 stories · Aug 15

▶ Play today's brief Listen on Spotify

New every morning, and the back catalogue is archived by date.

Reporting from

Frore Systems published a white paper suggesting its LiquidJet coldplate technology can lower Nvidia Rubin GPU junction temperatures by up to 12°C, potentially increasing token generation by 15% with a 10°C reduction. This development addresses the critical need for improved cooling in high-power AI accelerators to maximize performance and efficiency in data centers.