New AI Supercharger: How Doudna Will Power Tomorrow’s Discoveries

The U.S. Department of Energy is set to unveil Doudna, a next-generation supercomputer named after Nobel laureate Jennifer Doudna, that will turbocharge both artificial intelligence and scientific research. Slated for deployment at Berkeley Lab’s National Energy Research Scientific Computing Center in 2026, Doudna will harness cutting-edge Nvidia “Vera Rubin” accelerator chips in Dell’s liquid-cooled servers—offering massive compute for fields from genomics to climate modeling and even national-security simulations.

What Doudna Brings to the Table

Scale and Speed: Housing thousands of Nvidia AI accelerators, Doudna will deliver petaflops of AI-optimized throughput, letting researchers train and run advanced machine-learning models at unprecedented scale.
Broad Research Impact: From mapping genetic interactions and designing new materials to simulating nuclear stockpile integrity, the system will support over 11,000 scientists across 17 national labs.
Energy-Efficient Design: Dell’s liquid-cooling and power-management tech aim to tame Doudna’s energy appetite—critical as AI workloads surge in size and complexity.

Energy Secretary Chris Wright says Doudna exemplifies America’s commitment to “uniting AI prowess with scientific rigor,” ensuring U.S. leadership in both technological innovation and critical research infrastructure.

Frequently Asked Questions (FAQs)

Q1: Why name the supercomputer Doudna?
A1: It honors Jennifer Doudna’s Nobel Prize–winning CRISPR work, reflecting the machine’s role in accelerating breakthroughs in genetics and beyond.

Q2: How does Doudna differ from previous DOE supercomputers?
A2: Unlike past systems optimized purely for simulation, Doudna is co-designed for AI and high-performance computing—balancing traditional modeling with large-language model training and inference.

Q3: Who will use Doudna and for what purposes?
A3: Over 11,000 researchers—spanning genomics, chemistry, physics, climate science, and national security—will access Doudna through DOE’s user facilities, tackling problems that demand both AI and extreme compute power.

Comparison: Doudna vs. Apple’s AI Chip Ambitions

Both Doudna and Apple’s upcoming custom AI server chips target the same challenge: smashing performance barriers for next-gen AI. Where Apple is weaving its M3 Ultra and A14X accelerators into consumer devices and private clouds, Doudna concentrates GPU-scale AI firepower in a public research hub—showing how bespoke hardware underpins breakthroughs, whether at your desk or at a national lab.