Forbes
Summary
Operators are rethinking every aspect of their data centers to ensure they maintain optimal performance levels, including developing new cooling systems and ensuring access to power. But the challenge, as always, is finding equilibrium. AI demands raw computational muscle, but what goes into producing it can't be an afterthought. How data centers push boundaries without faltering will determine just how far AI can go.
The Power Predicament
Power is the ultimate bottleneck for data centers.
While a standard Google search draws minimal electricity, a single ChatGPT query requires 10 times that amount, according to Goldman Sachs. That difference explains why data center power demand is projected to skyrocket 160% by 2030. But, as we see with recent innovations like DeepSeek, we are uncovering different ways to tackle efficiency.
Continue Reading
#News#DataCenter