Just a few years ago, most people may not have even known what a natural gas power turbine was, or what they're used for. Today, investors keeping tabs on the artificial intelligence (AI) revolution are almost certainly familiar with them, and the AI industry's lack of them.
See, gas turbines generate onsite electricity that AI data centers need, but utility companies aren't in a position to deliver. Anywhere from the size of a delivery truck to a train car, these massive machines can put out watts to power a small city, or -- obviously -- an AI data center. They just need a supply of natural gas, which is now proving easier to get than an institutional-scale hookup to a power grid.
Missed Nvidia in 2009? This Rare Signal Is Flashing Again. In 2009, a "Double Down" signal flashed for a little-known chipmaker called Nvidia. For the first time in years, that same "Total Conviction" signal is flashing for a company 1/100th the size of Nvidia. Continue »
And the AI industry is most definitely embracing the solution. Although the majority of them aren't yet operational, BloombergNEF reports that there are nearly 100 data centers with, or building, on-site natural gas turbine power infrastructure. Although they come with a higher upfront cost, owners/operators like their long-term cost-effectiveness and the self-sufficiency they enable. To this end, PwC expect the AI industry's consumption of natural gas to more than quintuple by 2035, with power turbines accounting for much of this growth.
There's just one not-so-small problem with the idea. That is, with demand greatly exceeding supply, prices of natural gas power turbines are soaring. As energy industry consulting and research firm Wood Mackenzie noted earlier this year, by the end of next year, the per-kilowatt cost of gas-powered turbines could be 195% higher than where it was in 2019.
#power #cost
See, gas turbines generate onsite electricity that AI data centers need, but utility companies aren't in a position to deliver. Anywhere from the size of a delivery truck to a train car, these massive machines can put out watts to power a small city, or -- obviously -- an AI data center. They just need a supply of natural gas, which is now proving easier to get than an institutional-scale hookup to a power grid.
Missed Nvidia in 2009? This Rare Signal Is Flashing Again. In 2009, a "Double Down" signal flashed for a little-known chipmaker called Nvidia. For the first time in years, that same "Total Conviction" signal is flashing for a company 1/100th the size of Nvidia. Continue »
And the AI industry is most definitely embracing the solution. Although the majority of them aren't yet operational, BloombergNEF reports that there are nearly 100 data centers with, or building, on-site natural gas turbine power infrastructure. Although they come with a higher upfront cost, owners/operators like their long-term cost-effectiveness and the self-sufficiency they enable. To this end, PwC expect the AI industry's consumption of natural gas to more than quintuple by 2035, with power turbines accounting for much of this growth.
There's just one not-so-small problem with the idea. That is, with demand greatly exceeding supply, prices of natural gas power turbines are soaring. As energy industry consulting and research firm Wood Mackenzie noted earlier this year, by the end of next year, the per-kilowatt cost of gas-powered turbines could be 195% higher than where it was in 2019.
#power #cost
2 months ago
