The issue of electrical energy production remains a top discussion point when considering all the factors involved in the AI Race that’s not just between the Outlaw US Empire and China as many think as Russia and India are both in the competition although omitted from artistic representations like the one above. Here’s a report published on 7 December at Guancha:
Driven by companies such as Nvidia, the AI boom is rapidly driving demand for data centers, but aging infrastructure and slow construction in the United States are dragging down the country’s data center construction. Bloomberg published an article on November 10 saying that in Santa Clara, California, where Nvidia’s headquarters are located, two data centers have been vacant for several years due to [the lack of] local power supply.
Jordan Sadler, a spokesman for Digital Realty, the developer of one of the data centers, said waiting three years for energy is “normal lead time in most of the United States,” and even longer in high-demand areas such as Silicon Valley and northern Virginia.
“In Santa Clara, if you find a piece of land today and start applying for new power, you may have to wait years,” Sadler said.”
Sadler’s “may have” ought to be “will have,” and that is primarily due to the lack of long-term planning by private energy firms like PG&E in California’s case that place profits over all else, including maintenance for which PG&E has lost many millions in lawsuit fines because its willful neglect caused recent devastating fires—nothing was learned from the contrived “Energy Crisis” in the late 1990s that showcased Enron and its “accounting” but also implicated PG&Es unwillingness to build more capacity and upgrade transmission lines, the latter being the cause of the fires. But California isn’t the only state with electrical power issues. Texas has major problems as does Virginia and most of the East Coast, New York in particular as was proven this past Summer. The reluctance to build is because of the profit motive of the mostly financial managers of energy companies, the engineers, as with Boeing, having been displaced from their Boards—as demand goes up for the limited supply on hand, the pricing also rises which boosts profits, which is shunted into the pockets of shareholders instead of building more capacity and upgrading transmission lines to meet the rising demand. Plus, within the Outlaw US Empire, there’s zero central planning when it comes to energy generation—it’s all state or regions. Washington can proclaim the need for more electrical energy generating capacity through the Department of Energy but has few means to do anything about it nowadays.
The nations remaining in the AI Race—China, Russia and India (Europe’s no longer a contender thanks to its amazingly idiotic energy policy and Biden’s destruction of the Nord Stream pipeline)—have governments that do long-term planning and have centralized Ministries that manage unitary systems, so they can plan and enact policies capable of providing for future energy demand. From the article. here’s Nvidia’s Huang:
“China has twice as much energy as the United States.” He said that the U.S. government is promoting the reshoring of manufacturing, “but without energy, how do we build chip factories, supercomputer factories and AI data centers?”
Although not associated with power generation, Huang makes this very important point:
Huang said that the United States still maintains a lead of about “six months” in AI cutting-edge models, but China is far ahead in open source models, “Without open source, startups cannot prosper, university researchers are difficult to conduct research, and scientists cannot use AI. Without open source, almost all industries in the economy cannot achieve fundamental progress. The different technologies that allow AI to thrive are open source, and China is ahead of the United States in open source.”
Warwick Powell has written some very valuable essays on this topic and technology in general. From his early November essay:
For large-scale models, energy costs, cooling overhead and total system throughput matter just as much; often more than peak chip efficiency. Here, the U.S. faces a structural disadvantage. Industrial-scale electricity prices in the United States are rising, currently averaging around $0.12 per kWh, while in China, large industrial regions serving AI clusters pay as little as $0.04 per kWh.
And this is from his excellent early October essay that deals exclusively with the power issue:
The United States faces deep structural bottlenecks in energy infrastructure development, ranging from regulatory delays and transmission constraints to capital costs, skilled labour shortages and technological readiness. Even in the absence of international complications, these domestic hurdles make a five-year timeline for 100 GW of firm power capacity effectively unachievable. The implications are brutal and stark. Curtailed and costly electricity supply for AI and manufacturing will impair American economic competitiveness, with knock-on effects for household affordability. These impacts are already becoming evident, with wholesale pool prices in the U.S. rising by 267% over the past 5 years, on the back of skyrocketing electricity demand from the AI sector.
The linked source’s data is alarming and shows that in the 25+ years since the Enron energy scandal nothing’s been done to improve the situation. And the cooling issue that requires vast amounts of water is another “domestic hurdle” that will generate all sorts of NIMBY protests.
Neoliberalism didn’t just hollow-out Outlaw US Empire industrial capacity; it’s focus on short-term profits at the expense of long-term planning has gravely wounded the Empire’s ability to compete as the AI Race lengthens into a marathon. And we see the billionaires hoarding their wealth instead of investing in more productive capacity and the energy systems to run them. Ultimately, they are proving to be worthless.
Those who read my report on Russia’s Energy Week annual forum will have learned what Rosatom has planned and is currently constructing, and not just in Russia—China and India, too. And then there’s a new energy storage technology—sodium ion batteries—that greatly enhances green energy capabilities. Both China and Russia still have hydropower resources to harness that the Empire doesn’t. The Empire could build more wind, but Trump killed its subsidy program. China’s pursuit of fusion is very intensive and it’s not just state-driven with the number of start-ups being part of China’s strategy of increasing the pace of innovation through competition, a policy that’s been implemented within all technological realms. When China formally releases its next 5-Year Plan early next year, many will be looking at its energy sector plans.
India has its own series of future energy production expansion plans that are centrally managed by the energy division of NITI Aayog which appear to mimic China’s path since it plans to expand 90 GW of coal-fired energy production by 2032 while also ambitiously building up its green energy sector, which is already quite large. Now that India has decided which geopolitical bloc it’s going to favor, it should continue its 8-9% annual development pace, and that will require energy.
I’ll let Warwick Powell make this closing statement:
The 100 GW target thus reflects an underlying recognition that America’s existing energy system cannot support the dual ambitions of powering AI and reviving manufacturing. But recognition of the problem should not be confused with feasibility. Even under optimistic assumptions, the scale of new firm generation required would be difficult to marshal in a decade. To attempt it in half that time demands not just an unprecedented construction boom, but also the resolution of entrenched structural bottlenecks that have long constrained U.S. energy development.
The lack of critical long-term planning is the Empire’s weakest point that it seems ideologically capable of improving. There’s no continental plan for the continental sized empire that says it wants to remain a leader but displays an incapacity to do so. The one major player constantly and publicly trying to improve the situation is Nvidia’s Jensen Huang. IMO, he needs to direct his critiques at the many tech billionaires, since there’s very little the federal government can do to improve the energy situation and keep American firms in the AI Race.
*
*
*
Like what you’ve been reading at Karlof1’s Substack? Then please consider subscribing and choosing to make a monthly/yearly pledge to enable my efforts in this challenging realm. Thank You!
First of all, let me say that A eye is A Lie, they should call it that.
Now, re the greed of gazillionairs in scAmerica, the great American way is to enrich oneself at the cost of others , most others, for the benefit of an increasingly vanishing few....
Tis a business model of the more corrupt the merrier. The more who suffer the more American it tis.
When folks die homeless on the streets in the cold while there are vacant shelters and spaces all around the city sitting empty, that is considered a stunning and smashing success. The fewer who can live reasonably, the more most suffer , the greater is America declared to be....
I've yet to hear any sound evidence that rapid AI advancement is good or bad for the Great Unwashed who now have an additional competitor for energy security, billions of microchips, all with the growing healthy appetite of any young farm labourer. lol
Sooner rather than later some important people at the top will have to decide on the merits of switching off the incubators or feeding the chip.
The rounding up of Undesirables is likely part of Big Tech's memo to Trump. The Undesirable pool will grow exponentially to include YOU.