惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

酷 壳 – CoolShell
酷 壳 – CoolShell
Microsoft Security Blog
Microsoft Security Blog
Recent Announcements
Recent Announcements
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
Last Week in AI
Last Week in AI
罗磊的独立博客
腾讯CDC
云风的 BLOG
云风的 BLOG
月光博客
月光博客
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
博客园 - 三生石上(FineUI控件)
宝玉的分享
宝玉的分享
U
Unit 42
I
InfoQ
D
DataBreaches.Net
Blog — PlanetScale
Blog — PlanetScale
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
V
V2EX
美团技术团队
IT之家
IT之家
Stack Overflow Blog
Stack Overflow Blog
F
Fortinet All Blogs
GbyAI
GbyAI
S
SegmentFault 最新的问题

Forbes - Innovation

Why Do Humans Have Fingerprints? Hint: It’s Not What You Think Booking.com Confirms Data Breach, Reservation PIN Codes Changed Why Major News Sites Are Blocking The Internet Archive’s Wayback Machine iPhone Fold Release Date: New Report Details Frustrating Apple News Comet Tracker: How To See Pan-STARRS And Three Planets On Wednesday NYT Mini Crossword Today: Tuesday, April 14 Hints And Answers Today’s NYT Strands Hints, Spangram, Answers: Tuesday, April 14 (It’s A Little Unclear) Today’s Wordle #1760 Hints And Answer For Tuesday, April 14 Most Of The Microplastics In Urban Air Come From Tires Today’s Wordle #1759 Hints And Answer For Monday, April 13 NYT Mini Crossword Today: Monday, April 13 Hints And Answers NYT Pips Today: Hints, Answers And Walkthrough For Monday, April 13 The YC Chief Who Codes 10,000 Lines A Day Has A Simple Secret Samsung Expands One UI 8.5 Beta To More Galaxy Owners Why You Should Stop Using Your iPhone If It’s On This List Chamath Says Firms That Treat AI As A Strategy Hand Rivals Their Edge 3 Unexpected Habits Of Secure Couples, By A Psychologist The First Lamp That Folds Your Clothes Samsung’s Disappointing Price Update For Galaxy Phone Buyers 3 Subtle Signs Someone Is Falling In Love With You, By A Psychologist Do Mantis Shrimp See More Colors Than Humans? A Biologist Explains NYT Connections Answers Explained For Monday, April 13 (#1,037) NYT Connections Hints Today: Monday, April 13 Clues And Answers (#1,037) LEGO Luigi & Mach 8 (72050) Review: 2026’s Best Set Yet? Marc Andreessen Says AI Productivity Will Trigger A Hiring Boom 3D Printing Is The Ultimate Hack To Reduce Household Spending Apple iPhone Fold: Striking Design Revealed In Leaked Photos Apple Smart Glasses: New Leak Reveals A Major Design Twist To Beat Meta Tested: The AI Coming To The Rivian R2 Quordle Hints Today: Monday, April 13 Clues And Answers
Startup Raises $27 Million To Solve Two Massive Data Cent...
John Koetsier · 2026-06-15 · via Forbes - Innovation
Data centers are one of the most controversial new builds today. They're also constrained by access to power.

Data centers are one of the most controversial new builds today. They're also constrained by access to power.

AFP via Getty Images

Green energy infrastructure startup TAR has just raised a $27 million seed round to provide scalable on-site plug-and-play power infrastructure to data centers with a heavy emphasis on renewable power, Forbes has learned in an exclusive. Given the current state of the public debate around data centers, the technology can’t come soon enough.

“We are building modular, scalable, behind-the-meter energy systems consisting of solar, batteries, wind energy and (for emergencies / long periods of unfavorable weather conditions) simple cycle gas turbines,” co-founder Pat Becker told me via email. “The main premise is being able to provide power 24/7 without relying on the grid at low cost, minimal environmental impact and extremely fast.”

That’s a potential step forward on a number of different levels.

The first is power supply for AI providers. At the recent NY Tech Week, prediction startup Principle ran a simulation of the enterprise AI market. In 110 of 123 possible futures, lack of power was the main constraint on growth: not lacking the right chips, not failing to develop the right models … just lack of power. Global demand for power is set to grow 50% faster than the previous decade, says the International Energy Agency, and China’s green energy revolution has allowed the country to add 543 gigawatts of power capacity – more than the total capacity added by the U.S. in its entire history – in 2024 alone, according to the IEEE.

That’s an issue for global competitiveness in AI.

But at the same time, data centers are becoming one of the most controversial new builds, and for good reason. They can be incredibly noisy – a data center in Dowagiac, Michigan emits a constant and irritating 60-decibel whine in neighboring homeowners’ yards – and that destroys not just property values but quality of life. Perhaps just as bad, data centers are also stealing power from communities. 50,000 Lake Tahoe residents just have to simply find a new power source after NV Energy decided to sell the power to data centers instead.

None of that is good for people. And yet, we all individually are using more and more AI. Companies are fueling demand much more, even firing people to fund their AI addictions.

Texas-based TAR might have a solution for at least some of those challenges.

It starts with providing on-site power so data centers don’t have to take it from residential communities. And that ability to provide onsite power means data centers aren’t as dependent on public grid access, so hopefully they can be built farther away from homes.

I had the opportunity to interview Becker on how TAR might be able to be a part of the solution: providing power that doesn’t rely on already-committed grid power, and making data centers at least somewhat more environmentally friendly.

TAR founders Leonhard Soenke and Patrice Becker

TAR

John Koetsier: You’re focusing on green solutions as much as possible?

Becker: We’re focusing on deployment speed and independent power generation, as little reliance on energy prices, gas prices or similar. Renewable energy works great for this from an energy security perspective and it’s great for the environment too.

John Koetsier: What’s new/different about your model?

Becker: For each power generation method we are using, conventional methods rely on outdated methods and a lot of onsite work. We focus on shifting work to factories where we can pre-wire, pre-assemble, pre-test and pre-commission energy generation methods significantly faster and cheaper. This means that each power generation method we are using requires changes to the deployment sequence.

John Koetsier: How much power will your pilot deliver? How does that compare to planned systems?

Becker: Our pilot site (on our own land) will deliver 10 MW of constant power (24/7). The nameplate capacity of the system is a lot bigger than 10 MW (due to fluctuations in solar and wind capacity throughout the year). We built extensive modeling systems to simulate capacity needs at various planned uptimes. Our first customer deployment will have a 24/7 constant power capacity of about twice that.

TAR will provide off-grid power for data centers, largely relying on green sources.

TAR

John Koetsier: Who are your customers?

Becker: The customer for our first deployment is a large neo cloud. We cannot disclose the name.

John Koetsier: How quickly can you get a site operational? Today versus when at scale?

Becker: Our system reduces onsite work to the absolute minimum and we don’t have to wait to be able to connect to the grid as we never tie into it. We are delivering a 20MW constant IT load system within 3 months. The system is modular and repeatable, we are deploying more than 200MW of constant load in 2027 and several gigawatts of constant load in 2028.

John Koetsier: What’s the strategic value here for the neocloud provider?

Becker: Building behind-the-meter energy systems that don’t rely on gas turbines (which are increasingly hard to get) is an incredibly difficult task - power generation via renewable energy requires securing power electronics, battery and energy management, extensive modeling as well as re-inventing how energy is deployed on-site. There is no one capable of building 24/7 energy systems fast enough. Time-to-token is the measure that dominates the industry now and will continue to dominate, so we will be the fastest on the market.

John Koetsier: How do the economics work out? Better than grid?

Becker: We are not cheaper than the grid, but we are much faster. TAR’s energy system is not affected by interconnection queues, curtailment, lengthy permitting cycles, or slow, labor-heavy on-site work. The pricing on the energy will depend on project to project since varying requirements affect how much system overbuild and thus CAPEX is required. It is not uncommon to see behind-the-meter PPA prices as high as $150-$160 per MWh. We are far below that.

John Koetsier: Does this help with the politics? Data centers aren’t popular these days.

Becker: I believe that there are valid concerns about data centers, pollution and impact on local communities. If you rely on building compute close to the energy grid, chances are you are closer to civilization and the impact on communities is larger. If a data center can be powered off-the-grid, far away from civilization in arid land, closed loop cooling and minimal pollution, the conversation around data centers would change significantly.

John Koetsier: How often do you expect the gas turbines to run?

Becker: This strongly depends on the uptime you are targeting and is a cost trade-off - you can get along with no gas turbines at all but the amount of overbuild needed is significantly higher than with (minimal) gas turbines. We expect to only need gas turbines for a few weeks per year if at all.

John Koetsier: What regions will you target? Will it depend on green energy availability?

Becker: We target sparsely populated areas with inexpensive land throughout the US. Since we don’t need a grid connection, a lot more locations are available to us to develop our system.

John Koetsier: Thank you for your time.