Forgot your password?
Close
typodupeerror
Power

Microsoft's $1 Billion AI Data Center Will 'Switch Off Half of Kenya' (tomshardware.com) 88

Posted by BeauHD from the power-constraints dept.
Microsoft and G42's planned $1 billion AI data center in Kenya has stalled amid disagreements over power commitments, with President William Ruto saying the country would need to "switch off half the country" to support the project at full scale. Tom's Hardware reports: The project, announced in May 2024 during Ruto's visit to Washington, was supposed to bring a geothermal-powered data center to the Olkaria region in Kenya's Rift Valley. G42 was to lead construction, with the facility running Microsoft Azure in a new East Africa cloud region. The first phase targeted 100 megawatts of capacity and was expected to be operational by this year, with a long-term goal of scaling to 1 gigawatt.

President Ruto isn't exaggerating about shutting off half the country's power. Kenya's total installed electricity capacity sits between 3,000 and 3,200 megawatts, and peak demand reached a record 2,444 megawatts in January, according to data from KenGen, the country's government-owned electricity producer. The full 1 gigawatt build would therefore have consumed roughly a third of the country's total capacity, and even the first 100 megawatts would have required a significant share of the Olkaria geothermal complex's output, which currently generates around 950MW across all its plants.

John Tanui, principal secretary at Kenya's Ministry of Information, told Bloomberg that the project hasn't been withdrawn and that talks are continuing, adding that the "scale of the data center they [Microsoft] wanted to do still requires some structuring." A separate 60-megawatt project with local developer EcoCloud is also still under discussion. [...] Microsoft is spending $190 billion on capex in 2026, and the company adds approximately 1 gigawatt of data center capacity every three months globally. But power constraints are proving to be a universal bottleneck: nearly half of planned U.S. data center builds this year have been delayed or canceled due to shortages of electrical infrastructure.

Microsoft's $1 Billion AI Data Center Will 'Switch Off Half of Kenya' More | Reply

Microsoft's $1 Billion AI Data Center Will 'Switch Off Half of Kenya'

Comments Filter:

  • Well, only 1/4 of Kenya (Score:5, Funny)

    by balaam's ass ( 678743 ) on Tuesday May 12, 2026 @03:02PM (#66140377) Journal

    since it's Azure, it'll be down half the time. ;-)

  • why not load shedding the DC and let them run off grid for some time.

    • Run off-grid?
      I assume you think solar and wind? Only if there is a decent-sized military standing guard 24/7, or you protect them somehow... otherwise, the panels and wind turbines will get taken overnight. The DC will need guard towers like a prison has.

      • Dude. Its Kenya, not Somalia. Whatever you might think of Kenya, it aint that. Any monster scale project will need security, but a handful of guys in security guard uniforms will cover it.

      • Run off-grid? I assume you think solar and wind? Only if there is a decent-sized military standing guard 24/7, or you protect them somehow... otherwise, the panels and wind turbines will get taken overnight. The DC will need guard towers like a prison has.

        Jobs for locals. Win/Win.

        • They'll replace the guy in the guard towers with motion-tracked guns, and the guys on the ground will be robots, so no jobs for locals.
          Yay! Introduce first-world problems to another country (build a DC that only lines a couple pockets and creates no new jobs), and let them deal with all the problems that come with it... oh, wait...

  • On the one hand it's a greedy western imperialist corporation, yada yada.

    But done right, upgrading the grid and capacity to support such a venture could provide economic benefits to a pejoratively 'developing' African nation.

    Who pays...

    • Re:RoI (Score:5, Insightful)

      by dfghjk ( 711126 ) on Tuesday May 12, 2026 @03:29PM (#66140425)

      "...upgrading the grid and capacity to support such a venture could provide economic benefits..."

      It could just as easily impoverish the nation. as the "upgrades" could easily be made to serve only the venture while being paid for by the locals.

      "Done right" is a matter of perspective. We know the perspective that matters, it isn't Kenya's.

    • On the one hand it's a greedy western imperialist corporation, yada yada.

      But done right, upgrading the grid and capacity to support such a venture could provide economic benefits to a pejoratively 'developing' African nation.

      Who pays...

      Typically in cases like this, the citizens of the area in question pay, while the benefits ultimately go to the corporation who is also getting massive tax breaks over other businesses in the area because of the huge positives they're theoretically developing for the area which, mysteriously, never materialize. At least, that's how we do it here in the states. I would assume that Microsoft would be expecting it to work the same there as it does here.

    • But done right, upgrading the grid and capacity to support such a venture could provide economic benefits to a pejoratively 'developing' African nation.

      How has building excess power infrastructure that sits unused ever benefitted a developing nation (or any nation for that matter). Power infrastructure must be designed and scaled to the consumption, anything in excess of that raises costs to the consumer without benefit.

      If the answer is jobs, well then you're asking the wrong question. If Microsoft wants to build a datacentre they should do what the mining industry, oil and gas industry, aluminium industry, and virtually EVERY OTHER HIGH POWER industry has

  • MS Power Plant Time? (Score:5, Insightful)

    by Gilmoure ( 18428 ) on Tuesday May 12, 2026 @03:17PM (#66140403) Journal

    Why doesn't MicroSoft just build a 2000GW power plant first?

    • The summary says that this thing is supposed to be geothermal powered.

      So they just have the cart before the horse here. They need to set up the geothermal power plant first, then build the datacenter after the power plant is operational.

      And everybody will be happy.

      • The summary says that this thing is supposed to be geothermal powered. So they just have the cart before the horse here. They need to set up the geothermal power plant first, then build the datacenter after the power plant is operational.

        The geothermal plant already exists: https://www.globalelectricity.... [globalelectricity.org]

        Apparently, Microsoft was proposing to build the data center there and tap into the existing geothermal power, not build new geothermal power (the summary was a little confusing about that).

        • Re: (Score:2)

          by Gilmoure ( 18428 )

          Maybe Microsoft should build their own geothermal source, then build a geothermal plant and then build a datacenter?

          • Why spend money, when they can just take from the rest of the area and screw them over?

            And, to Gilmore (up above): I'd love to see how big a 2000GW (2TeraWatt) power plant would be... imagine it'd just be a couple dozen nuclear reactors in parallel or some configuration.

        • Re:The geothermal plant already exists [Re:MS Pow. (Score:5, Interesting)

          by dgatwood ( 11270 ) on Tuesday May 12, 2026 @04:33PM (#66140555) Homepage Journal

          The summary says that this thing is supposed to be geothermal powered. So they just have the cart before the horse here. They need to set up the geothermal power plant first, then build the datacenter after the power plant is operational.

          The geothermal plant already exists: https://www.globalelectricity.... [globalelectricity.org]

          Apparently, Microsoft was proposing to build the data center there and tap into the existing geothermal power, not build new geothermal power (the summary was a little confusing about that).

          Yeah, that was confusing. But Kenya's president is almost certainly wrong. Here's why:

          1. It is not numerically correct, assuming the numbers in the summary are accurate. The country has a surplus adequate to power the data center at somewhere around half to three-quarters capacity even at peak power use, and probably at full capacity for 99 days out of 100. So even if they built it at full capacity right off the bat and did nothing else, you'd still only lose power to a small fraction of Kenya occasionally.

          2. They're not building it at full capacity. They're building a small data center at first, then building it up over time as more generating capacity comes online.

          3. They're a reliable customer of power. That means that they will alway pay the bill, even if it is high. The grid operators and generation plant operators can charge them a huge premium for bulk power, then use that extra revenue to build more power plants. By the time the data center is running at full capacity, they could have more than enough power to power it.

          4. Even if that extra investment in production doesn't happen, they can just refuse to provide the additional power from the grid. I'm sure Microsoft knows how to do solar + storage by now, and if not, they can pay someone to do it for them who does. Or they can build their own geothermal plant right next to the existing one. Or they can do any number of other things to produce power, like installing an SMR.

          5. Nothing inherently prevents them from reducing power usage during peak load periods. Service will get slower, but should gracefully degrade, assuming they're doing it right. Nobody will lose power, realistically speaking.

          It is unfortunate that so many people look at these data centers and the current worst-case state of resource availability and conclude wrongly that they are infeasible, but this is a common mistake made by planners, legislators, and members of the general public. They fail to account for how the existence of the data center with its need for resources will trigger the production of facilities to exploit previously unusable resources and make them available, and they fail to recognize that in a true power emergency, they can just turn 90% of it off and shift the load to other data centers.

          But the reality of the matter is that nobody is going to build a gigawatt of additional power capacity in Kenya unless the government or some private company that needs power pays them to do it. They already have a 23 to 30% surplus compared with their worst-case power consumption. That means that adding more production will just drive power prices down, so they'll get less money for the power they produce.

          But as soon as someone like Microsoft starts needing enough power to pull those margins down, suddenly additional capacity becomes economically feasible, and you'll see either existing power companies expanding or new power companies entering the market. And the existence of an all-but-guaranteed higher future demand is the key to making that happen. Without the data center being approved, that motive to expand does not exist, and the grid will likely stay at or near its currently levels unless the government forces the hand of the market by paying someone to build more generating capacity.

          • Re: (Score:2)

            by sjames ( 1099 )

            Assuming that no plant ever goes offline for any reason and that there will be no other growth in Kenya ever.

          • 3. They're a reliable customer of power. That means that they will alway pay the bill, even if it is high. The grid operators and generation plant operators can charge them a huge premium for bulk power, then use that extra revenue to build more power plants.

            I needed a good laugh, but that is exactly the opposite of how it actually works. They will be a discounted bulk price, or they'll build somewhere that will. That discount will delay the building of any new generating capacity, because the utility doesn't have the income. And while they will reliably use power, big customers generally get - because again, if they don't, they'll go somewhere that will - generous payment terms (you have to pay within 30 days of receiving at statement, they may have months, or more), and often don't live up to those.

            All of those fairly standard business practices are easier to arrange in third world countries. That's why they're building there, and not in the US.

            • Re: (Score:3)

              by dgatwood ( 11270 )

              3. They're a reliable customer of power. That means that they will alway pay the bill, even if it is high. The grid operators and generation plant operators can charge them a huge premium for bulk power, then use that extra revenue to build more power plants.

              I needed a good laugh, but that is exactly the opposite of how it actually works. They will be a discounted bulk price, or they'll build somewhere that will.

              You're describing the situation in places where demand doesn't exceed capacity. I'm describing how any sane person running an electrical utility would bill things in a place where a company wants to put in a data center that exceeds available capacity. They'll hit them with capacity charges based on their usage during peak demand periods, or they'll make them pay for capacity improvements as part of the connection charge, or both.

              And the "or they will go somewhere that will" part shouldn't really be a con

              • They want to build there because there's far less restrictions on building size and they don't have to worry about EPA regs.

          • 3) is the key point. This adds a big stable customer, which is exactly what African businesses need. Guaranteed money. This is a very good thing for Africa, even if MSFT doesn't directly build the new power plants needed.

            • African businesses = gangs and warlords, forcing kids to work in gemstone and rare earth mines

              Sure, the President or whoever gets to line their pockets a bit more, but Joe Blow down the dirt street isn't getting hired for more than sweeping the dust away from the gate. He's not sitting inside monitoring racks of GPUs all day or whatever... they'll bring in people from India.
              I don't know what you imagine this DC to be... it'll be one company that owns the racks just to run "everyone's favorite AII" on.; Cla

          • The country has a surplus adequate to power the data center at somewhere around half to three-quarters capacity even at peak power use, and probably at full capacity for 99 days out of 100. So even if they built it at full capacity right off the bat and did nothing else, you'd still only lose power to a small fraction of Kenya occasionally.
            Well, or the grid connection is not good enough where the plant is planned.
            And, or they do not have the balancing power near by, that is needed to react quickly to load c

    • Re:MS Power Plant Time? (Score:4, Insightful)

      by ArchieBunker ( 132337 ) on Tuesday May 12, 2026 @03:45PM (#66140461)

      Because it's cheaper to lobby the government and let the tax payers fund it.

    • Re: (Score:2)

      by sjames ( 1099 )

      That costs money. They were hoping the locals would foot that bill for them.

  • Ob. (Score:5, Funny)

    by greytree ( 7124971 ) on Tuesday May 12, 2026 @03:26PM (#66140417)
    640 watts should be enough for anyone. And their family.
  • I'd build a solar+battery powerplant plus a backup (natgas, oil, something that can be trucked in if necessary) for when there's a run of cloudy weather, and sell the excess back into the Kenyan grid. If the grid can cover the shortfall I'll use that, but I'd still want the backup in case of a grid outage during a storm.

    • Re: (Score:2)

      by Gilmoure ( 18428 )

      If they also build an electric vehicle factory, they can then use the vehicles for powering the datacenter.

      • Re: (Score:2)

        by taustin ( 171655 )

        They don't need to build an electric vehicle factory for that. I hear Tesla has a metric crapload of Wankpanzer's available real cheap.

    • Re: (Score:2)

      by Junta ( 36770 )

      Did you read the part where they want this to be a Gigawatt scale facility? That means you would need about 25 *square miles* of land for the solar....

      • There are several dozen solar installations in the world in the gigawatt range, so it's hardly impossible (especially in a place like Kenya that is generally sunny with a lot of open land).

        • Re: (Score:2)

          by Junta ( 36770 )

          Note that you need a significant multiplier of gigawatt scale to get gigawatt scale throughout the night.

          But either way, the point is that the magnitude wouldn't be just slapping panels on property already planned, they'd have to have a massive solar install bigger than many cities.

          • Storage would be a bigger problem than generating the power. The largest grid-scale battery storage facility in the world only stores 3.2Gwh. You'd need more like 15Gwh to reliably get through the night and early/late low-generation hours, and of course solar generation would go way down during cloudy weather.

        • Re: (Score:2)

          by pjt33 ( 739471 )

          While it's true that Kenya does have a lot of open land, there are some rather large animals which are used to that land being open, and taking city-sized chunks away puts more pressure on already declining populations.

      • Not challenging your number of 25 square miles. (However does instantly trigger the bullshit trigger in my head)

        But: are you aware how farking small 25 square miles are?

        Hint: it is a square of 5 miles versus 5 miles.

        So, lets see what my BS trigger tells us: 25 square miles is the equivalent of 6,475e+7 square meters.
        Why do I use meters? Oh, because the standard panels are roughly 1m wide and 1.5m long, and yield 650Wp (electric peak).
        So 1.5 sq43166666666.7m are roughly 1.3kWp - obviously the yield of such a

        • Re: (Score:2)

          by Junta ( 36770 )

          And double it to get through the night.... I was calculating based on kwh per day of expected solar against kwh of consumption for a gigawatt (so... 24gwh).

          It wasn't a random ass guess, I did the math.

          5 miles by 5 miles is a huge installation. Far from the suggestion that they could just slap some panels down on their facility and even have surplus for the grid..

          • I was calculating based on kwh per day of expected solar against kwh of consumption for a gigawatt (so... 24gwh).
            Then explain the math ... as I don't get that sentence.
            Why would you need 24GWh solar capacity when you only need 1GWh?

            5 miles by 5 miles is a huge installation.
            I guess this is certainly in the eye of the beholder.

            If you look at a random wheat field in a random part of USA, it is most certainly bigger.

    • But... then they'll be polluting!
      And, we're talking GW... how big does the solar field and warehouse-sized battery backup have to be to keep it working 24/7?
      If the grid can cover the shortfall?? The grid is basically lamp cord and wire nuts (they might even have a ground wire). This ain't NYC we're talking about, here.

  • Never mind the power... (Score:4, Interesting)

    by YuppieScum ( 1096 ) on Tuesday May 12, 2026 @03:42PM (#66140451) Journal

    ...what about the cooling? Equatorial Africa is not exactly known for its temperate climate.

    Worse, half the population of Kenya don't have access to clean drinking water, and low rainfall for the last year means that many areas are in drought conditions.

    It's long past time that some decision-makers got a good beating with the clue-stick.

    • So, they'll just use phase-change cooling, and just use even more power... they don't care... gotta have "Clod"!

    • Worse, half the population of Kenya don't have access to clean drinking water

      You don't need access to water for cooling. The question is only if Microsoft is a responsible party here, building a datacentre with closed loop cooling instead of evaporative air cooling. It's perfectly viable, but ultimately boils down to an economic decision.

      I'm going to take a guess that building, owning and operating a datacentre in a 3rd world nation is cheap compared to a western nation so there is more opex budget available to pick a cooling solution that allows a non-evaporatively cooled datacentr

  • Let them build them, but they must build the power for it first - built and paid for by themselves before the data center itself can break ground.

    Leaches and damned leaches, they even stole my cheap RAM.

    • Re: (Score:2)

      by taustin ( 171655 )

      And subject that power generation to the exact same regulations, environmental and otherwise, as any other. Do that, and we won't be seeing many new datacenters for the next 20 years.

      • Re: (Score:2)

        by PPH ( 736903 )

        Well, before AI, data centers were just another customer of the local utility. So, the "exact same regulations, environmental and otherwise" was sort of a given. That starts to break down when the DCs build and operate their own power plants.

        The whole "Aunt Millie in upstate New York buys her power from a New Mexico solar plant" is sort of bullshit. When she turns her lights on, that power comes from Indian Point.

  • I am getting SO tired of hearing the stories about the excesses of AI. OpenAI wants all the water and all the power and all the data centers and all the capital markets (ALL of them, ALL $5 TRILLION) in investments so that they can make a thing that's going to cost every man, woman, and child on this planet $550-ish a month AT A MINIMUM, just so they can break even. Microsoft wants all the things. nVidia wants all the things. Apple just wants to borrow all the things from someone else who knows how to

  • As of May 12 2026 Microsoft's market capitalization is just over 3 trillion USD. The World Bank estimate for Kenya's GNI (Gross National Income) in 2024 was 123 billion USD. [tradingeconomics.com]

    The value of the data center alone is about 0.81% of the GNI of the entire country. The value of the data center is about 0.033% of the value of Microsoft, so the ratio of Microsoft/Kenya relative value is 2454. Note this is not a percentage, but the approximate ratio of net worth. As a percentage it's 0.04% USD of Kenya per 1 USD of Mi

  • Sorry, this availability zone is down due to: Lions

    https://weebls-stuff.com/toons... [weebls-stuff.com]

  • Microsoft is a bad actor for datacentre builds (Score:4, Informative)

    by thegarbz ( 1787294 ) on Wednesday May 13, 2026 @11:00AM (#66141521)

    MS has a history of using lawyers and their influence to go against the spirit of being a good economic contributor, and commercial partner with countries and communities. If not the spirit, then in some cases they are intentionally abusing wording in laws to violate their spirit*. I have no doubt that they will fuck over the people of Kenya to get their way.

    * Example from the datacentre in Amsterdam: The local laws forbid the construction of a new hyperscale datacentre, but unfortunately they assumed that all datacentres are designed the way that ... well all datacentres are designed and defined hyperscale datacentres in terms of power consumption and land area use (note: not floor space). So along comes Microsoft and says, ... well if we build our hyperscale datacentre 20 floors high then we can fit it in an area under 8 acres and as such are exempt from hyperscale datacentre bans. *Lawyers celebrate*, *Microsoft profits*, *local communities get fucked*.

Slashdot Top Deals

It is not every question that deserves an answer. -- Publilius Syrus

Close