Blackstone wants to plug hyperscale datacenter into former Britishvolt battery site US private equity investor Blackstone has plans for a £10 billion ($12.45 billion) hyperscale datacenter in northern England on a site formerly owned by battery startup Britishvolt. Blackstone, which styles itself as the world's largest alternative asset manager, is looking to acquire the site near Blyth in Northumberland in …
There are a significant number of wind turbines in the sea around Blyth.
Blyth is billed as a centre of excellence for wind power.
The site used to be Blyth power station so should be easy to have significant grid access.
And yes, it's not in Blyth but Cambios. (Pronounced Camis)
Yes, they will want to tap into existing infrastructure as it will be cheaper for them…
They could however:
Lay their own interconnect
Build a zero carbon power generator on the site…
I bet their plans don’t include solar panels, wind turbines, heat recovery or reuse (market gardening)…
There is something not right about the assumption that a private developer can build a facility that will have a significant power draw and they expect the public/taxpayer to pay the capital costs of building the necessary generating capacity. I expect they will also assume someone else (taxpayers) to fund their coolant/water supply and waste water processing…
Genuine question, how do VM hypervisors allocate, schedule and prioritise GPU resources? Since a 750MW bitbarn sounds like it will involve a lot of "AI" crap ..
At best, I suppose when the wind stops blowing, your GPU app could be unloaded and replaced by someone else's GPU app, but if someone else was offering to pay more than you, then the GPU would already be running their app and not yours, regardless of the power source. The bit barn won't want to have idle silicon -ever-, since it is so expensive.
Guessing they may be working on the assumption that SMRs become a thing. But even then it'd need a few for both resilience and simply provide 750MW. If it's near the coast though, cooling should be more doable, if they can overcome objections to the 'n' word. But Blackstone has been very good at buying influence.
Why's that?
Also, could a large-scale (2GW) plant not be built out of several ~300MWt reactors? It would still mean the reactors are smaller and easier to build/maintain, while sharing security and waste handling on the one site
Do you know how long Hinkley C took to plan/agree, and the build is still going on, now expected to complete in 2027, although they're talking about another delay to 2029-31 now.
And the cost is now projected to be £46bn, which dwarfs the investment Blackstone expects to put into the datacentre.
Personally, I think EDF is taking the UK government for a ride, with the ever greater projected costs. I'm sure that they are financially benefiting somewhere, and will probably expect the UK government to pick up the pieces when it eventually proves unprofitable because it's "too big to fail".
I am pretty certain that the UK government bought the Chinese share out a while back.
And I believe that the Sizewell contract that the Chinese wanted 100% of, plus the ability to provide their own labour and project management is never going to be awarded to them.
Plus the fact that there are two new reactors being built, not just one.
Water moderated nuclear is stupidly expensive to build and even more so to actually run thanks to the high maintenance costs on the non-nuclear side resulting from the technology's inability to provide dry steam
I'm beginning to suspect that SINAP's TMSR series will reach production models before Hinckley Point actually sells any joules - and if those work/scale according to predictions they'll be 1/5 or less of the cost of Hinckley Point to build/operate whilst producing 98% less nuclear waste - and thanks to not needing stupidly large containment vessels/buildings, they can be built in less time than the associated steam turbine installations (in fact, MSRs can act as drop-in heat source replacements at coal plants, which may turn out to be the perfect bridging technology needed to maintain energy production)
For a massive improvement even after a meltdown / vastly enhanced criticality, I suggest siting it in Cumbernauld, though had they used some foresight, instead of building houses on the old British steel site in Motherwell they should have built several atomic reactors, given the already present grid connections to power a steel mill
Hinkley C - our only new Nuke plant actually under construction was started in 2018 and is now not scheduled to be operational until 2027. It will add 3.2GW of capacity.
None of the other UK "new nuke" plants have even broken ground.
In *1 year* between 2021 and 2022 the UK added 4GW of renewables. So in the same timeline you could add 40+GW if you put your mind to it.
Before anyone says baseload. its a myth. No-one has even even put a number on what Baseload needs to be, and it can be addressed by renewables just like mass cheap storage does - by extensive overprovisioning.
'Before anyone says baseload. its a myth. No-one has even even put a number on what Baseload needs to be'
Curious how you state categorically that Baseload is a myth and can be addressed by renewables when you openly claim that it hasn't or can't be quantified.
Even massive over provisioning still wouldn't cope, a completely impractical and nonsense idea.
Really?
Show your working that its impractical.
(A simple google will show you baseload is a concept with no objective reality.)
The *worst* renewable generation day last year in Feb produced 33% of demand. Ergo 3-5 times the current capacity (most of which has been added in the last 10 years) is achievable between now and 2050.
Yes really....
I don't need to show my working, I'm not the one making clearly contradictory statements.
There is a whole global electricity market based on the concept of base and transitory load and I spent 15 years of my career within the generation industry so I have practical experience and first hand knowledge of power generation on a national scale as opposed to reading and believing the opinion of some obscure Looney Tune with an axe to grind on Google.
A reliable power generation scheme requires dispatchable and controllable generation capability A. To power the grid itself ( switching etc) B. To meet essential power load, and C. to manage and control grid frequency, all these represent the minimal base load requirements.
There's nothing mythical about that...
"So in the same timeline you could add 40+GW if you put your mind to it."
Except the principle of "low hanging fruit" kicks in
This is already happening around the world. For a standout example see the Tararua windfarms in New Zealand. The first one was a spectacular sucess thanks to the particular local geography acting as a wind funnel but the investor pile-on into building more windfarms along the same mountain range has resulted in massive losses and at least one of the companies going bankrupt
That would take actual forward planning and an acceptance that an all-electric decarbonised future requires 6-8 times more total annual generation than currently exists
This is something that greenwashers, NIMBYs and subsidy farmers won't tolerate and I'm sure they'll find a way of virtue-signalling the inevitable rolling brownouts which at the current trajectory are coming in the 2030s
Nah we are heading for the future so the "dog" will be a weaponised variant of something from Boston Dynamics, several roaming with both visible spectrum vision and IR and armed with either hydraulic jaws and razor sharp tungsten carbide teeth or an m249....the guard being a touchscreen and parcel drop box
As part of Mr Lee’s good neighbor policy, all Rat Things are programmed never to break the sound barrier in a populated area. But Fido’s in too much of a hurry to worry about the good neighbor policy. Jack the sound barrier. Bring the noise.
Well, yeah, it was obviously going to need more than two security dogs and a guard regardless(!)
But that marginal exaggeration- and the joke as a whole- wasn't really about security. It was that, despite the hype, most data centres tend to end up providing relatively few jobs locally- and even fewer high-quality ones- for something that large and disruptive.
And that even if it matches the abandoned factory proposal in terms of scale, it's going to be a piss-poor replacement that's unlikely to provide more than a small fraction of the former's originally-promised 3000 jobs.
Can't run VMs without breaking a few icebergs?!
Sorry but the planet is doomed, it's a fact as saving it has no financial incentive for the great and good. Personally I do recycle, cycle or walk place and keep my energy usage down but i know that 2 doors up my neighbours, who are builders with white vans, always come out at 7am and start their deisel engines and then bugger off indoors for 10 mins most days leaving their vans to pump crap!
The best we can hope for is that all the tech investment will yield a way to get future generations off this dying ball of rock and out into space.
proliferation of all these bitbarns seems suspicious to me. I remember previously people set up a computer company and invested heavily and just sat there with one or two customers till they went broke.
I imagine now they set up and just start mining bitcoins.
Harmless? No as we keep being told about rationing and energy shortages. get on it sunak. Do one thing at least with that chess club brain.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
