Provcence in southern France, known for its picturesque villas and vineyards, is also home to a vast building site - about the size of 60 soccer pitches.
Before work began in 2007, the dead-flat 42ha platform at Cadarache, about an hour's drive from the French Riviera, was forested. Now it is set to become one of the biggest nuclear research facilities in the world.
In all, says my guide Topher White, 90ha of woods have been cleared, something he suggests has been accomplished with a certain ruthless efficiency.
I don't mention to White that New Zealand has had its own brush with French determination to push ahead with nuclear experiments, when its secret service blew up the Rainbow Warrior in Auckland harbour.
The Cadarache site, adjoining an existing nuclear research centre, will house the world's largest fusion reactor, built with the backing of the European Union, the United States, Russia, China, India, South Korea and Japan.
Called Iter (International Thermonuclear Experimental Reactor), the project has been decades in the planning, is costing tens of billions of euros and, it is hoped, will prove fusion's feasibility as a clean source of electricity.
White is an American, helping confirm the international nature of the undertaking. He is an astrophysics graduate, concerned with the physics of the universe, a fitting background for a project that aims to mimic on Earth the reaction that makes the stars shine.
He is part of a communications team that hosts more than 10,000 site visitors a year. Three years into a five-year contract, unless he signs on for another stint, he won't even be around when the reactor's on switch is thrown for the first time.
Yes, he concedes, results from the Iter experiment remain a frustratingly long way off. However, the future of world energy supplies hinges on the outcome.
Other countries might be reluctant to play host to nuclear facilities, but not France. About 80 per cent of its electricity comes from nearly 60 reactors, after it threw itself into the nuclear industry's arms in response to the 1973 oil crisis.
France beat Japan, also firmly in nuclear power's embrace - it has more than 50 reactors - for the right to host the Iter facility. As Japan struggled to contain the damage at its Fukushima Daiichi nuclear power plant after last March's huge earthquake and tsunami, scientists from Cadarache travelled to the country to help out.
That says a couple of things about nuclear research: when a reactor mishap occurs, things can go catastrophically wrong, endangering millions of people and doing the industry long-term harm; and nuclear physics is a collaborative field, with high stakes in both cost and potential payback.
Iter, however, is different from the scores of French and Japanese fission reactors that produce electricity by splitting uranium atoms. Fission plants use hundreds of kilograms of fuel, produce large amounts of radioactive waste and their reactions can go out of control, releasing harmful radioactivity into the environment.
Iter, White says, is intended to be the first demonstration of a fusion reactor - one that produces more energy than is put in to trigger the reaction.
Hydrogen, helium, lithium ... anyone who can name even the first three elements in the periodic table is acquainted with the essential ingredients for nuclear fusion.
Iter's reactor will consume just grams of deuterium and tritium fuel (both forms of hydrogen), releasing energy and producing harmless helium as a byproduct.
Fusion does have its hazards, White acknowledges, but they are minor compared with fission.
Iter's reactor will become mildly radioactive, but for a fraction of the thousands of years that fission waste takes to decay. And a fusion reactor cannot melt down - the big fear at Fukushima - because the reaction simply stops when it is no longer controlled.
Fusion has been the Holy Grail of nuclear physicists and the energy industry for more than half a century.
Gazing over the Cadarache site, however, and hearing White recite the extraordinary facts and figures of the project, makes clear that Iter is a serious attempt at cracking the problem.
WHEN France was declared the experimental reactor's host in 2005, it launched into the project by spending €110 million on 100km of roads to transport reactor components from Fos-sur-Mer on the Mediterranean to Cadarache. The heaviest loads will weigh 900 tonnes and take as much as three days to make the trip.
Since completion of the 1km by 400m site in 2009, which involved shifting 2.5 million cubic metres of earth, a 250m-long magnetic coil winding facility has been built, a headquarters building for 500 people is under construction and a 1.5m-thick concrete pad for the reactor to sit on has been completed.
The level of engineering is extreme, but so is the mass of equipment that will occupy the site, which happens to be in an area of "moderate" seismic activity, near to a fault that caused France's worst recorded earthquake, a magnitude-6 jolt a century ago. The weight of the reactor and its complex of buildings will amount to about 360,000 tonnes.
The doughnut-shaped reactor itself, or tokamak, will weigh 23,000 tonnes, several times heavier than the Eiffel Tower. Its components will be built by the 34 Iter countries, the whole thing is supposed to be place by 2018, and it should be ready for switching on two years later.
If the deadline is met - and there has already been slippage - that will be a mere 35 years after the project was first proposed.
Mikhail Gorbachev, head of the Soviet Union, is credited with putting the suggestion for an international fusion project to his US counterpart, Ronald Reagan, at a 1985 superpower summit.
It was the ultimate expression of the Cold War's end. The fusion reactor, which will operate at a temperature 10 times hotter than the sun, will be a peaceful application of the same technology that for decades cast a shadow over world affairs.
"It took since then to get all of the countries on board," White says.
Not least of the problems has been getting financial commitments for a project with a construction price tag of about €13 billion ($20 billion). Willingness to contribute has proved closely related to fuel prices.
"When fuel prices are low, there is less political drive for the project."
Funding issues have already meant a four-year delay to the moment of switching on - known as "first plasma" - which will be the initial one of three phases of the experiment.
Money and political will are minor matters, however, compared with the technical challenges that lie ahead. Iter is attempting to do what no other fusion reactor has done: produce more energy than it consumes.
The best any previous experimental reactor has managed is to return about two-thirds of the energy required to trigger a fusion reaction. That record was achieved at Jet (Joint European Torus), a reactor at Oxford, in England, 15 years ago.
Jet, as Iter will be, is a tokamak, or doughnut-shaped vacuum chamber surrounded by powerful electromagnets.
Soviet physicist and dissident Andrei Sakharov is jointly credited with the tokamak's invention in the 1960s.
FOR FUSION to occur, the fuel must be heated to about 150 million degrees celsius, at which point it becomes plasma, an electrically charged gas referred to as the fourth state of matter, consisting of electrons and nuclei.
Heating the fuel - hydrogen in the experiment's initial phase, then hydrogen and deuterium, and deuterium and tritium in the last phase - will rely on methods not unlike some used in the home kitchen, with Einstein, and his E=mc2 formula, playing the part of master chef.
Fusion temperature will be reached by a combination of firing high-speed particles into the plasma, like the steam that heats milk in a cappuccino machine, and using radio waves, similar to cooking with a microwave oven.
To complete the analogy, the electromagnets act a bit like the teflon on a non-stick frying pan: they prevent the plasma from contacting and melting the vacuum vessel. But they can't stop neutrons, liberated when the hydrogen nuclei in the plasma fuse to form helium, from hitting the vessel walls.
These collisions are both a good and a bad thing. According to Einstein's recipe, each new helium atom is lighter than the two hydrogens, and the mass it loses is converted into energy, most of which is taken on board by the neutron.
As the neutron hits the vacuum vessel wall, that energy is converted to heat, which can be used to produce steam to drive a turbine. But the neutron also makes the vessel radioactive, which is not so desirable.
White says harnessing fusion energy is not Iter's goal. The experiment will be deemed a success if a fusion reaction lasting 10 minutes can be established, and if there is a 10-fold net energy gain from the reaction - 500 megawatts' output (about half as much as the Huntly power station) from 50MW expended on plasma heating.
"First we will build the machine, run it at a lower level, then insert new components and raise the energy level, before inserting the final pieces and running it at the high level." By that time the calendar will have advanced to 2026, later than anyone originally thought.
"This is simply because we've come face to face with the practicalities of building a machine like this," White says.
One of the difficulties is a result of the way the project is set up. To share the economic benefits among all the funders, each contributing country is playing a part in building the machine, so that 90 per cent of what they put in is spent in their domestic markets.
THAT makes it hard to know the exact project cost. And it also presents enormous logistical challenges, ranging from transporting parts that weigh hundreds of tonnes across the world, to ensuring all the components fit precisely together.
"Although this approach is at least twice as expensive, takes much longer and complicates the machine's assembly, the upside is that every country involved walks away from the project with the full infrastructure and know-how to build a machine on their own," White says.
The Iter tokamak's scale and complexity can be appreciated from the fact that its 6m-diameter vacuum vessel will be surrounded by more superconducting magnets than the 3km-diameter particle accelerator at Cern, in Switzerland.
According to White, most Cadarache locals back the project. A consultation process that ended late last year found fears of a nuclear accident were less on people's minds than ensuring they got their share of jobs.
Economic spinoffs are helping win local acceptance. About €1.2 billion of project spending has gone into the region, which stretches from Nice to Marseille. And about a quarter of 500 Iter staff positions are allocated to French nationals.
The French aren't completely blase about radiation risks, White says, but fusion is a different kettle of fish from fission. Whereas last year's Fukushima disaster in Japan has revived fears of reactor meltdown, a malfunction at Iter would result in the fusion reaction simply stopping.
But getting the fission-fusion distinction across can be a challenge, particularly since the French term for meltdown is "fusion de reacteur". "So with all the Fukushima coverage this has been quite difficult to handle."
Will the machine work? Iter will prove whether fusion is viable, and knowing the outcome either way is a good investment, White says, when the global energy bill runs to trillions of euros a year.
Sir Chris Llewellyn Smith, who headed Iter's supervising council for two years to the end of 2009, and led Britain's fusion programme before that, thinks likewise. He sees hope in recent results from Jet. "It's hard to put a probability on Iter bearing fruit, but if it does the returns would be enormous, which makes the investment worthwhile."
By the time the reactor's 30-year operational phase is over, we should know the answer.
Anthony Doesburg is an Auckland freelance writer on science and technology
Kiwi graduate sees future in fusion
Jonathan Squire Although New Zealand is not part of the Iter experimental fusion effort, Kiwi talent is not completely absent.
Jonathan Squire is a University of Otago physics graduate two years into study towards a fusion-related PhD at Princeton University in the US. The Fulbright Science and Technology Award recipient says Iter is the biggest game in town when it comes to fusion research.
"It will be bigger than anything that's been before and will almost certainly do much better than all the previous experiments, unless something goes really wrong," says the 23-year-old.
Much of the work is done using computer simulations - the theoretical side of fusion that Squire is attracted to - but there comes a time when theory has to be put to the test. Iter - it stands for International Thermonuclear Experimental Reactor, and also means "the way" in Latin - is the crucial next stage.
The huge facility, which is being built in a €13 billion ($20 billion) project in the south of France, will advance the work done at the Jet (Joint European Torus) reactor in Britain, which has been operating since 1983.
Iter's doughnut-shaped reactor - or tokamak - has eight times the volume of Jet's, giving it a better chance of being able to sustain a reaction once its hydrogen fuel has been heated to the 150 million degrees Celsius needed to trigger fusion.
Keeping the reaction going is one of the many challenges to be overcome before fusion can be tapped as a low-carbon, renewable source of electricity.
Another challenge is producing a supply of tritium, an isotope of hydrogen that is likely to be the fuel source for a successful fusion reactor, with deuterium, a second form of the gas. Deuterium is abundant in sea water, but tritium doesn't exist in nature.
Squire says one promising idea is to produce tritium within the reactor itself in a reaction with liquid lithium, which would double as the material for absorbing the energy released by the fusing hydrogen atoms. There is thought to be enough mineable lithium to produce all the world's electricity from fusion for centuries.
Once Squire has completed his PhD, he would like to be working on research that supported Iter. However, if he doesn't get a role there, other fusion projects are under way in China and South Korea.
"There are quite a lot of big experiments that are coming online."
The next stage after Iter, Demo, is already being planned. Demo is intended as a demonstration fusion power plant, putting power into the grid by 2040.
The Iter organisation, however, doesn't forecast the world entering the "age of fusion" before the last quarter of the century. Squire thinks that sounds about right.
"Such big experiments always take a bit longer than everyone initially hopes so it's tough to imagine a scenario in which it comes quicker than that."
By that time Squire will be in his 80s.