ITER is one the largest physics experiments ever attempted. In short, it should be the first research fusion reactor that actually will succeed to produce power. ITER attempts to operate in the 500 MW (thermal) range, which is around the output of a small nuclear reactor. 

If you look at the design or operating parameters, everything it is just mind-boggling. The torus dimensions are just huge - 12 meters in diameter, and six meters height. In total, a little less than 1000 cubic meters of plasma (mostly, rarefied tritium + deuterium gas). The plasma is kept at 100 million celsius degrees, yet at a meters away, there are huge superconducting coils that are cooled down to just a few degrees above absolute zero.

If you are concerned about safety - this is much safer than a fision-based reactor. The fusion reaction will abruptly stop as soon as the carefully controlled conditions are not met anymore. There are no more issues around super-criticality, etc. that are so sensitive for a normal nuclear reactor.

It's total cost would be around 3 billion dollars, and there are several countries that are helping with this project, notably EU, Japan, US and others. This all looks interesting, but it looks like the participating countries can't decide where to build this thing. There seems to be a fight between Japan and France for the last months, and there are apparenly some recent rumors that the reactor might be built in France.

 

P.S. Maybe my calculations are wrong. One gram of Ra-226 has one curie, and a half-life is 1600 years. One gram of Tritium (H-3) has 12 years half-life will have 1600/12 * 226/3 = 10,000 Curie. Now, since there are maximum 450 grams of H-3 in the reactor at any time, if it completely blows up you get around 4.5 millions Curie in the atmosphere, which is much less than from an fision-based reactor. It's still a significant number, though.