Normally, only the innermost kilogram of uranium in an implosion-type bomb actually undergoes fission (and only 0.6 to about 5g of mass are ultimately converted to energy - the sources don't agree on this), because the explosion itself vaporizes and blows apart the rest and the fission cannot proceed further. This, by the way, was the main limit to the size of fission devices, that led Sakharov, Ulam et al. to design fusion devices.
In this case, though, practically the whole mass of uranium undergoes fission within a millisecond, while the outer shell (while still exploding) acts as a tamper. Only a very thin layer will neither undergo fission nor provide inertial confinement. Immediately after, the enormous neutron flux will trigger fission in that thin layer too.
At the same time, the neutron density will be enough to achieve a very high burn everywhere inside the sphere. So, while Little Boy had an efficiency of about 1.4%, and we can expect a fission efficiency of 50% (same source) due to the size, the actual energy efficiency will be much higher because the reaction will proceed further. A rough (probably incorrect) estimate based on the hypothesis that all the byproducts will be made up of the lowest (most "burnt") isotopes from the Hiroshima explosion tells us that the per-mass energy release should be about twenty times higher. And the mass of the sphere is around 9,948,368,333,000 higher than the fissioned mass at Hiroshima; we can then estimate an energy release to the tune of three billion megatons.
I wouldn't be surprised in the least if the radiation density at the center of the fireball allowed photodisintegration (photo-fission) and further increased the energy release (or maybe the reaction would find an equilibrium between photofission and nucleosynthesis, which is endothermic).
There is no need to calculate the exact magnitude of the explosion: it would not be enough to vaporize the Earth, but the radiative shock would scorch the target hemisphere and supply enough secondary activation to thoroughly poison the whole atmosphere as well as the oceans, as if the primary onslaught weren't enough.
The atmospheric blast reaches the other side of the planet after about sixteen hours - actually way less than that due to the speed of sound being higher in the high atmosphere - and increases constantly for at least eight hours; but the havoc it causes is probably nothing compared to the reentry of the incandescent, radioactive tephra from ground zero. Oh, the storm winds are radioactive too.
Primary concussion should be more than enough to trigger most existent fault lines, and seismic waves travel faster than sound in air; so for those countries that did not see the blast aurora, the first hint something went very, very wrong (apart from communications gone awry and possibly a sudden "sour, metallic taste in the mouth") might be a series of catastrophic earthquakes and volcanic eruptions.
Also, atmospheric pressure will finally stabilize to much lower value, possibly all the way to half of the former total, due to the gases being ionized and stripped off the Earth. In the worst scenario, oxygen partial pressure goes from 21% to about 11-12%, below the level sustaining current terrestrial life. I'm uncertain whether to count oxidation and burning - it might well be the case that there is not very much left to actually burn. If there is, then the oxygen goes still lower, and CO2 contents goes up.
Plus, the firestorm triggers the grandfather of all nuclear winters. Photosynthesis stops on the whole surface, probably for several years.
Plus the neutrino flux might supply enough inverse-beta decay to sterilize the whole planet, both hemispheres at once, or at least ensure death by cancer to all superior organisms.
Deinococcus radiodurans might yet survive.
Nothing else will.