A nuclear heating device with a mass of around 1000kg wouldn't make a measurable difference on the martian atmosphere in it's total (nuclear) lifetime. If you use MMRTG's, the same power source that the Curiosity rover and the Voyager probes have, you could bring around 20 of them as they mass around 50 kg a piece.
Your average MMRTG puts out 125 Watts of Electrical power and 2kW of Thermal power. If we use the total power, and use the electrical power to heat, of all 20 MMRTG's, that's a thermal power output of 42.5 kW. Now 42.5 kW of heating power sounds like a lot but it really isn't on a planetary scale.
On Mars, sunlight strikes the surface with an energy of around 500 W/sqm. If mars were 100% sunlight absorbent it would require around 85 square meters of sunlight to equate 42.5 kW of power. Since it isn't, lets say we need around 200 square meters of sunlight heating to equate the nuclear heating--that's only a ~14 meter square patch of Martian soil. This means a small landslide uncovering darker sand has a larger effect on the climate than a small heater melting some c02CO$_2$.
If you really want to start terraforming with 1000kg, it would be more effective to bring a lot of black paint to increase heat absorbance or use your 1000kg as a small nuclear warhead.