In this case, it's absurdly beside the point as well. The expenditure of nuclear materials is utterly irrelevant to the problem.
The way to rocket works it to use nuclear-generated heat to expand and accelerate a working fluid (usually hydrogen) and shoot it out the nozzle. What matter is the mass of the working fluid expended per impulse (force x time) - the specific impulse (lb-sec/lb or kg-sec/kg, for units of seconds) or ISP. T
housands and thousands of lbs of the working fluid will be consumed, the fact that it also consume a few ounces of nuclear material, too, is utterly in the noise.
A very good chemical rocket will have an ISP of 450-460 seconds. A nuclear thermal rocket will have an ISP of around 900-1000, or roughly twice as "good". "Good" is defined by the amount of impulse/momentum change you get for a given amount of fuel consumed.
A nuclear thermal engine can be built to provide almost any desired thrust level, with 25000 lb thrust engines actually built and tested.
By comparison, a Hall Current or other ion engine will have an ISP of around 1800, but use vast amounts of electrical power for extremely feeble thrust of far less than a pound in typical cases.