These, in turn, will help extend the uptime of devices like laptops. And in the neighboring lithography department, amazing work is being done to ensure that those aren’t the only types of computers destined to become more and more powerful. “This is where we make exposure systems for the microchip industry,” says Markus Wiederspahn, a press officer at Carl Zeiss, as he points to lenses about 30 centimeters in diameter. These lenses are produced, ground, coated, and assembled into extraordinarily precise objectives at the plant. The tolerances involved here are measured not in millimeters but nanometers.

Only with this level of precision is it possible to project the ever finer structures onto the photoresist of a silicon disc known as a “wafer”. Wiederspahn, who had academic training as an eyeglass optician, knows the challenges of this sort of optical system only too well. “The objectives that project the ‘slide’ have to be designed for very specific wavelengths — the shorter those wavelengths are, the smaller the semiconductor structures can be on the chip later on.”

Carl Zeiss checks the individual lens elements of the objectives by means of wavefront measurements on the nanometer scale. At every glass-air surface, however, the light loses four percent of its power through reflections, which then travel around in the objective and degrade the contrast and definition of the image. But ultra-thin coatings reduce these reflections to the extent that it becomes possible to exploit the benefits of ultraviolet light with a wavelength of only 193 nanometers.

Moreover, production can also take place at a very brisk pace: A “stepper” creates 180 wafers per hour, and each one is exposed 30 to 60 times in order to generate the three dimensional structures for future microprocessors. Eventually, some of these will in turn be installed in electron microscopes. The cycle is completed in this hall, which has a 1.5-meter-thick concrete floor and is thus well suited to absorb almost any vibration.

Instruments that expand our knowledge of the world will continue to emerge from these halls in years to come. And they will help ensure that Moore’s Law remains valid for a long time: in 1965, Gordon Moore observed that the complexity of electronic circuits was increasing exponentially.

In its current version, the “law” named after him states that the number of transistors that can be created on a surface will double every two years. Mastering even shorter wavelengths is one key to the continued accuracy of this law. Recently, Zeiss took a crucial step toward this goal with the delivery of the first systems that expose wafers using wavelengths of 13.5 nanometers — less than a tenth of the wavelength commonly used today.