We report on two ultrastable lasers each stabilized to independent silicon Fabry-Pérot cavities operated at 124 K. The fractional frequency instability of each laser is completely determined by the fundamental thermal Brownian noise of the mirror coatings with a flicker noise floor of $4\times {10}^{-17}$ for integration times between 0.8 s and a few tens of seconds. We rigorously treat the notorious divergences encountered with the associated flicker frequency noise and derive methods to relate this noise to observable and practically relevant linewidths and coherence times. The individual laser linewidth obtained from the phase noise spectrum or the direct beat note between the two lasers can be as small as 5 mHz at 194 THz. From the measured phase evolution between the two laser fields we derive usable phase coherence times for different applications of 11 to 55 s.

Crystalline optical cavities are the foundation of today’s state-of-the-art ultrastable lasers. Building on our previous silicon cavity effort, we now achieve the fundamental thermal-noise-limited stability for a 6 cm long silicon cavity cooled to 4 K, reaching $6.5\times {10}^{-17}$ from 0.8 s to 80 s. We also report for the first time, to the best of our knowledge, a clear linear dependence of the cavity frequency drift on incident optical power. The lowest fractional frequency drift of $-3\times {10}^{-19}/\mathrm{s}$ is attained at a transmitted power of 40 nW, with an extrapolated drift approaching zero in the absence of optical power. These demonstrations provide a promising direction to reach a new performance domain for stable lasers, with stability better than $1\times {10}^{-17}$ and fractional linear drift below $1\times {10}^{-19}/\mathrm{s}$.

We report on a laser locked to a silicon cavity operating continuously at 4 K with $1\times {10}^{-16}$ instability and a median linewidth of 17 mHz at 1542 nm. This is a tenfold improvement in short-term instability, and a $10^4$ improvement in linewidth, over previous sub-10-K systems. Operating at low temperatures reduces the thermal noise floor and, thus, is advantageous toward reaching an instability of ${10}^{-18}$, a long-sought goal of the optical clock community. The performance of this system demonstrates the technical readiness for the development of the next generation of ultrastable lasers that operate with an ultranarrow linewidth and long-term stability without user intervention.

We report on the first timescale based entirely on optical technology. Existing timescales, including those incorporating optical frequency standards, rely exclusively on microwave local oscillators owing to the lack of an optical oscillator with the required frequency predictability and stability for reliable steering. We combine a cryogenic silicon cavity exhibiting improved long-term stability and an accurate ${}^{87}\mathrm{Sr}$ lattice clock to form a timescale that outperforms them all. Our timescale accumulates an estimated time error of only $48\pm 94\mathrm{ps}$ over 34 days of operation. Our analysis indicates that this timescale is capable of reaching a stability below $1\times {10}^{-17}$ after a few months of averaging, making timekeeping at the ${10}^{-18}$ level a realistic prospect.