With their lightweight yet sturdy build, solid performance, and pleasing buttons, the LAMZU Atlantis and Atlantis Mini have been among the most successful mouse releases in recent times. The Atlantis 4K builds on this success by mostly updating the internals, most notably including the ability to enable true 4000 Hz wireless polling. Whereas the original Atlantis and Atlantis Mini were USB full-speed mice and thus limited to a polling rate of 1000 Hz, the Atlantis 4K is a USB high-speed device and thus natively capable of polling rates of 2000 and 4000 Hz.

While the Atlantis 4K still comes with PixArt's PAW3395 sensor, it is paired with a Nordic nRF52840 instead of a CX52850 MCU, though the firmware/software solution is still provided by CompX. Compared to competitors using the same solution, such as the VGN Dragonfly F1 Pro Max, the Atlantis 4K fares a decent bit better. There is no CPI deviation at all, general tracking is flawless regardless of settings, and polling is stable almost throughout. That is, occasionally polls are delayed, resulting in small outliers in tracking. This affects all polling rates, but only happens infrequently, and I found in-game performance unaffected. The originally present polling instability at polling rates of 1000 Hz and lower has been addressed in the latest firmware update. In terms of motion delay, the Atlantis 4K at 1000 Hz and without MotionSync already achieves parity with the G403 (control subject), and the lead only increases at 2000 and 4000 Hz, as the Atlantis 4K is ahead by almost 1 ms at the latter polling rate. Unlike the Atlantis and Atlantis Mini, the Atlantis 4K allows disabling MotionSync. MotionSync is a sensor feature that improves tracking consistency by synchronizing SPI reads with USB polls, but introduces a minor motion delay penalty roughly equal to half of the set polling interval; i.e., 0.5 ms at 1000 Hz, 0.25 ms at 2000 Hz, and so on. As such, in order to achieve the lowest possible motion delay, and get the most out of 4000 Hz polling compared to a 1000 Hz mouse without MotionSync, disabling MotionSync may be preferable.

In terms of click latency, the Atlantis 4K likewise delivers excellence. When using the lowest possible debounce time (0 ms), click latency at 4000 Hz sits at just 0.8 ms, which is only marginally shy of the numbers mustered by the likes of Razer or G-Wolves. At 2000 Hz, 1.0 ms are averaged, and at 1000 Hz, click latency sits around 2.4 ms. Higher debounce time values will result in higher click latency, which is relevant in that the Atlantis 4K already slam-clicks at values of 0 and 1 ms. Slam-clicking describes the inadvertent actuation of the main buttons upon "slamming" or gently repositioning the mouse after lifting it off the pad. In-game, this is undesirable, since people using their wrist for aiming often resort to rapidly repositioning their mouse instead of moving their entire arm, and one's weapon going off in an FPS in such instances is certainly not welcome. In my testing, I've found that only by increasing debounce time to 2 ms such behavior could be prevented, resulting in accordingly higher click latency. Another clear improvement over the original Atlantis and Atlantis is the lack of any idle delay on the Atlantis 4K; i.e., there is no additional latency on the first click after not moving the mouse for several seconds.

Battery life is a bit of a mystery on the Atlantis 4K. LAMZU provides no numbers, and while the percentage-based battery life indicator in the software suggests single-digit accuracy, I didn't find the readings particularly reliable. Considering that the VGN Dragonfly F1 Pro Max, which utilizes the same firmware solution in conjunction with a larger 500 mAh battery, musters around 30-35 hours at 4000 Hz, I would expect the Atlantis 4K with its 300 mAh battery to land somewhere in the range of 20-25 hours. Battery drain at 2000 and 1000 Hz is roughly two and four times lower, so battery life can be expected to be two and four times higher, respectively. Charging is decently speedy on the Atlantis 4K, and while the charging cable isn't particularly flexible, the angled connector helps a bit, making playing while charging quite possible, though it may feel somewhat restrictive still. Speaking of the software, this one has all the important settings and pleasantly low resource usage, and the only bug I've found relates to inputting CPI values manually, which simply doesn't work, so one has to use the slider instead.

In terms of build and buttons, the Atlantis 4K is just as solid as the original Atlantis. On my sample, there is no rattle when shaking, no flexing of the shell and only minor creaking when applying lateral pressure, and actuating the side buttons by pressing below them is impossible. The Atlantis 4K still utilizes an open-bottom design, much like Pulsar or G-Wolves mice, and the PCB is designed in such a way that it aids with structural rigidity. As a result, the Atlantis 4K still weighs 55 g, much like the original Atlantis. Huano switches (blue shell, pink plunger) are still used for the main buttons, which provide a snappy and firm button response, some pre-travel notwithstanding. Compared to the Atlantis, the side buttons have been improved, as post-travel is quite a bit lower now, and the Huano switches (white plunger) used for these deliver a very satisfying button response, along with even actuation. The same switch type is also used for the scroll wheel click instead of the tactile switches seen on previous releases of the Atlantis and Atlantis Mini. For the scroll wheel, LAMZU has moved from a TTC Gold encoder to a TTC Silver, which is said to be less susceptible to dust, and therefore likely more durable. In terms of tactility, I prefer the previous TTC Gold encoder, though separation between steps is still quite good, although noise levels also seem to be higher than with the TTC Gold one. In terms of materials, the feet are unchanged and still made of pure PTFE, gliding as well as ever as a result, but their shape is different now. LAMZU not only includes a set of replacement feet in the box, but also a set of grip tape, which is commendable.

Overall, the Atlantis 4K improves on its predecessors pretty much across the board. Click latency in particular is significantly lower, even when using 1000 Hz, and motion delay likewise a good bit improved. Considering the uplift in performance regardless of polling rate, a premium of $10 over the regular Atlantis is more than fair, and considering that Razer or G-Wolves are charging upwards of $150 for their 4K mice, $99.99 for the Atlantis 4K is perfectly fine, too. Accordingly, the Atlantis 4K earns our Editor's Choice.

Looking at the competition, one is spoiled with choice. For $179.99 including the HyperPolling Wireless Dongle enabling 4000 Hz wireless polling, the Razer Viper V2 Pro has a slightly higher weight, slightly better performance, similarly good build quality, and comes with optical main button switches. For $159.00, the G-Wolves Hati-S Plus 4K likewise comes with true 4000 Hz wireless polling, similar performance, similarly good buttons, and a slightly lower weight at 51 g. The $119.99 VAXEE XE Wireless has slightly worse performance, a significantly higher weight, slightly better buttons and build, and extensive on-board configurability. For $99.99, the Xtrfy M8 Wireless comes with worse performance, similar weight, great buttons and build, and no need for software. The $94.95 Pulsar X2 and X2 Mini have worse performance, a similar weight, and similar button quality, but some issues with quality control. For $79.99, the Endgame Gear XM2we has worse performance, a lower weight, optical switches for the main buttons, and perfectly functional software. Lastly, for $69.98 including the 4K Dongle, the VGN Dragonfly F1 Pro Max has virtually the same performance, same weight while also featuring a full bottom plate, similarly good buttons, along with true 4000 Hz wireless polling.