Fixing Apple's Engineering in an Hour

The video opens by establishing the MacBook Air 2020 as a sleek machine with a notable design flaw: Apple places the cooling fan away from the CPU, limiting cooling capacity to around 10 watts. The hosts propose a series of unconventional cooling experiments to test how much performance is really left on the table. They assemble a variety of cooling apparatus, including nontraditional stands, fans in odd configurations, and attempts at water cooling, all to quantify the impact on CPU performance and thermals. The baseline measurements show the stock machine pulling around 40 to 45 watts initially, then thermally throttling as temperatures hit 100 degrees Celsius, with noticeable loss of performance as the CPU power is constrained. The discussion then shifts to how much improvement each radical approach can deliver, and whether the MacBook Air’s design choices are inherently preventing meaningful performance gains, regardless of cooling modifications. A core segment explores cooling method number one, a simple stand with an extra fan. The team notes that this adds a fan but fails to effectively direct airflow over the heatsink, leading to only modest gains in performance and limited reduction in CPU temperature. The crew then analyzes the CPU location and cooling path, puzzled by the design choice that keeps the heatsink and fan from working together. As the tests progress, the water cooling experiment is introduced, with a tongue-in-cheek nod to past experiments from the channel, testing whether pouring water around the chassis or submerging the laptop could yield higher sustained performance. Despite the dramatic setup, the results show the machine remaining power-limited rather than thermally limited, with only incremental gains and surprising limits in how much coolant or surface area can realistically improve performance. In a later phase, the team experiments with expanding cooling surface area via thermal pads and chassis insulation deletion. They remove a thermal insulator and place pads to boost conduction, only to discover that while the bottom temps fall into the 40s, sustained performance still struggles to climb beyond a marginal uplift. They then revisit a more extreme approach by attempting direct water cooling, but the pump power and coolant flow fall short of expectations, leaving the system thermally or power-limited rather than fully dissipation-limited. The session wraps with a synthesis: the MacBook Air’s CPU power delivery and overall thermal design severely cap performance, and the planned future tests reveal that even aggressive cooling modifications cannot unlock the full potential promised by the hardware alone, though the experiments provide insightful lessons on laptop thermal design and the limits of DIY cooling experiments. The video ends by acknowledging the sponsor Ting and framing the day as a learning experience rather than a definitive hardware fix, inviting viewers to consider the broader implications for how laptops are engineered and sold.

Topics · technology · hardware · laptops · thermals · engineering · performance · computing · science-and-technology