I believe any product should be designed to benefit its user to the maximum possible extent. For us, this means:
Maximum accessibility: The product should "just work" in a way that the user can pick up the product and use it with a minimal learning curve.
Minimal upkeep: The product should be able to sustain itself without significantly inconveniencing the user. This means minimizing direct monetary costs (add-ons, subscriptions, etc.) and the opportunity cost of any required maintenance.
Uncompromising quality: We will never compromise the quality of our products for any reason. Not only do we believe in the fundamental right of the customer to a high-quality product, we also believe that providing customers with high-quality, affordable products results in significant additional marketing for us, as sufficiently-satisfied clients will go out of their way to recommend us to other people with computer issues.
With these three principles in mind, the benefits to a Flagship Tuneup performed on your ordinary laptop or desktop are obvious. Better thermal performance results in improved temperatures, processing speeds, and component lifespans, while reducing noise, temperature, and dust buildup.
These benefits help users focus better with lower surface temperatures and less distracting fan noise and help professionals converge simulations and finish exports more quickly, all while lowering downtime by reducing maintenance requirements from cleaning and re-pasting and reducing long-term costs by extending the life of your device through reducing the likelihood of component failure from overheating.
However, Flagship Tuneups aren't just reserved for your ordinary computer: my hands-on thermal expertise works across a wide variety of devices! here's a look at one of our more recent (and more unconventional) Flagship Tuneups. While it doesn't involve surface lapping, it still requires the high degree of experience I've gained in heat transfer from years of working on computer cooling.
Videography and content creation is a huge industry. From vloggers to Youtubers to streamers to movie producers, there's hundreds of billions of dollars all revolving around animated pictures you can project on a sreen. If you're a professional videographer, you need the best, most reliable tools to produce content and make a living.
Enter the Canon EOS R5, a $3,900 mirrorless camera released in 2020. On paper, it's a nearly ideal tool for photographers and videographers alike, featuring a huge selection of EF- and RF-mount still-photography and cinema lenses, in-body image stabilization to reduce motion blur and smooth video without cropping the image, and a durable all-magnesium chassis with weather-sealing.
Crucially, it's the first camera of its type to offer 8K video recording. While there's hardly a perceptible difference between 4K and 8K on-screen, 8K can make a huge difference when editing, as it lets you zoom in on footage to a greater degree without pixellation.
Unfortunately, Canon gave the R5 an Achilles' heel: a poorly-designed thermal system resulting in overheating after ~20 minutes of 8K video recording (or after ~25 mins of 4K60 video recording), requiring up to 2 hours to cool back down enough to continue recording. [1]
The short recording duration and extended cooldown period makes the R5 wholly unsuitable for professional high-resolution videography workloads. However, if you look at Canon's product portfolio, the R5's tendency to overheat was baked in for a very specific reason: to sell the EOS R5 C.
The EOS R5 C is a modified EOS R5 more focused for videographers, offering improved cooling with a heatsink and fan pre-installed for extended recording times. However, the R5 C has a variety of drawbacks compared to the standard R5:
No in-body image stabilization (IBIS): The standard R5 physically moves its sensor in five axes to counteract camera shake to smooth out photos and videos, providing some of the benefits of a tripod and gimbal head without the need for additional expensive equipment. However, the R5 C doesn't feature IBIS at all, significantly reducing its ability to smooth motion, especially when using lenses with reduced or nonexistent optical image-stabilization capabilities.
Bulkier and heavier: Due to the built-in heatsink and fan, the EOS R5 C is 30 grams heavier and features a large, unwieldy protrusion on the rear of the camera that makes it significantly less portable than a standard R5.
Added point of failure: The R5 C's reliance on active cooling adds a point of failure, especially if you're shooting in dusty or icy conditions. Replacement fans are virtually nonexistent for the end-user to buy, so any fan replacment is dependent on sending the R5 C back to Canon, resulting in downtime for an expensive repair.
More expensive: The EOS R5 C was launched at $4,500, nearly $700 more than the standard R5. As I'm writing right now (Feb. 2026), the R5 is down to $2,900 on eBay, whereas the R5 C still costs $4,250.
Clearly, the EOS R5 C is simply not a viable cinematography option for a large handful of potential users, and its lack of IBIS even limits its practicality as a stills camera.
Shortly after the EOS R5 was released, the company Kolari released an overheating modification service for the EOS R5. When launched, it cost $400 ($300 if you just want the necessary parts). It replaced the EOS R5's standard aluminum internal heat spreader with a larger, copper heat spreader that helped transfer heat into the magnesium chassis. Here's their self-proclaimed results:
8K 30FPS @20C ambient temperature: 44 min. (compared to 23 min. stock)
While this is enough time for the R5 to fully deplete its battery before it overheats, this extended recording time isn't enough for people who power the battery externally through its USB-C port for extended recording sessions, nor does it consider camera usage in hotter environments or on sunny days (this IS a pure black camera, after all).
An EOS R5 recently landed in my hands, and I knew I could do better than Kolari.
In addition to offering EOS R5 cooling modification as a commercial service, Kolari released it as a do-it-yourself guide (scroll halfway down the page for the relevant bits).
The EOS R5's motherboard contains the main processor, DIGIC X, surrounded by four memory chips. The heatspreader of DIGIC X is almost exactly 0.5mm higher than the top of the memory chips. From the factory, the R5 has two 1mm-thick thermal pads that bridge the gap between DIGIC X and the aluminum heatspreader (left). Kolaris' replacement copper heatspreader (right) is almost exactly 1mm thicker. Despite this, the official modification guide instructs you to simply move the thermal pads from the CPU to the memory. If you do the math, this means the thermal pads on the memory now sit 0.5mm above the processor's heatspreader. This has a few consequences:
Suboptimal thermal performance: Because of the increased height above the processor more thermal paste is needed to bridge the gap, resulting in a larger ΔT between the processor and heatspreader. Because the processor is the main heat-generating component that governs overheating based on its temperature, a large ΔT results in premature shutdown from overheating.
Excessive and incorrect materials: Kolari instructs you to fill this gap with a tube of thermal paste that they include with the replacement heatsink. However, thermal paste is simply not the correct material to use for large gaps due to the possibility of thermal paste becoming less viscous at elevated temperatures, running out from the application area and causing a mess while worsening temperatures. The correct material would've been gap fillers, which are designed to thermally connect two surfaces that aren't directly pressing against each other.
Fortunately, there's an easy improvement on Kolari's procedure. By removing the factory-applied thermal pads and replacing them with aftermarket 0.5mm thermal pads directly applied only to the memory, the heatsink sat low enough to completely eliminate the gap between the CPU and heatsink, allowing me to use a small amount of thermal paste to bridge the gap. I ensured the memory thermal pads still made excellent contact with the heatsink as well.
Next, I did something counterintuitive: increase the thermal pad thickness between the top of the heatsink and the inside of the magnesium chassis. Originally, Kolari instructs to use a 1.5mm thermal pad. However, a DC-DC converter board sits on top of the heatsink, and this 1.5mm thermal pad barely provides any contact with the tallest of components. Increasing the thermal pad thickness above the DC-DC board to 3.5mm significantly increases the pressure on the thermal pad, providing significantly better contact with components on the DC-DC board, as evidenced by the impressions seen on the large thermal pad (bottom-right image, left side). This helps wick away heat from the DC-DC board and the heatsink under it, reducing equilibrium heat transfer. (I added thermal paste under the DC-DC board to help conduct heat away from the heatsink.)
Because the chassis is made of thick magnesium, it can easily handle the increased mounting pressure from thicker thermal pads without deformation, which helps maintain the external appearance of the camera while leaving the camera's weather seals untouched.
As a reminder, here's Kolari's stock testing results:
8K 30FPS @20C ambient temperature: 44 min.
Here's our testing results under nearly-identical conditions:
8K 30FPS @21C ambient temperature, no forced convection: ∞ min. (okay I tested it for 5 hours but it just would not overheat, so it's definitely reached thermal equilibrium)
These results are huge! At 20 C, the near-overheating indicator never even shows up, indicating that this R5 has considerably extra thermal headroom compared to even Kolari-modified units. It means that...
this modified EOS R5 is now much more suitable for videography usage even in higher-temperature and sunny environments.
Videographers can record for far longer on external power without worrying about overheating.
There is simply no cooldown period to worry about, so this $200 modification prevents videographers from having to purchase multiple $3,900 EOS R5 bodies to work around the cooldown period.
This also prevents issues like footage-syncing, transportation and storage, and settings-syncing that would occur with multiple cameras.
The removal of practical video-recording limits on the standard, IBIS-equipeed EOS R5 lets value-conscious content creators have a single device capable of both stabilized stills and continuous stabilized video, something that neither an unmodified EOS R5 or EOS R5 C can do.
While this Flagship Tuneup isn't exactly the standard surface-lapping procedure, the same experience and intuition with heat transfer still applies. This is evidence that I can apply my years of experience with cooling and heat transfer in a manner that can significantly enhance the value and utility of a device to end-users in a meaningful way, saving significant amounts of time and money.
Flagship Tuneups are a service truly unique to us. If you're interested, please check out our Flagship Tuneups page, and contact us at berkeley.the-it-club.org!