It wouldn’t really surprise me if the chips in the modems were capable of faster speeds than what they are running but most manufacturers don’t shoot for a performance mark two times greater than what they actually have to hit. The area that is most often left lacking in order to cut production costs is almost without fail cooling. Spending money to add extra cooling measures at the factory can run contrary to profit margins by increasing production costs (for the extra cooling components and the time required to install them) as well as by making the part last too long thereby cutting down the market for replacements (as long as it makes it past the warranty period, the cooling is adequate — been there, seen that). They clearly didn’t add cooling adequate for the speeds they are supposed to run not to mention double that speed. For my modem to run as hot as it was getting does not exactly lead me to be impressed with the cooling design of the modem (especially considering it is off the first split from the main cable that comes into the house, everything in between was replaced with the highest quality components available including the cable and the splitter [the cable guy was more than pleased with my upgrades in the attic ] and it was on a well ventilated shelf).
Anyway, For there to be not even so much as a thin-line passive heatsink on a chip that would run 45C/113F without one is just silly and has the overall feel of penny pinching. IMO, to have a bare chip that runs 55C/131F without a sink falls into the category of negligent cooling design. Temps that are well under the max operating temp are much more reassuring than those on the edge of the max storage temp range. hmmm…I will have to check the temps on my friend’s modem for the fun of it – his house has more splits in the cable line than can be counted and I suspect it is on a line at least 3 down from the main.
the idea here is not to get some amazing increase in speed from the modem – that is obviously the result of the service provided by the ISP. As stated previously, the plan was to keep the chip from overheating and possibly make it a bit more stable from moment to moment under heavy traffic.