How relevant is the "myth" about bad GPS in MTK chips?

Question

satellite navigation

One of the first MTK chips left, at one time, a deep trauma in the souls of users with their terrible GPS. No one to whom he (gps) is somehow important wants to mess with them.
How relevant is this for the chips they produce now?
Should we now be afraid of Chinese chips?
And how striking is the difference between the same "snapdragon" and "mt"?

Answer 1

In short: yes, still relevant.
Because it is difficult to understand the units in which to measure the difference, I will try to answer in detail:
The difference is not only in the signal processing algorithm. The general engineering and execution of devices based on MTK chips also largely influences. Strange antenna placement, solderless, etc. more typical for such devices. Secondly, in addition to the chips themselves, there are also firmware. I know several examples when complete nonsense in gps.conf (launched by someone on XDA) wandered from firmware to firmware of a completely official, fairly well-known manufacturer.
As for the chip itself, yes, the algorithms are different. MediTech, at least, came up with (EPO, it seems) their own version of optimization a la A-GPS, and something is "not very" with it.
However, through the engineering menu, as a rule, you can turn it off and return to the classic A-GPS. If after that you fix the settings files (sometimes convincing the firmware that you don’t need to return everything back) and correctly calibrate the GPS, then there will be no noticeable problems with the algorithm.

Answer 2

The technology and implementation of MTK geopositioning itself is excellent compared to the proprietary HA allowed by Sirf, which, for example, still limits the number of processed satellites and accuracy for unclear reasons.
As an example, you can search the description and specifications of the MTK3333 and compare them with other solutions.
But there was, and continues to be, the problem of implementing solutions from MTK by third-party (Chinese) manufacturers who use MTK chips in their devices, and there you can really see a deviation from the MTK recommendation.
MTK doesn't make devices for end users, they offer KITs.
Kir aka Dober mentioned gps.conf, this file is not from MTK module firmware, it is from Android firmware.
How Android is screwed to Chinese devices is a separate, very sad topic.
In general, the solutions of American and European engineers are very stable and expectedly reliable, but somehow they are limited either by licenses, or by militarists, or by conspirators, I don’t know.
With China, it’s more difficult, it’s how you get there, but you already come across solutions from which you get both speed and accuracy and pleasure from implementation and use.