Спокойно може да се използва този въпрос и на интервюта за работа за да отсеете надъханите пишлемета и зубрачите от тия на които ембедед софтуера им е в кръвта.

...
С две думи - не виждам никаква реална полза от подобна "врътка" - напротив дори да има то прави кода "чуплив". А за четимост и поддръжка изобщо не става дума.

4-ти случай.
Това е за наистина ама наистина разбирачи. Които правят тънки сметки и подравняват кода в кеша с добавяне на една-две, привидно излишни, инструкции тук-таме из кода.

As a disadvantage we should refer to the fact that FRAM reads are destructive.
That is, every read must be followed by a write operation (with the same
data). However, this is automatically handled by the FRAM controller, and the
end user does not need to pay any attention to this. Therefore, the frequency of
FRAM read operations are limited to the write speed. Due to this limit, TI has
integrated a read cache in front of the FRAM to accelerate the operations in case
the MCU operates at a higher frequency than the FRAM. In the MSP430FR
family, the FRAM can be operated at up to 8 MHz without use of this cache.
When the MCU operates at a frequency of 16 MHz (the maximum operation
frequency of the MSP430FR family), the cache is utilized.
The integrated cache is a two-way associative cache containing two cache
sets [17]. Each of these sets consists of two lines of four words (64 bits). The cache
controller selects one of the cache lines to preload FRAM data and preserves
recently-accessed data in the other cache line. If one of the four words stored in
one of the cache lines is requested (a cache hit), no FRAM access occurs, and
the requested data is read from the cache with full system speed. However, if
none of the words that are available in the cache is requested (a cache miss),
a wait state (one clock cycle at 16 MHz) controls the CPU to ensure proper FRAM access. Therefore, memory read accesses on consecutive addresses can be
executed without wait states when they are within the same cache line.
Each 64-bit location in FRAM can be cached in only one of the two sets in
the cache. As shown in Figure 1, the most common scheme is to use the least
significant bit of the FRAM location’s address as the indicator to the corresponding
cache set. We should emphasize that FRAM contains both program
code and data, e.g., look-up tables, which are to be stored in the non-volatile
memory. Hence, frequent jumps and frequent accesses to the pre-stored tables
in FRAM can negatively affect the cache performance.

During the mission, the cause was diagnosed as the rendezvous radar switch being in the wrong position, causing the computer to process data from both the rendezvous and landing radars at the same time.[118][119] Software engineer Don Eyles concluded in a 2005 Guidance and Control Conference paper that the problem was due to a hardware design bug previously seen during testing of the first uncrewed LM in Apollo 5. Having the rendezvous radar on (so it was warmed up in case of an emergency landing abort) should have been irrelevant to the computer, but an electrical phasing mismatch between two parts of the rendezvous radar system could cause the stationary antenna to appear to the computer as dithering back and forth between two positions, depending upon how the hardware randomly powered up. The extra spurious cycle stealing, as the rendezvous radar updated an involuntary counter, caused the computer alarms.

...
Може да звучи странно, ...

PDP-11, VT102 и BSD 2.11 с работещ етернет и ТСР срек
качено на ESP32 (~$1) с 512КВ PSRAM и 4/8МВ флаш... (още $2), сторидж на SD (8GB за 5лв)

първото PDP, до което ме пуснаха, имаше 64КВ феритна памет, 2х10+10МВ 30" дискове (или бяха 5+5?), 2х38МВ 1/2" ленти и оригинален телетайп...
няма и 40 години от тогава

Въпрос е много полозен: директно разбираш, че е по-добре да продължиш с търсенето на друго работно място. Изобщо тази мода да се задават подобни въпроси е силно токсична. За мен по-доброто е: покажи какво си правил.

Всъщност от време на време има проблясъци - новият ARM MacBook е проектиран доста дорбе и с мисъл, работи като пушка, не грее, батерията му държи на цял ден работа, че и на два ако не си даваш много зор да работиш (има и такива дни, уикенди). ....

А каква е тая набираща популярност напоследък мания, GUI за Уиндоус Софтуер да почва да се пише на.... HTML и Java Script.

До тоя извод сам ли стигна или от ревюта в нета? Наистина ли е по-добър от х86 в ежедневна работа и в частност - програмиране (докери, джави, змии)?