In recent years, with the gradual maturity of processes and IP, 32-bit MCUs have grown rapidly, to the point where 16-bit MCUs are almost skipped. Nowadays, when talking about embedded MCUs, it is either 8-bit or 32-bit, with very few 16-bit MCU product models available.So, what about the situation with 8-bit MCUs? Many embedded engineers have some misunderstandings, and here we will briefly analyze them.1. 8-Bit MCUs Are Being Phased OutThis is the most common misunderstanding. First, let’s look at the facts. According to the latest Gartner market report, the market revenue and growth of 8-bit MCUs are only a few percentage points behind 32-bit MCUs. Considering that 8-bit chips are significantly cheaper than 32-bit chips, the shipment volume of 8-bit MCUs is actually much higher than that of 32-bit MCUs.To put it intuitively, just because we have high-speed trains now, does that mean all traditional trains should be immediately phased out? Clearly, that is not the case, and there are many reasons for this. The reality is that the application fields of 8-bit MCUs cannot be directly replaced by 32-bit MCUs.2. 8-Bit Processors Lack InnovationMany believe that since the market’s darling is the 32-bit MCU, manufacturers must not be investing R&D resources into 8-bit products. In fact, those who think this way might immediately recall the “classic 8051” with the 40DIP design when they think of 8-bit MCUs. However, chip manufacturers have not stopped innovating.For example, the CIP-51 core has increased the performance of the 8051 at the same frequency by 12 times due to its design where one clock cycle equals one instruction cycle. Some domestic semiconductor manufacturers have also innovated based on the 8051 or other 8-bit cores.3. 8-Bit Processors Are Difficult to Program in C/C++If you understand the design principles of Arduino, this misunderstanding will be dispelled. To be honest, programming 8-bit MCUs in high-level languages is indeed more challenging than with 32-bit MCUs, primarily due to the lack of unified memory addressing. For instance, the memory addresses of the 8051 core are divided into CODE, data, sfr, idata, and xdata. If banking is involved, it becomes even more complex.8-bit PICs also have hardware stacks that are more “non-mainstream” designs, but these obstacles can be alleviated through tool optimization.4. 8-Bit Processors Are Only for Simple ApplicationsThis viewpoint holds some truth, but embedded applications are generally simple. The characteristics of embedded system applications determine that 8-bit MCUs still have many applications. The evolution of peripherals and compilers will gradually expand the application scope of 8-bit processors.5. 8-Bit Processors Cannot Meet IoT Application NeedsIoT applications are not a standalone application but a composite one. Smartwatches, smart speakers, main controllers, and gateways certainly require complex processors. However, IoT applications also include numerous sensor nodes, actuator nodes, and converter nodes. These nodes are more suitable for implementation with low-power 8-bit processors.6. 8-Bit Processors Are Slow to RespondThis is a complete misunderstanding. In typical embedded applications, response speed is mainly related to interrupt response and wake-up delay. 8-bit processors have inherent advantages (less address conversion workload, fewer IP unit implementations), and at least do not lag behind 32-bit processors.7. 8-Bit Processors Have Lower Energy Efficiency Than 32-Bit ProcessorsI once read a book by an authoritative engineer from ARM, which stated that the energy efficiency of 32-bit processors is higher than that of 8-bit MCUs because 32-bit processors can quickly complete tasks, resulting in a larger proportion of sleep time. However, this conclusion contains an assumption that the tasks have a certain level of complexity.If the tasks themselves are very simple, the power consumption during the wake-up process can also be significant, then this assumption does not hold. For different application scenarios, one cannot simply say which has a higher energy efficiency, 8-bit or 32-bit. At least in very simple applications, the energy efficiency of 8-bit processors is higher. If we add tasks that can respond independently without CPU intervention, the energy efficiency of 8-bit processors can be significantly higher.8. 32-Bit Processors Offer Superior Functionality at the Same PriceThis is also somewhat credible, but do not forget that a considerable number of applications are sufficiently met by 8-bit MCUs. In such cases, purchasing a slightly more expensive 32-bit MCU will only increase costs.For many embedded products that have basically been standardized, 8-bit MCUs still have a certain cost advantage.9. Applications Designed with 8-Bit Processors Cannot Adapt to Future ChangesThis is a matter of perspective. As an embedded programmer, one should focus more on the current tasks. Regardless of the type of MCU, if the product form changes or the requirements change, it needs to be redesigned. The future is unclear for anyone; why worry about so many forward-looking considerations that have no practical significance?10. Development Work for 8-Bit Processors Is More Burdensome and Lacks Upgrade Path32-bit processors are more software-centric and allow for more code reuse. In contrast, 8-bit processors rely more on hardware peripherals to complete tasks. Overall, there is no absolute difference.As long as it is an embedded processor, the upgrade path is not very clear. If you use manufacturers that offer both 8-bit and 32-bit products, you will find that many peripherals are quite similar. Considering the current trend of graphical configuration of peripherals, the upgrade path is becoming less important, as it is all about generating basic driver code through graphics or scripts.Source: The Last BugThis article is reproduced from The Last Bug , and the Electronic Engineering Magazine remains neutral on the views expressed in the article, which are for learning and communication purposes only.
