In the realm of custom Android hardware development, the array of potential product use cases is virtually boundless. While every product is unique, the process of creating these products adheres to a consistent methodology.
When conceptualizing the architecture of an Android product, the initial step involves selecting the appropriate CPU, which stands for central processing unit and serves as the device’s ‘brains.’ While users may not directly perceive the intricacies of the CPU, they interact with all the components controlled by it.
The optimal approach to choosing a CPU begins with a thorough understanding of the product’s use case. CPUs exhibit varying levels of capabilities, encompassing factors such as support for different screen or camera resolutions, WiFi-only or WiFi+Telecom functionalities (including mobile data and calling), memory constraints, processing power, and other distinguishing features. By comprehending the product’s use case, specifications can be defined, paving the way for the identification of a CPU that aligns with these requirements.
Generally, the preferred choice is a CPU that meets the specified requirements without surpassing them excessively. In scenarios where the client envisions future products with heightened performance demands, evaluating the feasibility of developing two distinct architectures (utilizing different CPUs) becomes crucial. Incorporating high-end components into a lower-end product can inflate the unit price unnecessarily, but there is also a cost associated with development and potential revenue loss from not having a product in the market. Key considerations in this decision-making process include volume, timing, and upfront capital.
When assessing different CPU brands and models, the initial consideration is whether the CPU supports telecom functionalities. For products not requiring calling or mobile data, the three most notable CPU suppliers are Rockchips, Mediatek, and, to a certain extent, Allwinner. Despite offering relatively similar chips, discerning the best match necessitates an evaluation of finer details and intangibles, such as the quality of technical documentation and sales support, among other factors. For custom Android products requiring telecom functionalities, Mediatek typically emerges as the preferred option for most customers. Qualcomm’s custom development costs are prohibitive for quantities less than 100k units, while Spreadtrum competes on the low end in terms of both price and performance. Mediatek occupies the middle ground, providing world-class CPUs at competitive prices, meeting the needs of most companies that don’t require extremes in performance or cost.
A valuable factor to consider is the expected duration of the chip’s mass production. Opting for a chip that is not the newest but still has at least 3-4 years of mass production ahead is prudent. Avoiding the newest CPUs and selecting one that has been popular for two years allows engineers to be already familiar with it, and other companies have undergone the initial mass production runs. The prolonged mass production of a chip contributes to amortizing the initial development into future runs and postpones the need for redevelopment, offering economic benefits.
While there are additional considerations in the CPU selection process, the high-level points encompass:
1. Define the product specs.
2. Identify a CPU that aligns with those specs without unnecessary capabilities.
3. If faced with numerous options, delve into the niche performance details and intangible values of each CPU relevant to your product.
Managing the CPU selection process is an integral part of the development work conducted by Hatch for all new projects. After perusing this article, it is hoped that you now possess a better understanding of the intricacies involved in this selection process. Should you have any questions, feel free to reach out.
