Are you creating software? If so, did you spend a few minutes to think about the hardware?
With the right package, your software product’s worth has major potential to climb the ranks.
Trends in the industrial computing market:
The hardware market is currently faced with several challenges that can change it beyond recognition.
First, cloud services are taking up more and more volume. The Network Functions Virtualization (NFV) technology is one step away from making a massive entry into the market, and is “threatening” to turn all client-side hardware to virtual machines in the cloud.
Even so, you are still required to have an embedded hardware solution to serve as local series providers for the client side. Examples include embedded computers with integrated screens to operate as beverage machines, SBC cards like COMEXPRESS to perform calculations, an in-vehicle computer connecting and receiving data from the Internet of Things (IoT), a local NAS storage platform, or a client-side NFV unit loaded by storage and security applications. In essence, hardware isn’t going completely out of business.
The market today is led mostly by software companies, and these software companies can have two options to view their product:
- As a storage/cyber/networking/etc. solutions as software only.
- As a complete solution that contains software and hardware, aka a “BLACK BOX.”
These companies are faced by a number of strategic decisions:
- Is the “BLACK BOX” solution combining software and hardware worthwhile?
- What is the added value of marketing and selling complete solutions?
- Is specific hardware required, or maybe it is right to rely on the standard hardware on the market?
- Should I choose general purpose processing or ARM?
So what should a software company do: software only packages or all-in-one software + BlackBox solutions?
This dilemma is one of the largest of its kind that software companies face before releasing a new product to the market. Both pros and cons inevitably exist, and eventually each company has its own priority.
Here are a few of the decisions that need to be made:
- For new start-up companies, it’s not easy to sell expensive Branded Box Solutions. In this case, most end customers will prefer to allocate virtual servers instead of real hardware.
- Complete solutions including both software & hardware require appropriate logistical adjustments, which can possibly lengthen launching expectations and perhaps lower your overall net income since you’re now working with another company.
On the other hand, to-market and to-sale complete solutions have significant benefits:
- Selling Branded Box Solutions as service providers increases the value of the product, making it easier for marketing and pricing.
- Known and proven hardware configuration reduces the feasibility of performance/compliance issues, and also allows more reliable research and monitoring performance.
- Even the most efficient virtual machine will be limited in terms of the data transfer rate and the efficiency of real hardware.
Okay, so now you’re convinced that the Black Box Solution is the better bet (even though it might be a little more expensive). What should you do with the “dedicated hardware solution” or “shelf product” dilemma?
This question doesn’t get the proper attention. Naturally, software development is a step or two away from the hardware platform selection. As a result, the hardware selection and adjustment part of the progress don’t get enough attention until the final stages of the development process.
In order to make the right decisions, hardware selection considerations should be intertwined within the software development processes.
The prevailing view is that hardware should be as standard as possible in order to avoid dependence between manufacturers and hardware creators. While this consideration is legitimate, it doesn’t take additional important parameters into account, those of which include:
- Configuration control (a point that I’ve expanded on in a previous post)
- Hardware performance and software resources
Now let’s see how hardware design can help lower the cost of the product and give added value:
Design to Cost
My goal in writing this post is to offer an additional perspective, to present other action options. and finally to make an informative decision by considering where the future is taking us.
First and foremost, the engine that drives the project is costing viability. The questions here are whether my final product will be attractive enough to be sold, and how can I lower the costs.
It’s important to understand that shelf products are built according to the broadest common denominator.
This means that most likely, the nearest shelf product to our needs may contain not required abilities and components (things that increase the prices of hardware) or missing features. Therefore, sometimes dedicated hardware design becomes a real option.
Of Course, design to cost mainly refers to big projects and not for few samples… However, what was once believed to be relevant only for thousands of units (MOQ), is now appealing manufacturers for much smaller quantities, due to new production technology & technique.
When the hardware is part of the considerations during software development, the hardware can be optimized to whole system activity. Dedicated hardware design uses only the necessary components on the board, and on the other hand allows existing components run in harmony, helping to achieve maximum performance with minimum cost.
It’s necessary to understand that the hardware contains more than a few lines of software code in the BIOS and in the drivers, and that these code lines have greater significance maximizing hardware performance.