The VLSI industry has a number of upcoming challenges, keep Moore’s law going. At the same time, the VLSI industry needs more engineers and designers. Specifically for digital VLSI design, the hardware for any software algorithm is going to determine the efficiency and the performance. Let’s look at it from different sides. Is there a future for a VLSI design engineer?
The future of VLSI is bright. VLSI, the technology, nears the limits of shrinking the transistors in the X- and Y-direction. But 3D stacking, different silicon substrates and interposers (interconnecting silicon inside the same package) is a path forward. Analog, digital or mixed signal is in demand everywhere. Decades ago, a car had just a few electronic components, mainly analog. Nowadays a car has thousands of chips inside. Automation and smart devices are the future. Robots, drones, they all need hardware and software to operate. Hence, there will be a wider variety of solutions needed and this is good news for the VLSI industry. There will be demand.
Which kind of demand?
R&D is an expensive for a company. The pipeline for chips needs to be loaded to make money in the future. While the stock price is based on the current cost and revenue. The revenue now is based on R&D of two, three years ago. One of the best tricks in the top management book to boost the stock price and to make their stock options worth more, is to cut R&D. They don’t care about the revenue in three years, they care about now. Many of these successful corporate ladder rockets cut R&D versus revenue so they look good. The they go to another department, go higher on the ladder or go to another company. R&D needs to be low cost. It means excel sheets with these columns:
- Number of resources.
- Monthly cost per resource.
- Average cost per resource.
The pressure on employee salaries is high, for VLSI design engineers. The same goes for freelancers and consultants. The race to the bottom lowers the average cost of resources. But the quality suffers quite heavily. So, more resources are needed. This requires more budget. Big companies complain that a sizeable ASIC in todays smallest technology node costs more than 500 million USD. The reason is exactly the push to low cost, low quality that makes the budget explode. It is counter intuitive to explain that selecting experts in key positions and paying them well seems to increase the budget. However, if you compare a team led by experts, it is clear they need a lot less resources to complete the same task. It is an optic illusion that low cost will lower the overall budget of a project.
The efficiency in a big company for VLSI design is bad, sometimes only 10%. Chip startups are able to beat the big companies to the market with 10% of the budget. Incompetence is inherent to a big company. It is no secret for people that worked a few years in those companies. Hence, the way to look at a VLSI design engineer path is that there is demand and that will not disappear soon. The knowledge of hardware is the foundation for all above it, like firmware and software. Enter this field with the purpose of a skill set. To learn and understand. And move on. This knowledge is much more valuable to do something else. It is not a career in itself. It is a path towards something else.
Interesting as well: why good employees always leave.