Accelerating the Production of Semiconductor Engineers for the Global Job Market
The target of developing 15,000 semiconductor talents should not stop at the number of training participants. The measure of success must be more concrete: can an Indonesian fresh graduate present a portfolio that makes global semiconductor companies understand what they are truly capable of doing?
The Ministry of Manpower can utilise Competency-Based Training (PBK) with universities acting as training nodes. The programme could begin with 200 engineers every month, for example, through four universities with 50 participants each in an intensive three-month programme.
However, the output should not be a certificate of attendance. A participant in Digital IC and SoC, for instance, must leave with a processor they understand down to the RTL level. That processor must not merely have been simulated.
Participants must create a verification environment, encounter intentionally inserted defects, identify the first divergence, fix it, run regressions, implement it on an FPGA, and then make that processor boot and run its own firmware. Their portfolio could include architecture descriptions, RTL, testbenches, instruction tests, synthesis reports, FPGA resource utilisation, timing results, boot logs, firmware source, defect-debug traces, and regression evidence.
When a recruiter asks, “What have you worked on?”, the answer should no longer be “I have studied RISC-V.” They should be able to say: “This is my processor. This is the RTL. This is how I verified it. This is the defect I diagnosed. This is the FPGA implementation. And here is my firmware booting on it.”
Analog engineers must have equally strong evidence. Participants could build a PLL (Phase-Locked Loop) from requirement to verification. They must determine the frequency range, understand the reference, phase-frequency detector, charge pump, loop filter, VCO, and divider; perform hand analysis and independent modelling; and then run SPICE.
Their portfolio would demonstrate lock acquisition, lock time, frequency error, stability, PVT corners, Monte Carlo variation, and failure cases. At an advanced level, participants could show layout, parasitic extraction, and the differences between pre-layout and post-layout.
Thus, when global companies view such a portfolio, the candidate does not merely write “understands PLL” on their CV. They possess evidence that they have designed, broken, diagnosed, and verified a PLL.
The Firmware and FPGA tracks must also produce something demonstrable. Participants receive a processor/SoC, understand reset and memory maps, create boot code, control MMIO, timers, and interrupts, write peripheral drivers, and then perform hardware-software debugging on an FPGA. The final output could be a processor → boot ROM → firmware → peripheral → measurement, complete with source code and evidence.
This is the difference between a training certificate and an engineering portfolio. PBK enables this approach because its orientation is indeed competency. Universities provide lecturers, laboratories, and mentoring. The industry would not need to guide thousands of participants through transistors or C programming from the very beginning.
Industry engineers can concentrate on curriculum reviews, engineering challenges, design reviews, and advanced participants. Incentives for participants are also vital. The 2026 Ministry of Manpower government assistance guidelines already recognise a training allowance of Rp50,000 per person per day for PBK/OJT, subject to programme requirements.
For an illustration of 60 effective days, this amounts to approximately Rp3 million per participant. While not an automatic right, it shows that the supporting instruments are available. The ultimate goal can even transcend domestic industry needs.
Semiconductor engineering is part of the global labour market. RTL, verification, FPGA, embedded firmware, analog design, and various engineering workflows use international technical languages, methods, and tools.
Therefore, PBK portfolios should be created in English from the outset, using structured repositories, requirements, test reports, design reviews, and reproducibility evidence.
Imagine if every month Indonesia produced 200 candidates with portfolios such as:
“My RISC-V processor boots my firmware.”
“My PLL locks, and here is its PVT and Monte Carlo evidence.”
“Here is the defect I found, the first divergence, the repair, and the passing regression.”
That would mean a potential 2,400 engineer portfolios per year. Certification remains important, but to enter the global job market, certificates must be reinforced by something much harder to forge: engineering evidence.
Consequently, the target of 15,000 semiconductor talents should be translated into 15,000 verifiable portfolios. The Ministry of Manpower should not only help create trainees. Through university-based PBK, Indonesia has the opportunity to build a pipeline of semiconductor engineers who can demonstrate to the global industry: This is what I can actually build.