Let’s start with the conclusion:If you are exceptionally talented, have a strong background, and possess great ambition, then without a doubt—design (DE) is your first choice. However, if you are an ordinary person looking to secure a position in a large company, earn a good salary, and live a decent life, then verification (DV) may be a more realistic path.
Don’t rush to criticize me; let me finish my point.
When it comes to the “ceiling”, design undoubtedly has a higher one. This is indisputable.
First,architects and project leaders are mostly those who have risen from design positions. How many system architects have you seen who come from a verification background? Very few. It’s not that it’s impossible, but the probability is too low. In verification, we constantly focus on test cases, coverage, and UVM sequences; no matter how skilled we are, we are at most a “verification expert” or “methodology expert.” On the other hand, designers are involved in drawing circuits, timing adjustments, and power consumption considerations, where every decision directly impacts whether the chip can be taped out and mass-produced. This level of technical authority makes it easier for them to reach the top of the technical ladder.
Second,the aspects that designers care about can directly translate into financial gains. If the area increases, costs go up. If power consumption is high, customers won’t buy. If timing doesn’t converge, tape-out fails directly. If temperature reverses, chips burn during mass production. These are not just things written on a PowerPoint; they represent real financial losses. In verification, if we succeed, management says, “That’s expected”; if we miss something, they say, “Why weren’t you more careful?” You might work hard to improve coverage from 98% to 99.5%, but your boss might not even glance at it. However, if a designer successfully tunes a critical path, the entire project schedule can advance by a week, which is a significant contribution.
Third,in leading foreign enterprises or large companies, the technical level ceiling is higher for design than for verification. There are plenty of P8 verification engineers, but above P9, the vast majority come from design backgrounds. If you look at the list of titles like Principal Engineer or Fellow, how many started in verification? None. It’s not discrimination; it’s reality. Resource allocation, technical weight, and authority naturally place design in a more central position.
But
what I’m saying is about the “ceiling.”
The question is,can you really reach that ceiling?
Let’s ask ourselves: are you a top student from prestigious universities like Xidian, Chengdu, Fudan, or Tsinghua? Have you won first prize in EDA competitions during your undergraduate studies, published papers at top conferences during your graduate studies, or interned at Huawei HiSilicon or AMD working on core modules? If not, the threshold for entering a large company as a designer may be much higher than you think.
As for verification?It remains the most realistic stepping stone for ordinary students to enter leading IC companies. Many companies have relatively relaxed educational requirements for verification, with a high demand for hiring and positions, especially for skills like UVM and SystemVerilog, which can be learned in a few months. You might say this is “low-end”? I don’t deny it, but can you get into a design position? If you can’t, how can you talk about development?
Moreover, don’t underestimate the “system perspective” of verification.
While we don’t draw circuits, we look at the entire module’s data flow, protocol interactions, and state machine transitions. A DE working on PCIe design might be stuck with PCIe for their entire career; however, a DV working on PCIe verification can switch to Ethernet, MIPI, DDR, or CPU core after completing one project. Because we don’t care about the underlying timing, only the function and protocol,the cost of switching modules is much lower than for design.
I have examples around me:
- A, who has a master’s degree from Xidian, has been working on CPU front-end design for three years and is still modifying the CPU’s branch predictor;
- B, a classmate from the same year, worked on CPU verification, switched to SOC verification after two years, and in the third year, changed jobs to work on data flow verification for AI chips, doubling their salary.
Who do you think has more flexible development?
Furthermore,most people will only ever be a Senior Engineer, with a maximum annual salary of around 700,000 to 800,000. In this case, the quality of life, work intensity, and promotion speed between design and verification are not that different. If you insist on saying design has “more technical content,” that’s fine, but you also have to deal with more overtime, higher pressure, and greater risk of blame. Is it worth it? You decide.
Finally, let’s talk about some realities:
- Salary growth: When starting out, design may be a few thousand higher than verification, but after three years, it generally levels out. After five years, who advances faster depends on projects, luck, and alliances, not positions.
- Career planning: If you want to pursue a technical path, design has an advantage; if you want to transition to management or systems, verification offers a broader perspective; if you want to switch to software, algorithms, or system architecture, the abstract skills from verification are easier to transfer.
- Industry trends: As chip complexity increases, the workload of verification has surpassed 60%. The high demand for hiring indicates a significant need, but it also means intense competition. In the future, those who will survive are the“senior verification engineers” who understand systems, protocols, methodologies, and can write some Python/C++ automation, rather than just being cogs who can set up platforms and write sequences.
Therefore, my advice is:
👉 If you are a top student with a strong background and aspire to become a chip architect or a technical expert, choose design and go for it.👉 If you are an ordinary student from a 985/211 university, looking to enter a large company, earn a good salary, and maintain career flexibility, verification is a more pragmatic choice.
Don’t just focus on the “ceiling”;first, consider whether you can reach the second floor.
When you truly stand at the door of P9 one day, it won’t be too late to ponder “why verification doesn’t advance.”
Until then,master UVM, understand the protocols, and develop the ability to work diligently—this is the only path for ordinary people to succeed.
Let’s encourage each other.