Hardware Engineer at @Microsoft His Journey, Interview Exp 45+LPA & Career Tips | ft. Somit Jaiswal

Main ideas, concepts, and lessons

Guest background & motivation

• Sumit is a hardware engineer at Microsoft.
• Education path
  • B.Tech Electrical Engineering from an AKTU-affiliated college (AK Shri Ram Swaroop College). He started developing interest in hardware in the 3rd year.
  • Unilever: worked 1 year in marketing, which helped improve communication/personality and provided early industry exposure when core-core opportunities were limited.
  • NIT Jalandhar for M.Tech, achieved via GATE, specializing in Electronics and Communication Engineering, with focus on Signal Processing and Machine Learning.
• Career interest triggers
  • Early decision to avoid “software-only” and pursue core hardware/VLSI.
  • During his NIT period, the semiconductor boom and incentives for semiconductor setups reinforced his direction toward semiconductors/VLSI.

What hardware engineering means at Microsoft (in his role)

Hardware is described as broad. Major categories include:

• Chip design
  • Example from an internship at NXP Semiconductors:
    • RTL design: writing logic in code; tools convert logic into chip layout (the EDA flow).
• Azure / cloud server validation
  • He works in Azure hardware / infrastructure teams.
  • He validates new data-center server rack setups that will be deployed in the next 1–2 years.
  • He validates end-to-end system readiness for AI workloads, including:
    • Machine learning training
    • LLM training
    • Inference
  • Server rack components mentioned:
    • CPU, GPU
    • Drives/SSD storage
    • Interconnects (communication between components)

GPU-focused experience (internship/benchmarking)

In a previous Microsoft internship, he:

• Did GPU profile benchmarking for projects using NVIDIA GPUs.
• Ran multiple workload types such as:
  • Graphics rendering
  • Gaming-like workloads
  • Matrix multiplication / linear algebra
  • LLM inference
• Produced benchmark results assessing system performance.
• The work increased his interest by showing GPU architecture “how GPU works” in detail.

How AI is used to accelerate hardware validation work

• His core work is validation/testing: run tests, detect bugs, fix issues, and retest until the system is production-ready for customer data centers.
• He emphasizes that AI helps fast-track workflow, especially through Microsoft Copilot:
  • Teams define:
    • What test cases/workloads to run
    • How workloads should be implemented
    • What kinds of bugs should be detected
  • Then teams use Copilot (Microsoft internal) to generate the scripts for running workloads on the server side.
  • Engineers still debug scripts and run tests, but AI reduces time spent on manual scripting.

Copilot vs ChatGPT: different use cases

He frames them as serving different environments:

• ChatGPT: more B2C oriented, general-purpose use.
• Copilot: B2B oriented and tightly integrated into Microsoft Teams and Microsoft Outlook.
• Copilot benefits from access to internal documents (within the organization), improving usefulness for enterprise workflows.

Explanation of RTL design

• RTL = Register Transfer Level.
• Process described:
  • Write logic code corresponding to logic-gate behavior (gates producing outputs based on inputs).
  • Use EDA tools (he mentions Siemens) to generate layout.
  • Proceed to back-end physical design, verifying constraints such as:
    • Electromagnetic interference constraints
    • Wire spacing/placement constraints

Advice for students who want hardware roles (especially at Microsoft)

• Foundational requirements
  • Earn a degree in Electronics or Electrical Engineering.
• Knowledge areas he recommends
  • Strong base in core degree subjects
  • Embedded systems fundamentals
    • Example: a washing machine—buttons + automatic process = embedded logic on a chip
  • C++ basics
  • Python scripting (helpful for tooling/scripts)
  • Computer architecture (to understand/optimize performance)
  • Core electronics
  • Low-level languages/tools used in hardware/VLSI workflows
  • Digital design (recommended; whether it appears in curriculum depends on program year/track)
• Learning approach
  • Build breadth from degree foundations, then deepen into architecture, digital/VLSI, and low-level toolchains.

Student journey lessons

• COVID impact
  • His first/second-year experience was heavily impacted by COVID, limiting normal exposure.
• GATE path
  • From third year: began GATE preparation using YouTube + course curriculum.
  • Qualified GATE in fourth year and achieved campus placements:
    • Selected in three companies (two core + one marketing at Unilever).
• Personal habits & interests
  • Sports/activities: dance, chess, table tennis, pool
  • Later free time: LinkedIn scrolling + learning latest tech; sometimes guitar, and gym
• Mindset/psychology
  • Emphasizes destiny + hard work, but above all: execution matters, not only planning.
  • Notes negative emotions when outcomes aren’t as expected: self-doubt, negative self-talk, and effects on surroundings.
  • Comparison: healthy comparison is okay, but don’t compare paths as identical timelines—everyone reaches milestones differently.

Hardware job market companies mentioned

Examples of companies hiring in hardware/semiconductors:

• NXP
• NVIDIA
• Qualcomm
• TSMC
• Samsung
• Nokia
• Analog Devices
• Texas Instruments

He also notes that Microsoft is relevant, though less popularly known compared to some others.

Microsoft interview experience + strategy

• Application timing
  • Interviews may be scheduled ~1 year before hiring.
  • He applied early after starting masters to position himself.
• Interview structure
  • Two rounds, about 1 hour each
    • Round 1: technical
    • Round 2: technical + managerial
• Topics asked (as he recalls)
  • Computer architecture
  • Digital electronics
  • Basics of VLSI design
  • Power electronics
    • He believes it gave him an advantage due to his electrical background and B.Tech project:
      • converters: DC→AC, AC→DC, DC→DC
    • He links power electronics to data centers because power supply is critical.
• How to respond
  • You don’t need to answer everything.
  • Aim to answer ~70% confidently, and start/structure the solution even if you’re not sure about everything.
  • For unfamiliar problems, interviewers value your approach and reasoning.
• Preparation/fit with the company
  • Before the interview: check Microsoft’s company page, and text/latest innovation sections to learn what technologies match your interests.
  • During the end interview: ask questions to show engagement.
  • He specifically mentioned:
    • Learning about Microsoft work in quantum computers (including a reference to a “topological cubic quantum chip”)
    • Understanding how AI drives company growth
• Questions he asked
  • What success looks like in the role/career if selected
  • How the interviewer evaluates performance and what feedback would help improve

---

Methodologies / instruction-style guidance (detailed bullets)

A) Using AI (Copilot) to accelerate hardware validation scripting

• Define the workload/test plan
  • Decide which test cases to create (e.g., for CPU validation)
  • Decide how the workload will be implemented
  • Identify what bugs the tests should detect
• Once the rack/system setup exists:
  • Prepare to run scripts on the server side
• Use Copilot to generate the script
  • Explain/author the workload structure to Copilot
  • Let Copilot generate the script portion
• Human steps still required
  • Debug generated scripts
  • Run tests on the server
  • Fix issues
  • Re-validate until stable/ready for data center deployment

B) How to become competitive for hardware roles (student roadmap)

• Get the right degree
  • Electronics/Electrical Engineering
• Build a strong base
  • Know the core subjects from your curriculum well
• Learn embedded systems
  • Understand embedded logic as chip-based controllers for real devices (e.g., appliances)
• Programming skills
  • C++ basics
  • Python scripting (for tooling/workflows/scripts)
• Study performance fundamentals
  • Computer architecture
• Add hardware/VLSI learning
  • Electronics + awareness of low-level tool/language workflows used in front-end VLSI design
  • Digital design (if not covered, learn via electives/self-study)
• Target role-relevant areas
  • RTL, VLSI flow, and how architecture connects to performance

C) Microsoft interview approach (practical checklist)

• Timing
  • Start early (often ~1 year before the hiring window)
• Prepare likely topics
  • Computer architecture
  • Digital electronics
  • VLSI basics / front-end knowledge
  • Power electronics (especially if you have an electrical background)
• During Q&A
  • You don’t need to answer everything
  • Aim to answer ~70% correctly and show confidence
  • Always begin with your solution/outline your approach
  • For unfamiliar problems, explain your reasoning process
• Company-specific prep
  • Review Microsoft’s latest innovations/technology pages
  • Study topics aligned with your background
• End-of-interview questions
  • Ask what success looks like and how feedback/evaluation works
  • Use relevant Microsoft tech examples to demonstrate engagement

---

Speakers / sources featured

Speakers

• Ram (podcast host; asks questions)
• Sumit (Sumit Jaiswal in title) — hardware engineer at Microsoft; guest interviewee

Sources mentioned

• NXP Semiconductors (RTL design work during internship)
• Unilever (marketing role)
• Microsoft Copilot (used/mentioned)
• ChatGPT / OpenAI (compared against Copilot)
• NVIDIA GPUs (workloads/benchmarking)
• Siemens EDA tools (RTL/VLSI flow explanation)
• GATE (used for NIT admission and preparation)
• Microsoft Teams / Microsoft Outlook (Copilot integration mentioned)
• Microsoft (interview + quantum/AI context)