Interview experience
BARC Interview Questions and Answers (2026)
Overview
Section titled “Overview”BARC recruits Scientific Officers through the OCES/DGFS training scheme, screening candidates via either its own Computer-Based Test or a valid GATE score, with final selection resting entirely on a single interview.
BARC interview process at a glance
Section titled “BARC interview process at a glance”| Round | Duration | What it tests |
|---|---|---|
| Screening - BARC CBT or valid GATE score | Varies (CBT is discipline-specific, curriculum-level) | Core subject fundamentals and problem-solving (CBT), or your GATE score directly |
| Selection Interview | 30-45 min | Core technical depth, project discussion, research/scientific suitability |
| Medical fitness examination | - | Physical fitness for the role |
Screening - BARC CBT or GATE score
Section titled “Screening - BARC CBT or GATE score”BARC offers two parallel screening channels for the same shortlist: its own Computer-Based Test (discipline-specific, run across 50+ Indian cities, testing B.E./B.Tech-level core subjects only - no general English, GK, or aptitude sections), or a valid GATE score in the matching subject from the current or two preceding GATE years. Either route only shortlists you; the score itself doesn’t carry into final selection.
Common questions
- Core-subject fundamentals at B.E./B.Tech curriculum depth (CBT route) - e.g. core CS subjects for Computer Science candidates
- Problem-solving questions within your discipline, without general-aptitude or GK sections
Selection Interview
Section titled “Selection Interview”The entire final selection rests on this one interview, plus medical fitness - screening scores don’t carry forward. Panels go deep on your core discipline and cross-question your final-year project closely, alongside gauging your fit for a research-oriented government scientific career.
Common questions
- Core technical questions from your engineering/science discipline
- Detailed walkthrough of your final-year project or thesis - methodology, results, design choices
- Why a research career at a government scientific establishment like BARC over industry
- Questions assessing scientific curiosity and suitability for long-term, mission-driven research work
Sample answer frameworks are on the BARC HR interview questions page.
Why screening scores don’t carry into selection
Section titled “Why screening scores don’t carry into selection”Unlike most PSU or corporate pipelines where your test score contributes to a final weighted merit, BARC’s OCES/DGFS process treats the CBT or GATE score purely as a gate to reach the interview - once you’re shortlisted, that score is dropped and selection is decided entirely by interview performance (plus clearing the medical exam). That makes the interview far higher-stakes than in a typical weighted-scoring PSU process, so treat screening as a hurdle to clear, not something to bank marks on.
Common technical interview questions and answers
Section titled “Common technical interview questions and answers”Q: What are the four necessary conditions for deadlock, and how is deadlock handled?
Deadlock needs all four of Coffman’s conditions to hold simultaneously: mutual exclusion (a resource is non-shareable), hold and wait (a process holds one resource while requesting another), no preemption (a resource cannot be forcibly taken), and circular wait (a cycle exists in the wait-for graph). Break any one and deadlock cannot occur - requesting all resources at once breaks hold and wait, and imposing a global ordering on resource acquisition breaks circular wait. Avoidance uses the Banker’s algorithm, which grants a request only if the resulting state is safe, meaning some sequence exists in which every process can finish. Detection plus recovery instead lets deadlock happen, finds cycles in the wait-for graph, and kills or rolls back a victim - which is what most general-purpose operating systems effectively do by ignoring the problem.
Q: Explain paging, the TLB, and how effective memory access time is calculated.
Paging divides the logical address space into fixed-size pages and physical memory into frames of the same size, with a page table mapping page numbers to frame numbers, which eliminates external fragmentation at the cost of some internal fragmentation in the last page. A plain page table means two memory accesses per reference - one for the table, one for the data - so a Translation Lookaside Buffer caches recent page-table entries. If the TLB hit ratio is h, TLB access time is t, and memory access time is m, effective access time is h(t + m) + (1 - h)(t + 2m) for a single-level page table. With a hit ratio of 0.9, t = 20 ns and m = 100 ns, that works out to 0.9(120) + 0.1(220) = 130 ns. Multi-level page tables reduce the memory the table itself consumes, but add one more memory reference per level on a TLB miss.
Q: What are the ACID properties, and what anomalies do the SQL isolation levels allow?
Atomicity means a transaction executes fully or not at all; Consistency means it moves the database from one valid state to another respecting all constraints; Isolation means concurrent transactions do not observe each other’s intermediate state; Durability means committed changes survive a crash, typically guaranteed by a write-ahead log flushed before commit. The four SQL isolation levels trade isolation against concurrency: Read Uncommitted permits dirty reads, Read Committed prevents those but permits non-repeatable reads, Repeatable Read prevents those but permits phantom rows, and Serializable prevents all three. Stricter levels need more locking - Serializable typically requires range or predicate locks - so throughput drops as isolation rises.
Q: Why do databases use B+ trees for indexing rather than binary search trees or hash tables?
A B+ tree is a balanced multi-way tree in which every internal node holds many keys and children, so a node maps neatly onto a disk block or page. Because the branching factor is in the hundreds, the height is roughly log base b of n and a lookup on millions of rows costs only three or four disk reads, whereas a binary search tree of the same size would cost around 20. In a B+ tree all actual records live in the leaves and the leaves are chained in a linked list, which makes range queries and ordered scans a single sequential walk. A hash index gives O(1) equality lookup but cannot answer range or ordering queries at all, which is why B+ trees remain the default.
Q: Compare TCP and UDP, and explain how TCP achieves reliability.
TCP is connection-oriented and reliable: it establishes a connection with a three-way handshake (SYN, SYN-ACK, ACK), numbers every byte with sequence numbers, acknowledges receipt, retransmits on timeout or duplicate ACKs, and delivers bytes to the application in order. UDP is connectionless with an 8-byte header, no handshake, no acknowledgement, no ordering, and no congestion control - so it is faster and used for DNS, video streaming, and real-time voice where a late packet is worse than a lost one. TCP also does flow control with a receiver-advertised sliding window, so a fast sender cannot swamp a slow receiver, and congestion control with slow start, congestion avoidance, fast retransmit, and fast recovery, so senders back off when the network is loaded. The cost is per-connection state, handshake latency, and head-of-line blocking.
Q: State the master theorem and use it on merge sort and binary search.
For a recurrence T(n) = aT(n/b) + f(n) with a at least 1 and b greater than 1, compare f(n) with n raised to log base b of a. If f(n) grows polynomially slower, the answer is dominated by the leaves and T(n) is theta of n to the power log base b of a. If they grow at the same rate, T(n) is theta of that same term multiplied by log n. If f(n) grows polynomially faster and satisfies the regularity condition, T(n) is theta of f(n). Merge sort is T(n) = 2T(n/2) + O(n), where log base 2 of 2 is 1 and f(n) is n, so it falls in the second case and gives theta of n log n. Binary search is T(n) = T(n/2) + O(1), where log base 2 of 1 is 0, again the second case, giving theta of log n.
Q: What is instruction pipelining, and what limits the speedup you actually get?
Pipelining overlaps the stages of successive instructions - a classic five-stage RISC pipeline runs fetch, decode, execute, memory access, and write-back concurrently on five different instructions - so throughput approaches one instruction per cycle even though each instruction still takes five cycles end to end. The ideal speedup equals the number of stages, but three classes of hazard erode it. Structural hazards arise when two stages need the same hardware, solved by duplicating units or splitting instruction and data caches. Data hazards arise when an instruction needs a result not yet written back, mitigated by forwarding or bypassing and, where that is not enough, by stalling. Control hazards come from branches, where the pipeline does not yet know the next address, and are handled with branch prediction and delayed branches - a mispredict costs a full pipeline flush.
Q: How should you present your final-year project when the panel cross-questions it?
BARC panels probe the project harder than almost anything else, so structure the answer as problem, approach, alternatives rejected, results, and limitations. Be able to justify every design choice quantitatively - why that algorithm rather than a simpler baseline, what the measured improvement was, and what your test or validation setup was. Own your specific contribution honestly rather than describing the whole team’s work in the first person, because the follow-up question is always about a detail only the person who wrote it would know. Finally, state at least one limitation and what you would do differently with more time; panels read that as scientific honesty rather than weakness, which is precisely the trait a research organisation is screening for.
Frequently asked questions about BARC interviews
Section titled “Frequently asked questions about BARC interviews”What is the BARC interview process for freshers?
BARC hires Scientific Officers through its OCES/DGFS training-scheme route, not a corporate pipeline: 1. Screening - either BARC’s own Computer-Based Test (CBT, discipline-specific, held across 50+ Indian cities) or a valid GATE score in the matching subject; candidates choose whichever channel they qualify through. 2. A Selection Interview for shortlisted candidates. 3. Medical fitness examination. Screening scores (CBT or GATE) are used only to shortlist - final selection rests entirely on interview performance plus medical fitness.
Does BARC require its own written exam, or can I get in through GATE alone?
Both routes work. BARC runs its own Computer-Based Test as one screening channel, but candidates with a valid GATE score (from the current or two preceding GATE years) in the same subject as their qualifying degree can be screened directly on that score instead - skipping BARC’s CBT entirely. Either way, screening only shortlists you; it doesn’t count toward final selection.
What questions are asked in the BARC interview?
The Selection Interview goes deep on your core discipline at B.E./B.Tech curriculum level - for Computer Science candidates, for example, fundamental CS subjects and problem-solving rather than general aptitude or GK. Expect detailed cross-questioning on your final-year project, and questions probing suitability for a research-driven, mission-critical government scientific role.
How many rounds are there in the BARC interview process?
Effectively two stages: a screening step (BARC’s own CBT or a valid GATE score - either qualifies you) and a Selection Interview. There’s no separate HR round - motivational and fit questions are folded into the same interview alongside technical depth, since selection is scientific-suitability-based.
How should I prepare for BARC interviews?
If you’re going through the CBT route, revise your core discipline at B.E./B.Tech depth - the exam skips general English, GK, and aptitude entirely and tests only subject fundamentals and problem-solving. If you’re using a GATE score to screen in, focus your prep on the interview itself: your final-year project, core-subject depth, and a clear reason for wanting a research career at a government scientific establishment.
What is OCES/DGFS at BARC?
OCES (One Year Orientation Course for Engineering Graduates) and DGFS (DAE Graduate Fellowship Scheme) are BARC’s training programmes for recruiting Scientific Officers into the Department of Atomic Energy - OCES is for direct entrants, DGFS sponsors selected postgraduate study before absorption. Both share the same screening-plus-interview selection process; which scheme you’re considered under depends on your qualifying degree.