NVIDIA interview questions

Showing 61–90 of 95 results

• Hash-Partition an Array on the GPU in C++ and Return Partition OffsetsMedium

  A GPU programming question that asks you to hash-partition an array of keys into P partitions in CUDA C++, making each partition contiguous and returning each partition's starting offset. Follow-up questions test GPU optimization: atomic contention, memory coalescing, shared memory use and the number of passes over global memory.

  NVIDIASoftware EngineerSoftware Engineering FundamentalsTechnical Screen
• How AI Agents Fit Into Your Research Workflow: Ideas, Experiments, and Paper WritingMedium

  A research internship interview question about how AI agents change a researcher's daily workflow across generating ideas, running experiments, and writing papers. It tests concrete delegation choices, how agent output is verified, awareness of failure modes such as fabricated references, and judgment about confidentiality and reproducibility.

  NVIDIAApplied ScientistBehavioral & LeadershipTechnical Screen
• How AI Coding Tools Change Software Development and When to Trust Their OutputMedium

  An open-ended question from a director round asks how AI is changing software development and whether engineers should trust AI-generated code. It tests a balanced position on benefits and failure modes, how generated code should be verified, team practices for review and security, and how to measure real impact.

  NVIDIASoftware EngineerSoftware Engineering FundamentalsOnsite
• How would you optimize large-scale training/inference?Medium

  This problem tests your understanding of performance bottlenecks in large-scale machine learning systems, requiring you to prioritize optimization strategies across memory, compute, and communication dimensions.

  NVIDIASoftware EngineerML System DesignTechnical Screen
• Implement 2D convolution using NumPy slicingMedium

  This question evaluates a candidate's proficiency with array manipulation and vectorized numerical computation using NumPy, focusing on 2D convolution concepts such as slicing, broadcasting, and filter application.

  NVIDIASoftware EngineerData Manipulation (SQL/Python)Technical Screen
• Implement CUDA-tiled matrix multiplication and explain architectureHard

  This question evaluates a candidate's competency in CUDA GPU programming, parallel algorithms, and performance engineering for FP32 matrix multiplication, covering tiling strategies, memory hierarchy (global/shared/register), synchronization, numerical precision, and occupancy/resource analysis.

  NVIDIAData ScientistCoding & AlgorithmsHR Screen
• Implement a Python test harnessMedium

  This interview question evaluates SQL or pandas logic, joins, grouping, window functions, null handling, edge cases, and validation in a realistic interview setting. A strong answer for Implement a Python test harness states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

  NVIDIASoftware EngineerData Manipulation (SQL/Python)Online Assessment
• Implement a Thread-Safe Reference-Counted Smart Pointer in C++Medium

  Implement a reference-counted smart pointer class template in C++ from scratch, with copy and move construction and assignment, reset, dereferencing and a use count. It tests ownership rules, self-assignment and exception safety, atomic reference counting with correct memory ordering, and extensions such as weak pointers, base-class conversion and reference cycles.

  NVIDIASoftware EngineerSoftware Engineering FundamentalsOnsite
• Implement and Explain a C++ Smart PointerMedium

  Tests understanding of C++ memory management, RAII, explicit ownership semantics, and the Rule of Five by implementing a reference-counted shared smart pointer.

  NVIDIASoftware EngineerSoftware Engineering FundamentalsTechnical Screen
• Implement core graph algorithms for graphicsMedium

  This interview question evaluates algorithm design, data structures, correctness, complexity, edge cases, and implementation details in a realistic interview setting. A strong answer for Implement core graph algorithms for graphics states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

  NVIDIASoftware EngineerCoding & AlgorithmsOnline Assessment
• Implement short algorithms on logs, grids, and stringsHard

  This problem tests your ability to handle multi-stage data processing, specifically focusing on greedy interval selection in logs, stack-based parsing of nested strings, and constrained combinatorial optimization on a grid.

  NVIDIASoftware EngineerCoding & AlgorithmsOnsite
• Introduce yourself for a senior roleMedium

  This question evaluates leadership, communication, ownership, and impact-articulation skills for a senior software engineer, focusing on the ability to concisely present scope, technical and business outcomes, and future alignment.

  NVIDIASoftware EngineerBehavioral & LeadershipTechnical Screen
• Market a new AI techniqueMedium

  This interview question evaluates requirements, scale assumptions, API/data design, architecture, trade-offs, failure modes, and rollout in a realistic interview setting. A strong answer for Market a new AI technique states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

  NVIDIASoftware EngineerSystem DesignTechnical Screen
• Merge Overlapping Time IntervalsMedium

  Practice a NVIDIA coding interview problem focused on merge overlapping time intervals. The prompt emphasizes edge cases, clean implementation, and verifiable test behavior without revealing the solution.

  NVIDIAData EngineerCoding & AlgorithmsTechnical Screen
• Normalize an Absolute Unix-Style File Path to Canonical FormMedium

  Convert an absolute Unix-style file path into canonical form by collapsing repeated slashes, dropping current-directory markers, applying parent-directory moves that stop at the root, and treating other dot-only names as ordinary directories. Tests careful string parsing and edge-case handling of path components.

  NVIDIASoftware EngineerCoding & AlgorithmsOther
• Optimize CUDA GEMM with tiling and coalescingHard

  This question evaluates proficiency in GPU programming and performance optimization, assessing understanding of the CUDA execution model, memory hierarchy and bank conflicts, occupancy limits, and practical kernel design concerns for single-precision GEMM including tiling, coalescing, vectorized memory accesses and benchmarking.

  NVIDIAData ScientistCoding & AlgorithmsHR Screen
• Optimize a small-string C++ classMedium

  This question evaluates low-level C/C++ systems programming skills, including memory management, data layout, small-string optimization, efficient copying semantics, alignment/padding, and cache-aware performance reasoning.

  NVIDIASoftware EngineerSoftware Engineering FundamentalsOnsite
• Plan discovery and adoption strategyMedium

  This question evaluates a candidate's skills in product discovery, stakeholder alignment, technical communication, and developer adoption strategy for integrating a new ML technique, including defining use cases, constraints, baselines, and success criteria.

  NVIDIASoftware EngineerBehavioral & LeadershipTechnical Screen
• Reason About Endianness, Ring Buffers, and Stack Behavior in FirmwareMedium

  Tests firmware-level engineering concepts across low-level binary data parsing, lock-free ring buffer design for ISR-to-thread communication, and static/runtime stack depth analysis under nested interrupts.

  NVIDIASoftware EngineerSoftware Engineering FundamentalsTechnical Screen
• Reconstruct tree from inorder and postorderMedium

  This question evaluates understanding of binary tree traversal relationships, recursive construction techniques, and algorithmic time and space complexity, reflecting core data structures and recursion competencies.

  NVIDIASoftware EngineerCoding & AlgorithmsTechnical Screen
• Reflect on interview takeaways and adaptationMedium

  This question evaluates self-awareness, growth mindset, adaptive communication, and the ability to identify skill gaps and define measurable improvements within a Data Scientist interview context.

  NVIDIAData ScientistBehavioral & LeadershipHR Screen
• Research Project Deep Dive and Your AI-Agent Research WorkflowMedium

  Present a 20-minute deep dive into one research project covering motivation, research question, training pipeline, results and takeaways, then extend it to the internship problem. The round also asks how AI agents fit into your daily research workflow for writing, experiments and ideation, testing technical depth, honesty about results, and verification habits.

  NVIDIAApplied ScientistBehavioral & LeadershipOnsite
• Resolve conflict and learn from failureMedium

  This question evaluates a data scientist's cross-functional conflict resolution, stakeholder management, decision-framework literacy, quantitative impact measurement, ownership of failed projects, and ability to run postmortems and implement guardrails; it is categorized as Behavioral & Leadership within the Data Science domain and tests both conceptual understanding of decision-making frameworks and practical application of communication and metric-driven processes. It is commonly asked to assess interpersonal influence, accountability, learning from failure, and the capacity to convert lessons into measurable process, monitoring, or risk-control improvements that enhance decision quality.

  NVIDIAData ScientistBehavioral & LeadershipHR Screen
• Restaurant Table Reservations: Availability Search and Conflict-Free BookingMedium

  System design question asking for a restaurant table reservation service where diners search availability, book, change and cancel, and restaurants manage tables and hours. It tests the data model, preventing double-booking under concurrency, holds and retries, fast availability search and handling booking spikes.

  NVIDIASoftware EngineerSystem DesignOther
• Return all file paths via DFSEasy

  This question evaluates understanding of tree traversal and depth-first search along with competence in handling hierarchical data structures and constructing full file paths.

  NVIDIASoftware EngineerCoding & AlgorithmsTechnical Screen
• Reverse linked lists, including k-groupHard

  This question evaluates understanding of linked list manipulation including pointer operations, in-place reversal techniques, recursion and stack-depth considerations, cycle detection, and time/space complexity analysis.

  NVIDIAData ScientistCoding & AlgorithmsHR Screen
• Sell GPUs to a retail CEOMedium

  This question evaluates a candidate's ability to build a business and technical case for deploying GPUs across enterprise analytics and AI workloads, testing competencies in ROI/TCO modeling, workload discovery, pilot planning, stakeholder alignment, risk identification, and crafting data‑backed rebuttals.

  NVIDIAData ScientistBehavioral & LeadershipHR Screen
• Solve a Difficult Problem and Navigate DisagreementMedium

  Assesses technical problem-solving rigor, hypothesis testing, and the ability to resolve technical disagreements using data, empathy, and shared goals.

  NVIDIASoftware EngineerBehavioral & LeadershipTechnical Screen
• Solve anagram grouping and in-place allocatorMedium

  This problem tests string hashing/sorting techniques for grouping anagrams and the ability to manage fixed-size memory blocks in-place without auxiliary data structures.

  NVIDIASoftware EngineerCoding & AlgorithmsTechnical Screen
• Solve small string and API tasksMedium

  This problem tests your ability to handle nested structural validation using stacks and the design of a robust API integration function that aggregates paginated responses.

  NVIDIASoftware EngineerCoding & AlgorithmsOnsite