There are N computer centers, each with multiple facility rooms offering various types of facilities. Users can request bookings for these rooms across 24 time slots in a day. Each booking specifies the computer center, facility room, starting slot, and number of consecutive slots required. Requests are processed concurrently using GPU cores to maximize throughput and performance.
-
Users interact with an application to book facility rooms:
- Specify computer center, facility room, starting time slot, and number of slots.
- Slots are booked consecutively from the starting slot.
-
Concurrency Handling:
- Ensure consistency and avoid race conditions when multiple requests target the same facility.
- Prioritize requests with smaller IDs if multiple requests target the same facility and the room is at capacity.
-
GPU Utilization:
- Implement parallel processing using GPU kernels for efficient handling of multiple requests.
- Avoid print statements inside GPU kernels to maintain performance.
After processing all requests:
- Print the total number of successful and failed requests.
- Print the total successful and failed requests for each computer center.
- Race Condition: Ensure mutual exclusion to prevent race conditions when multiple requests access the same facility concurrently.
- Priority Handling: Requests with smaller IDs should be prioritized if the facility is overbooked.
- GPU Implementation: Use GPU kernels effectively to achieve parallelism and improve performance.
- Sequential Implementation: Minimize sequential execution to maximize marks; focus on leveraging GPU parallelism.
- Submit a
.cufile namedRollNumber.cuon Moodle.
- Develop a CPU version of the code to compare performance with the GPU implementation.
- Measure and analyze performance differences between CPU and GPU implementations using large inputs.
- Explore synchronization mechanisms to manage concurrent access to shared resources effectively.