Virtual Labs

Directory-Based Cache Coherence Protocol Simulator

After using the simulator, what did you observe when multiple processors read the same memory block?

a: Only the first processor gets the data; others are blocked Explanation

Explanation

b: All processors receive data from memory and the directory state becomes Shared Explanation

Explanation

c: Each processor gets a different copy of the data Explanation

Explanation

d: The directory rejects additional read requests Explanation

Explanation

In the simulation, what visual indicator shows which processors currently share a memory block?

a: The cache state changes to red color Explanation

Explanation

b: The directory's sharer vector bits are set to '1' for sharing processors Explanation

Explanation

c: The network visualization shows permanent connections Explanation

Explanation

d: The performance counters increment Explanation

Explanation

Based on your simulation experience, what happens to the message count when using cache-to-cache transfers versus memory access?

a: Cache-to-cache transfers require more messages Explanation

Explanation

b: Both require the same number of messages Explanation

Explanation

c: Cache-to-cache transfers are more efficient and reduce overall network traffic Explanation

Explanation

d: Memory access is always faster Explanation

Explanation

During simulation, what did you observe about the directory state when a processor writes to a shared block?

a: The directory state remains Shared Explanation

Explanation

b: The directory state becomes Uncached Explanation

Explanation

c: The directory state changes to Exclusive with the writer as the new owner Explanation

Explanation

d: The directory state becomes permanently locked Explanation

Explanation

What did you notice about network message patterns when comparing read-heavy vs. write-heavy workloads?

a: Both workloads generate identical message patterns Explanation

Explanation

b: Read-heavy workloads generate fewer messages once sharing is established Explanation

Explanation

c: Write-heavy workloads require fewer directory lookups Explanation

Explanation

d: Network traffic is independent of workload characteristics Explanation

Explanation

From your simulation observations, when does the directory become a potential bottleneck?

a: When processors access different memory blocks Explanation

Explanation

b: When all processors repeatedly access the same popular memory block Explanation

Explanation

c: When the system is idle Explanation

Explanation

d: When processors only perform local computations Explanation

Explanation

In the simulator, what performance metric best indicates the efficiency of the directory protocol?

a: Total number of operations performed Explanation

Explanation

b: The ratio of cache hits to cache misses Explanation

Explanation

c: Number of processors in the system Explanation

Explanation

d: Amount of memory installed Explanation

Explanation

After extensive simulation, which scenario would you expect to show the greatest scalability advantage of directory-based over bus-based protocols?

a: Two processors sharing a single memory block Explanation

Explanation

b: Sixteen processors with mixed read/write patterns across multiple memory blocks Explanation

Explanation

c: One processor performing only local computations Explanation

Explanation

d: Four processors accessing completely disjoint memory regions Explanation

Explanation

Based on your simulation analysis, what modification would most effectively reduce directory bottlenecks for frequently accessed data?

a: Increase individual cache sizes Explanation

Explanation

b: Implement distributed directory organization with multiple directory nodes Explanation

Explanation

c: Reduce the number of processors in the system Explanation

Explanation

d: Use slower processors to reduce access frequency Explanation

Explanation