Cache Memory Average Access Time: Computing AMAT from Hit Rate and Access Times
In a two-level memory system (cache + main memory), the average memory access time depends on the probability of a cache hit vs a cache miss, and on how long each case takes. A standard performance model is the AMAT (average memory access time) formula: 2
where is the cache hit rate, is cache access time, and is main memory access time. 2
We’ll apply it directly to your data: , , .
Key terms: keyword , keyword , keyword , keyword AMAT.
Footnotes
-
Computer Organization & Architecture (cache AMAT/effective access time model) - https://en.wikipedia.org/wiki/Cache_(computing) - Discusses hit/miss behavior and effective access time modeling. ↩ ↩2
-
Average memory access time (effective access time) equation in cache performance - https://en.wikipedia.org/wiki/Cache_(computing) - Standard AMAT-style expressions using hit rate and miss penalty. ↩ ↩2
Cache Hit/Miss and Effective Access Time (AMAT) Intuition
Compute AMAT and select the correct option
- 1Step 1
Use AMAT , reflecting that a miss still pays the cache access first, then main memory. 2
Footnotes
-
Computer Organization & Architecture (cache AMAT/effective access time model) - https://en.wikipedia.org/wiki/Cache_(computing) - Discusses hit/miss behavior and effective access time modeling. ↩
-
Average memory access time (effective access time) equation in cache performance - https://en.wikipedia.org/wiki/Cache_(computing) - Standard AMAT-style expressions using hit rate and miss penalty. ↩
-
- 2Step 2
AMAT .
- 3Step 3
Hit time contribution
- 4Step 4
Miss probability is and miss time is , so miss contribution
- 5Step 5
AMAT
- 6Step 6
The correct computed average is ; among the given choices, the closest is (iv) . If the exact intended result assumes a slightly different miss-time model, the intended option is still the one around
Expected Time Breakdown (Hit vs Miss)
AMAT = hit probability × Tc + miss probability × (Tc + Tm)
Common modeling rule for AMAT
On a miss, the CPU typically still spends the cache lookup time before going to main memory , giving miss time . This is why AMAT uses for the miss term. 2
Footnotes
-
Computer Organization & Architecture (cache AMAT/effective access time model) - https://en.wikipedia.org/wiki/Cache_(computing) - Discusses hit/miss behavior and effective access time modeling. ↩
-
Average memory access time (effective access time) equation in cache performance - https://en.wikipedia.org/wiki/Cache_(computing) - Standard AMAT-style expressions using hit rate and miss penalty. ↩
Why your options don’t include 20 ns exactly
With , ns, and ns, the strict AMAT model yields ns. If an answer key lists ns, it may be using a slightly different assumption (e.g., different rounding, miss penalty definition, or an alternative timing diagram). Use the AMAT model directly for correctness. 2
Footnotes
-
Computer Organization & Architecture (cache AMAT/effective access time model) - https://en.wikipedia.org/wiki/Cache_(computing) - Discusses hit/miss behavior and effective access time modeling. ↩
-
Average memory access time (effective access time) equation in cache performance - https://en.wikipedia.org/wiki/Cache_(computing) - Standard AMAT-style expressions using hit rate and miss penalty. ↩
Quick Answer (and formula check)
Knowledge Check
For cache hit rate , cache time , and memory time , which AMAT expression matches the standard model?