11
In order to find avg memory access time we have the formula :

Tavg = h*Tc +(1-h)*M

where  h     = hit rate
       (1-h) = miss rate
       Tc    = time to access information from cache
       M     = miss penalty  (time to access main memory)

I have been solving quite a few problems on this concept recently. At times I find that there is this disturbing inconsistency : 

Case 1: M = Tm + Tc
Case 2: M = Tm

Meaning, the solutions indicate that value of 'M' is calculated for some question X as in 'Case 1' above, while in some other question Y, the same is calculated as in 'Case 2' above. I tried my best analyzing these questions to find what is that factor which makes the calculation different. No help. I've encountered situations where X & Y are exactly the same, just that the values are only different, yet calculation is done for X as in Case 1 and for Y as in Case 2.

Any other reason that I am not aware of which makes the calculation different? Thank you in advance.

Sreekanth
  • 111
  • 1
  • 1
  • 3

3 Answers3

7

The difference comes from when the latency of a miss is counted. If the problem states that the time is a miss penalty, it should mean that the time is in addition to the time for a cache hit; so the total miss latency is the latency of a cache hit plus the penalty. (Clearly your formula and variables do not take this approach, labeling M--which is really total access time on a miss--as the miss penalty.)

Sadly, if a problem says "memory access latency" or "L2 access latency", it is even less clear whether total access latency is meant (i.e., including the time for an L1 hit) or the extra time required by an L1 miss. The former has some conceptual advantages (e.g., it can hide details like L2 access beginning before data return would occur on a hit--say by using early miss detection or miss prediction or even parallel tag look-up for L1 and L2). The latter may make explaining latency effects of L2 size or associativity simpler (e.g., if doubling size increases L2-only access latency by 50%, it may be easier to understand L2-only latency increasing from 8 cycle to 12 cycles by doubling size and to 18 cycles by quadrupling size than to see total latency increase from 10 cycles [with Tc = 2] to 14 cycles and to 20 cycles.)

(Also, using a miss penalty number allows a slight simplification of the access time formula-- Tavg = Tc + (1-h)Tm --because Tc is always spent.)

A similar issue occurs for execution latency. With a scalar pipeline an instruction that takes one cycle to execute is often said to have zero latency because there is no delay in executing a subsequent dependent instruction. However, this use of latency can be confusing when considering a superscalar implementation.

2

Average memory access time = % instruction * (hit time + Instruction miss rate * miss penalty) + % data * (hit time + Data miss rate * miss penalty)

B-Y
  • 187
  • 1
  • 2
  • 12
2

A related question was raised in comments:

Comment : Why there is no hit rate in the formula? – enitihas May 13 at 18:09

The reason there is no hit rate in the formula is every access whether it hits or misses the first level cache has to access the first level and thus has the hit time i.e. Access latency of the first level cache.

So irrespective of whether the access is hit/miss, hit time (Access latency of the first level cache) needs to be included in the formula to calculate the Average access time.

Peter Cordes
  • 328,167
  • 45
  • 605
  • 847
Surya
  • 21
  • 1
  • The real reason is that the miss penalty is relative to an L1 hit, which is why it's called a "penalty" rather than total time for an L1 miss/L2 hit. You're correct that most L1 cache designs fetch data in parallel with tags, so there's no extra time for the data to be ready after hit/miss is determined. But that's not necessary for the formula to work. The formula even still works in an insane system where requests are sent to L2 in parallel with checking L1 (so L1 hits doesn't reduce the volume / bandwidth of requests hitting the rest of the hierarchy at all). – Peter Cordes Sep 24 '16 at 07:41
  • hrm, after reading the question carefully, it appears the main subject of the question is whether the miss penalty is "total time for a request that misses", "time beyond what an L1 hit takes", or "time from when the rest of the cache starts checking" (which as Paul Clayton's excellent answer points out, isn't the same thing as L1 hit time, because miss detection could be slightly faster than retrieving the data from a hit). – Peter Cordes Sep 24 '16 at 07:58