Part of the Technology & IT suite · 35 calculators

Cache Hit Ratio Calculator

What a cache is actually buying — hit ratio, the effective average latency, and how much load it keeps off the origin.

Effective latency is the weighted average of hit and miss latency.

%
Results update as you type
Results
Effective latency (ms)
11.96
Requests reaching the origin (per second)
Served from cache (per second)
Latency with no cache (ms)
Speed-up against no cache
Effective latency at the target (ms)
Origin load at the target (per second)
Origin load removed by reaching the target
Reviewed September 2026. Computing arithmetic: bytes, bits, seconds and probabilities do not change by country. Ofcom requires UK providers to quote a minimum guaranteed download speed.
No account required · Google Analytics off unless allowedCalculator arithmetic runs in your browserResults update as you type
All calculations run 100% in your browser. The calculator code does not submit your figures to GlobalCalc to obtain a result.
About cache hit ratio

How the cache hit ratio calculator works

Effective latency is the weighted average of hit and miss latency. The non-obvious part is how sharply the returns diminish: going from 0 to 50% hit rate halves the origin load, but going from 90 to 95% halves it again — the same 5 percentage points near the top are worth ten times what they are worth in the middle.

That is why the last few per cent of hit rate are worth chasing on a hot cache and worthless on a cold one.

Formula: effective latency = h × hit + (1 − h) × miss

Worked examples

InputsEffective latency (ms)Note
88% hit rate11.9612 ms effective, 600 req/s to origin
A cold cache51.8origin load six times higher
Near-perfect2.83origin barely touched

Frequently asked questions

What is a good cache hit ratio?

For a CDN serving static assets, 90 to 98%. For an application cache it depends entirely on the access pattern — 70% can be excellent.

Why are the last few per cent worth so much?

Because origin load is the miss rate, not the hit rate. Going from 90 to 95% halves the misses; going from 40 to 45% removes only 8% of them.

Does a higher hit ratio always help?

Only if the misses were the problem. If the origin is comfortable and latency is already low, more cache buys nothing and costs memory.

What lowers a hit ratio?

Short TTLs, high cardinality in the cache key, and cold starts after a deploy. The cache key is usually the fixable one.

Is effective latency what users see?

It is the average. Users experience the distribution — the 12% who miss still wait the full 85 ms, and they are the ones who complain.

Where these figures come from

Last checked: September 2026. Units follow the SI decimal convention (IEC 80000-13 defines the binary alternatives); network and security figures cite the defining standard.