RAM Latency Calculator

Convert CAS Latency (CL), tRCD, and tRP clock-cycle timings into true latency in nanoseconds using the standard DDR memory formula.

Quick Facts

Formula
True latency (ns) = CL x 2000 / Data rate (MT/s)
DDR is double data rate, so one clock cycle lasts 2000 / MT/s nanoseconds.
Why it matters
A higher CL number is not automatically slower
DDR4-3200 CL16 and DDR5-6000 CL30 both compute to about 10 ns of CAS latency.

Your Results

Calculated
True CAS latency (tCL)
-
CL converted to nanoseconds
RAS-to-CAS delay (tRCD)
-
Row activate to column access, in ns
Row precharge (tRP)
-
Time to close and reopen a row, in ns
Full random access latency
-
tCL + tRCD + tRP combined

Ready

Enter your memory's data rate and CL-tRCD-tRP timings, then press Calculate.

About the RAM Latency Calculator

RAM speed is usually advertised two ways at once: a data rate like "DDR4-3200" and a set of timings like "CL16-18-18-38." Neither number alone tells you how long a memory access actually takes. This calculator converts the clock-cycle timings — CAS Latency (CL), RAS-to-CAS Delay (tRCD), and Row Precharge (tRP) — into true latency in nanoseconds, using the same formula memory manufacturers and reviewers use to compare modules across generations.

The standard latency formula

DDR memory is double data rate, meaning it transfers data on both the rising and falling edge of the clock signal. So the real clock frequency is half the quoted transfer rate (MT/s), and one clock cycle lasts:

Clock cycle time (ns) = 2000 / Data rate (MT/s)

Each timing parameter (CL, tRCD, tRP) is specified in whole clock cycles, so multiplying by the cycle time converts it to nanoseconds:

True latency (ns) = Timing (cycles) x 2000 / Data rate (MT/s)

Applied to CAS Latency specifically, this is the formula widely published as: True CAS Latency (ns) = CL / (Data rate / 2) x 1000, which simplifies to the same CL x 2000 / MT/s used above. This calculator applies it to CL, tRCD, and tRP, then sums the three for an estimate of full random access latency — the time to activate an unopened row and read from it.

Worked example

For DDR4-3200 running CL16-18-18: the clock cycle is 2000 / 3200 = 0.625 ns. True CAS latency is 16 x 0.625 = 10 ns. tRCD is 18 x 0.625 = 11.25 ns, and tRP is another 11.25 ns. Full random access latency is roughly 10 + 11.25 + 11.25 = 32.5 ns. A DDR5-6000 CL30-36-36 module works out to 10 ns CAS latency, 12 ns tRCD, and 12 ns tRP — about 34 ns full latency — nearly identical real-world delay despite the CL number being almost double.

Interpreting estimates vs. measurements

  • This calculator reports the rated (JEDEC or XMP/EXPO profile) timings converted to nanoseconds — it does not measure your system. Actual access latency also depends on the memory controller, whether a row is already open, and platform-level overhead.
  • Sub-timings beyond CL, tRCD, and tRP (such as tRAS, tRC, or command rate) also affect real performance and are not included here.
  • Benchmark with a tool such as AIDA64 or Intel MLC for a measured latency figure before making purchasing or overclocking decisions.

Good-enough vs. precise

For comparing two RAM kits or sanity-checking an XMP/EXPO profile, the nanosecond figures from this formula are the standard reference point used across the industry. For validating a specific build's real-world performance, pair it with a measured benchmark.

Frequently Asked Questions

How do you calculate true RAM latency in nanoseconds?
True latency (ns) equals CAS Latency (CL) divided by half the data rate, then converted to nanoseconds: Latency = (CL x 2000) / Data Rate (MT/s). This works because DDR memory transfers data on both edges of the clock, so the real clock frequency is half the quoted MT/s data rate, and one clock cycle lasts 2000 / MT/s nanoseconds.
Why does a lower CAS Latency number not always mean faster RAM?
CL is measured in clock cycles, not time, so it only tells you something once you know the data rate. DDR4-3200 CL16 and DDR5-6000 CL30 both work out to about 10 nanoseconds of CAS latency, even though the CL numbers are very different, because DDR5 runs at roughly double the data rate. Compare true latency in nanoseconds, not the raw CL figure, when judging speed across generations.
What is full random access latency?
Full random access latency approximates the time to open (activate) a new memory row and read a column from it: it adds the CAS Latency (tCL), RAS-to-CAS Delay (tRCD), and Row Precharge time (tRP), each converted to nanoseconds. It is a more complete picture of real-world access delay than CAS latency alone, though actual latency also depends on the memory controller and whether the row is already open.

Related calculators