True latency in nanoseconds

RAM Latency Calculator

Convert CAS latency and MT/s into nanoseconds, then put two kits side by side. Presets cover common DDR4 and DDR5 sticks. Change a RAM timing and both columns update. First-word latency is the CAS delay only.

Kit A
First kit
MT/s
CAS (CL)
tRCD
tRP
tRAS
Command rate
Channels
Kit B
Second kit to compare
MT/s
CAS (CL)
tRCD
tRP
tRAS
Command rate
Channels
⏱
Your result
First-word latency
Compare two kits. Lower nanoseconds win.
Faster kit
- ns

Kit A
First word
-
tRC (row)
-
Kit B
First word
-
tRC (row)
-
How they sit

By Mubashar Iqbal

How to use this RAM latency calculator

  1. Pick a preset for kit A, or type the MT/s and CL from the box. Add tRCD, tRP, and tRAS if the label has them.
  2. Do the same for kit B. Dual channel is the default. Switch to single channel only if you really run one stick.
  3. Read first-word latency, row cycle time, and peak bandwidth. The shorter nanosecond number wins on latency. The higher GB/s number wins on copy-heavy work.

First-word latency is CL x 2000 / MT/s. That is the CAS delay only. Real games also see tRCD, command rate, and whether the kit actually trains at the advertised EXPO or XMP profile.

Why CL alone lies

KitPaper CLFirst word
DDR4-3600 CL16168.89 ns
DDR4-3600 CL181810.00 ns
DDR5-6000 CL303010.00 ns
DDR5-6000 CL363612.00 ns

DDR5-6000 CL30 and DDR4-3600 CL18 can land on the same 10 ns first-word number. The faster bus on DDR5 still helps bandwidth. Intel Gear 2 and a weak IMC can erase a paper win.

RAM latency calculator photo of a DDR4 SO-DIMM showing memory chips used for CAS timing
A DDR4 module. This RAM latency calculator turns CAS and MT/s into nanoseconds. Photo: D-Kuru, CC BY-SA 4.0.

How the RAM latency calculator turns CL into nanoseconds

CAS latency (CL) is a cycle count, not a time. First-word latency in nanoseconds is CL times 2000, divided by the data rate in MT/s. DDR5-6000 CL30 is 30 x 2000 / 6000 = 10.00 ns. DDR4-3600 CL18 is 18 x 2000 / 3600 = 10.00 ns. That is why a bigger CL number on DDR5 is not automatically slower. JEDEC publishes the base DDR4 and DDR5 standards at jedec.org.

tRCD and tRP use the same conversion. tRC is tRAS plus tRP in cycles, then converted the same way. The RAM latency calculator shows first word and tRC for two kits so you can see when a ‘faster’ CL is just a faster clock.

XMP, EXPO, and what the board will actually train

The number printed on the heat spreader is an XMP or EXPO profile, not the JEDEC default the board uses with XMP off. If the motherboard QVL does not list the kit, it may boot at a slower JEDEC rate or refuse the profile. This page does not read your BIOS. It only does the arithmetic.

KitPaper CLFirst word
DDR4-3600 CL16168.89 ns
DDR4-3600 CL181810.00 ns
DDR5-6000 CL303010.00 ns
DDR5-6000 CL363612.00 ns

Need capacity before timings? Use the RAM size calculator. If the kit is fine and the game still stutters, check CPU versus GPU on the bottleneck calculator or read Guides.

What the RAM latency calculator will not tell you

Gear mode, Infinity Fabric, and whether the kit is on the motherboard QVL sit outside this page. A 10.00 ns first-word on paper can still fail EXPO and boot at JEDEC 4800. That boot is slower in nanoseconds than the heat-spreader math. AMD’s AM5 platform page is the place to confirm the CPU’s official memory ceiling, not a random Reddit screenshot.

tRFC and voltage are also out of scope. This RAM latency calculator is CL, tRCD, tRP, and tRC converted to time, plus a rough GB/s from the data rate and a dual-channel assumption. If the two kits tie on first word, pick the one on the board QVL, then confirm game FPS on the FPS calculator.

AM5 official notes: AMD AM5. Standards body: JEDEC.

If both kits share the same MT/s and CL, first-word latency ties and the RAM latency calculator will say so. Then the useful difference is capacity, QVL, or price, not another CL number.

RAM latency FAQ

DDR transfers twice per clock, so the real clock is half the MT/s number. The 2000 turns that clock and the CL count into nanoseconds.

For latency, yes. For copy-heavy work, bandwidth can matter more. A slightly slower first-word kit with more MT/s can still feel faster in some apps.

CL and MT/s are enough for a first-word compare. The other timings fill in row cycle time. If the box only prints CL, leave the extras on the preset.

The calculator does not train your IMC. Check the motherboard QVL and be ready to loosen timings if EXPO or XMP fails.

Use the RAM size calculator, then come back here to pick between two kits of that capacity.

Next step

Size the kit on the RAM size calculator, then confirm CPU and GPU balance on the bottleneck calculator.