What the CL number counts
CL is how many clock cycles the RAM waits between receiving a column address and starting to send data back. A kit sold as DDR5-6000 CL30 waits 30 cycles. The label gives CL first in a string like 30-38-38-78, and the other three numbers are the memory timings that follow it.
A count of cycles isn't a time. How long a cycle lasts depends on the clock, so the same CL means a different wait on different memory. CL is also exact, not a minimum: the memory controller and the module have to agree on one value.
Turning cycles into nanoseconds
DDR memory moves data on both edges of its clock, so the clock runs at half the data rate. DDR5-6000 has a 3000 MHz clock, and one cycle lasts 0.333 ns. At CL30 that's 30 cycles, or 10.0 ns.
The shortcut is CL × 2000 ÷ the data rate in MT/s. The 2000 comes from the clock being half the data rate: one cycle lasts 1000 ÷ (rate ÷ 2) nanoseconds, which is 2000 ÷ rate. Here are some common kits run through it:
- DDR3-1600 CL9: 11.3 ns
- DDR4-3200 CL16: 10.0 ns
- DDR4-3600 CL16: 8.9 ns
- DDR5-4800 CL40: 16.7 ns
- DDR5-6000 CL30: 10.0 ns
- DDR5-8000 CL38: 9.5 ns
Why a bigger CL isn't always slower
DDR5 kits usually carry bigger CL numbers than DDR4 kits, which looks like a step back. It isn't always. DDR4-3200 CL16 and DDR5-6000 CL30 both wait 10.0 ns, because each DDR5 cycle lasts 0.333 ns against 0.625 ns on DDR4.
The reverse catches people too. The same CL30 on DDR5-4800 would wait 12.5 ns, which is 25% more, because its cycles are longer. Compare kits in nanoseconds, never by CL alone. The nanosecond figure has its own name, true latency.
What CL leaves out
CL only covers the wait when the right row is already open. If the controller has to open a row first, tRCD and tRP add to it, and neither is in the CL number. Bandwidth is separate too. A DDR5-6000 kit peaks at 48.0 GB/s per channel and DDR4-3200 at 25.6 GB/s, whatever their CL.
How much a gap in nanoseconds matters depends on the workload. Programs access memory in different patterns, and more bandwidth can help a system even when latency goes up. The CAS latency chart ranks common DDR3, DDR4 and DDR5 kits by first word latency, so you can see where a kit lands before you buy.