F5F Stay Refreshed Hardware Desktop Performance of DDR3 CAS latency

Performance of DDR3 CAS latency

Performance of DDR3 CAS latency

D
Dat_Asian_
Member
146
11-17-2016, 07:22 AM
#1
Will a 2 ms latency gap between 9 and 11 impact performance significantly? Also, the initial word needs to be changed from "10" to "10" for consistency.
D
Dat_Asian_
11-17-2016, 07:22 AM #1

Will a 2 ms latency gap between 9 and 11 impact performance significantly? Also, the initial word needs to be changed from "10" to "10" for consistency.

E
evogeli
Member
249
11-17-2016, 03:41 PM
#2
Not really, but the CAS number by itself offers little insight without considering frequency. CAS is shown in clock cycles, and a higher frequency means each cycle lasts shorter, which can result in lower absolute latency when measured in nanoseconds. For example, CL9 at 1600 MT/s equals 11.25 ns while CL11 at 2133 MT/s equals 10.3 ns.
E
evogeli
11-17-2016, 03:41 PM #2

Not really, but the CAS number by itself offers little insight without considering frequency. CAS is shown in clock cycles, and a higher frequency means each cycle lasts shorter, which can result in lower absolute latency when measured in nanoseconds. For example, CL9 at 1600 MT/s equals 11.25 ns while CL11 at 2133 MT/s equals 10.3 ns.

_
_NeoBl0X_
Senior Member
635
11-24-2016, 08:43 PM
#3
The delay mainly reflects the time between a processor's request for data and when it’s available for sending. The speed indicates how quickly information travels along the wires to the CPU once the process begins. As mentioned, Eigenvector explains this measurement in terms of "ticks" or clock cycles; higher frequency reduces the real-world duration a value takes. When an app handles large data blocks—like 512 KB or more—the time for transfer becomes the dominant factor, making latency less significant. Conversely, if the app constantly fetches tiny data pieces (such as 1 KB or less), the random nature of these requests prevents effective caching, increasing the impact of latency.
_
_NeoBl0X_
11-24-2016, 08:43 PM #3

The delay mainly reflects the time between a processor's request for data and when it’s available for sending. The speed indicates how quickly information travels along the wires to the CPU once the process begins. As mentioned, Eigenvector explains this measurement in terms of "ticks" or clock cycles; higher frequency reduces the real-world duration a value takes. When an app handles large data blocks—like 512 KB or more—the time for transfer becomes the dominant factor, making latency less significant. Conversely, if the app constantly fetches tiny data pieces (such as 1 KB or less), the random nature of these requests prevents effective caching, increasing the impact of latency.

S
Smilesoldier
Member
60
11-24-2016, 08:50 PM
#4
It will remain secured at 1600mhz regardless of the RAM I install, since my motherboard only accommodates 1600mhz.
S
Smilesoldier
11-24-2016, 08:50 PM #4

It will remain secured at 1600mhz regardless of the RAM I install, since my motherboard only accommodates 1600mhz.

A
Avogadroz
Junior Member
7
11-28-2016, 05:13 AM
#5
This comparison involves 11.25 ns (CL9) versus 13.75 ns (CL11). If the cost gap isn't significant, I'd choose the option with less delay, though it might not justify the extra expense. Also, DDR3 won't deliver the highest performance regardless of the choice.
A
Avogadroz
11-28-2016, 05:13 AM #5

This comparison involves 11.25 ns (CL9) versus 13.75 ns (CL11). If the cost gap isn't significant, I'd choose the option with less delay, though it might not justify the extra expense. Also, DDR3 won't deliver the highest performance regardless of the choice.

K
kirby4344
Junior Member
3
12-09-2016, 09:24 PM
#6
It seems you're looking for some vintage Hypers or similar parts. I found a setup that spins at 1600 RPM with a 6-6-6-20 configuration and runs at 1.65v. Comparing CL6 versus 11 is quite interesting.
K
kirby4344
12-09-2016, 09:24 PM #6

It seems you're looking for some vintage Hypers or similar parts. I found a setup that spins at 1600 RPM with a 6-6-6-20 configuration and runs at 1.65v. Comparing CL6 versus 11 is quite interesting.

C
COD_Jaime12
Junior Member
12
12-13-2016, 04:20 AM
#7
I found a few Hyper X kits online.
C
COD_Jaime12
12-13-2016, 04:20 AM #7

I found a few Hyper X kits online.

M
minifig_cd
Member
168
12-13-2016, 01:13 PM
#8
Elpida Hypers
M
minifig_cd
12-13-2016, 01:13 PM #8

Elpida Hypers