F5F Stay Refreshed Hardware Desktop Spectrix D80 3600 CL17 model with memory OC capability

Spectrix D80 3600 CL17 model with memory OC capability

Spectrix D80 3600 CL17 model with memory OC capability

Pages (2): 1 2 Next
N
Narnab
Member
105
08-23-2019, 06:40 AM
#1
Hey everyone, I got my Spectrix D80 3600 CL17 running at 14-14-14-28 3600 1.45v. The AIDA scores look off—shouldn’t the latency be lower? Also, Power Down is turned off.
N
Narnab
08-23-2019, 06:40 AM #1

Hey everyone, I got my Spectrix D80 3600 CL17 running at 14-14-14-28 3600 1.45v. The AIDA scores look off—shouldn’t the latency be lower? Also, Power Down is turned off.

B
black_blaziken
Junior Member
14
08-23-2019, 09:41 PM
#2
Performance differences show latency tends to be greater for Ryzen compared to Intel.
B
black_blaziken
08-23-2019, 09:41 PM #2

Performance differences show latency tends to be greater for Ryzen compared to Intel.

F
FLPFive
Member
170
08-25-2019, 05:04 PM
#3
It seems reasonable to reduce it to 1.4 for me.
F
FLPFive
08-25-2019, 05:04 PM #3

It seems reasonable to reduce it to 1.4 for me.

S
SwerPent
Junior Member
36
08-25-2019, 07:50 PM
#4
Theoretical maximum bandwidth for a 3600MHz memory is 57,600 MB/s, which might make your findings appear modest—perhaps the secondary timing parameters are quite tight. Latency seems to be more typical.
S
SwerPent
08-25-2019, 07:50 PM #4

Theoretical maximum bandwidth for a 3600MHz memory is 57,600 MB/s, which might make your findings appear modest—perhaps the secondary timing parameters are quite tight. Latency seems to be more typical.

N
Niclasini
Junior Member
2
09-03-2019, 08:58 PM
#5
Looks like the signal is weak, but your 3600MHz CL18 model shares comparable delays too. Make sure no background processes are active during testing.
N
Niclasini
09-03-2019, 08:58 PM #5

Looks like the signal is weak, but your 3600MHz CL18 model shares comparable delays too. Make sure no background processes are active during testing.

P
PandaPantz
Member
59
09-04-2019, 12:03 AM
#6
You're right to verify this. It's common knowledge that 1.45V is safe for B-DIE, and even 1.5V works. Your adjustments to the timings and motherboard settings seem to have achieved solid performance—53655MB/s is impressive! Let me know if you need more details.
P
PandaPantz
09-04-2019, 12:03 AM #6

You're right to verify this. It's common knowledge that 1.45V is safe for B-DIE, and even 1.5V works. Your adjustments to the timings and motherboard settings seem to have achieved solid performance—53655MB/s is impressive! Let me know if you need more details.

A
alex68
Junior Member
36
09-04-2019, 04:58 AM
#7
It would reach its maximum voltage when timing isn't maintained at 1.4, but keep it at 1.45
A
alex68
09-04-2019, 04:58 AM #7

It would reach its maximum voltage when timing isn't maintained at 1.4, but keep it at 1.45

C
Challlenger
Member
67
09-10-2019, 02:52 PM
#8
At 1.4v the MOBO isn't posting. Latency is now 66.6ns with tRC 40 and tRFC 288. Which option—3600 14-14-14-28 or 3733 16-16-16-36—performs better overall?
C
Challlenger
09-10-2019, 02:52 PM #8

At 1.4v the MOBO isn't posting. Latency is now 66.6ns with tRC 40 and tRFC 288. Which option—3600 14-14-14-28 or 3733 16-16-16-36—performs better overall?

R
Rey_Gordo6
Member
51
09-10-2019, 10:31 PM
#9
It depends on the maximum speed of the FCLK
R
Rey_Gordo6
09-10-2019, 10:31 PM #9

It depends on the maximum speed of the FCLK

H
Hanzn
Junior Member
15
09-11-2019, 12:57 AM
#10
Typically 3600 works better, though a higher frequency needs a higher FCLK (provided the interconnect supports it)
H
Hanzn
09-11-2019, 12:57 AM #10

Typically 3600 works better, though a higher frequency needs a higher FCLK (provided the interconnect supports it)

Pages (2): 1 2 Next