F5F Stay Refreshed Hardware Desktop Reduced processing demand during HandBrake X265 conversion

Reduced processing demand during HandBrake X265 conversion

Reduced processing demand during HandBrake X265 conversion

Pages (2): 1 2 Next
T
tki_ip
Member
60
10-11-2016, 12:12 PM
#1
Hello everyone, your CPU runs at 11800H under Handbrake. When using medium encoding speed, your CPU usage drops to around 80, and it stays below 60 with slow settings. I've observed that when load is low, the CPU frequency fluctuates between its maximum and minimum values. Could this be due to a RAM bottleneck? You have 32GB of 2666MHz dual-channel memory (RAM type R1x8). Although my data is outdated, I'm curious—are there any tests that can confirm if RAM is the limiting factor?
T
tki_ip
10-11-2016, 12:12 PM #1

Hello everyone, your CPU runs at 11800H under Handbrake. When using medium encoding speed, your CPU usage drops to around 80, and it stays below 60 with slow settings. I've observed that when load is low, the CPU frequency fluctuates between its maximum and minimum values. Could this be due to a RAM bottleneck? You have 32GB of 2666MHz dual-channel memory (RAM type R1x8). Although my data is outdated, I'm curious—are there any tests that can confirm if RAM is the limiting factor?

E
explizip
Member
227
10-13-2016, 10:27 AM
#2
No, handbrake does not fully utilize your entire CPU.
E
explizip
10-13-2016, 10:27 AM #2

No, handbrake does not fully utilize your entire CPU.

M
MavrosGR
Senior Member
579
10-13-2016, 06:37 PM
#3
Adjusting the preset makes my CPU work harder and the clocks don’t spike much. With medium settings, even when using real-time, it only reaches about 60-70% of its capacity, but it fluctuates a lot—rising quickly to around 40-50% and then dropping again.
M
MavrosGR
10-13-2016, 06:37 PM #3

Adjusting the preset makes my CPU work harder and the clocks don’t spike much. With medium settings, even when using real-time, it only reaches about 60-70% of its capacity, but it fluctuates a lot—rising quickly to around 40-50% and then dropping again.

S
Sparkle_Mage
Member
206
10-13-2016, 10:43 PM
#4
Did you verify its effect on your data compression speed?
S
Sparkle_Mage
10-13-2016, 10:43 PM #4

Did you verify its effect on your data compression speed?

S
sacapatates
Posting Freak
843
10-14-2016, 06:48 AM
#5
Support on medium settings offers roughly double the speed compared to real time. On slower systems it runs about 25% slower, which means a 30-minute video might take around 45–50 minutes. I shared this info on Reddit and someone replied: same here with AMD3800x 32GB. My solution was turning on simultaneous encoding. Now the workload is well over 90%, and it seems the issue isn’t from my RAM.
S
sacapatates
10-14-2016, 06:48 AM #5

Support on medium settings offers roughly double the speed compared to real time. On slower systems it runs about 25% slower, which means a 30-minute video might take around 45–50 minutes. I shared this info on Reddit and someone replied: same here with AMD3800x 32GB. My solution was turning on simultaneous encoding. Now the workload is well over 90%, and it seems the issue isn’t from my RAM.

Y
yalo29
Senior Member
641
10-14-2016, 09:26 AM
#6
It seems your device handles workloads efficiently by utilizing all available cores or a full set of processors.
Y
yalo29
10-14-2016, 09:26 AM #6

It seems your device handles workloads efficiently by utilizing all available cores or a full set of processors.

F
Flazerblade
Junior Member
13
10-14-2016, 07:57 PM
#7
Typically all parts handle the workload uniformly. (I verified this) It's an 8-core setup with 16 threads.
F
Flazerblade
10-14-2016, 07:57 PM #7

Typically all parts handle the workload uniformly. (I verified this) It's an 8-core setup with 16 threads.

R
Rounyx
Posting Freak
838
10-15-2016, 03:43 PM
#8
It might be close to 100% with just 8 threads instead of 16. Unless the encoding is slower than expected, it shouldn’t be a major issue—especially if the delay is only a few minutes.
R
Rounyx
10-15-2016, 03:43 PM #8

It might be close to 100% with just 8 threads instead of 16. Unless the encoding is slower than expected, it shouldn’t be a major issue—especially if the delay is only a few minutes.

B
Bombartia
Senior Member
430
10-16-2016, 11:21 AM
#9
I faced a comparable problem after upgrading from 6700k to 3900x. I switched to another tool like Staxrip that handles chucks and parallel tasks. Typically I configure 4 chucks with an extra one for parallel work (4 for video encoding, 1 for audio). This led to all CPU threads reaching full capacity during encoding. The more chucks you set, the greater the memory demand, so monitor RAM usage particularly when working with higher resolution videos such as 4k.
B
Bombartia
10-16-2016, 11:21 AM #9

I faced a comparable problem after upgrading from 6700k to 3900x. I switched to another tool like Staxrip that handles chucks and parallel tasks. Typically I configure 4 chucks with an extra one for parallel work (4 for video encoding, 1 for audio). This led to all CPU threads reaching full capacity during encoding. The more chucks you set, the greater the memory demand, so monitor RAM usage particularly when working with higher resolution videos such as 4k.

X
xSudden
Member
228
10-16-2016, 04:01 PM
#10
Splitting the encoding does tend to reduce quality. Switching your RAMs to 3200mhz dual-rank could improve performance, but it might not raise your Cinebench score significantly—your current benchmark is already much lower than the top scores without any restrictions.
X
xSudden
10-16-2016, 04:01 PM #10

Splitting the encoding does tend to reduce quality. Switching your RAMs to 3200mhz dual-rank could improve performance, but it might not raise your Cinebench score significantly—your current benchmark is already much lower than the top scores without any restrictions.

Pages (2): 1 2 Next