F5F Stay Refreshed Hardware Desktop Are you on the correct path? 5950X PBO Tuning and CO.

Are you on the correct path? 5950X PBO Tuning and CO.

Are you on the correct path? 5950X PBO Tuning and CO.

Pages (2): 1 2 Next
G
gordo_craftr2
Member
200
07-08-2020, 11:16 AM
#1
The TDC and EDC values appear modest (my 5900X typically performs slightly better with around 180 on each), but overall your numbers seem accurate. You might want to adjust the core offset and keep increasing it until you no longer notice performance improvements.
G
gordo_craftr2
07-08-2020, 11:16 AM #1

The TDC and EDC values appear modest (my 5900X typically performs slightly better with around 180 on each), but overall your numbers seem accurate. You might want to adjust the core offset and keep increasing it until you no longer notice performance improvements.

K
kawaii_hunter
Junior Member
5
07-08-2020, 11:43 AM
#2
It seems the TDC and EDC values might be lower than expected, yet core frequencies aren't improving after adjustments. I'll experiment further and share the results.
K
kawaii_hunter
07-08-2020, 11:43 AM #2

It seems the TDC and EDC values might be lower than expected, yet core frequencies aren't improving after adjustments. I'll experiment further and share the results.

B
BlueStar_LH
Posting Freak
842
07-11-2020, 01:44 AM
#3
It's reasonable to compare results across different tools and platforms. Y Cruncher tends to be more demanding, often requiring higher benchmarks, which could explain the differences in scaling. Also, using performance metrics instead of just frequencies matters, since those Ryzen processors can handle higher clock speeds but that doesn't always translate to better overall performance. Adjusting other settings can also impact frequency-based comparisons.
B
BlueStar_LH
07-11-2020, 01:44 AM #3

It's reasonable to compare results across different tools and platforms. Y Cruncher tends to be more demanding, often requiring higher benchmarks, which could explain the differences in scaling. Also, using performance metrics instead of just frequencies matters, since those Ryzen processors can handle higher clock speeds but that doesn't always translate to better overall performance. Adjusting other settings can also impact frequency-based comparisons.

D
Droxus
Member
51
07-11-2020, 12:50 PM
#4
On a 5900X I've checked TDC and EDC from 140 to 180 with minimal changes, achieving a 4.55 across all cores at 84 to 88°C, never exceeding it, and a 4.95 on the best clocks with -15 on two cores, -20 on the third and fourth, and -30 on others. Adjusting with LLC didn't significantly affect results.
D
Droxus
07-11-2020, 12:50 PM #4

On a 5900X I've checked TDC and EDC from 140 to 180 with minimal changes, achieving a 4.55 across all cores at 84 to 88°C, never exceeding it, and a 4.95 on the best clocks with -15 on two cores, -20 on the third and fourth, and -30 on others. Adjusting with LLC didn't significantly affect results.

Z
Ziongeckos
Junior Member
15
07-11-2020, 09:41 PM
#5
Turning up the settings near 180 for TDC or EDC causes a drop in frequencies. Lowering EDC from 180 brings the values down to 4.28 and 4.31 respectively.
Z
Ziongeckos
07-11-2020, 09:41 PM #5

Turning up the settings near 180 for TDC or EDC causes a drop in frequencies. Lowering EDC from 180 brings the values down to 4.28 and 4.31 respectively.

A
ayapasdepb
Member
128
07-11-2020, 11:20 PM
#6
On my 5900X I run 235 PPT, 160 TDC, 190 EDC, and then apply a curve.
A
ayapasdepb
07-11-2020, 11:20 PM #6

On my 5900X I run 235 PPT, 160 TDC, 190 EDC, and then apply a curve.

R
RubenLexus
Member
63
07-12-2020, 10:48 AM
#7
I tried changing them more and adjusting with a curve and nothing really happened lmao. The processor is cool as hell, 56-62 o C and not boosting over 4.8GHz now on single core. Also, for some reason in Ryzen Master, it will start to boost single core not on the fastest core in the CCX, but the 2nd fastest core instead. I don't know if it's a bug or what but at this point, I give up lol (just thinking of forgetting bothering to do this and letting the mobo take control and just send it).
R
RubenLexus
07-12-2020, 10:48 AM #7

I tried changing them more and adjusting with a curve and nothing really happened lmao. The processor is cool as hell, 56-62 o C and not boosting over 4.8GHz now on single core. Also, for some reason in Ryzen Master, it will start to boost single core not on the fastest core in the CCX, but the 2nd fastest core instead. I don't know if it's a bug or what but at this point, I give up lol (just thinking of forgetting bothering to do this and letting the mobo take control and just send it).

D
DukeItOut
Junior Member
5
07-13-2020, 04:08 PM
#8
To achieve the large values you need, revert to AGESA 1203c. I often observed its peak boost reaching 5150 across several cores.
D
DukeItOut
07-13-2020, 04:08 PM #8

To achieve the large values you need, revert to AGESA 1203c. I often observed its peak boost reaching 5150 across several cores.

S
Square_Tv
Junior Member
4
07-13-2020, 06:34 PM
#9
BIOS version 4201 is the latest supported. I'm uncertain if Ryzen Master alters existing BIOS settings since I previously manually configured PPT, TDC, and EDC in RM. It seems they might already be set. Adjusting them in RM didn't seem to make a difference. I changed the curve and limited max OC to 200 in BIOS, as I decided not to rely on RM anymore. It did make a slight improvement, though not significantly.
S
Square_Tv
07-13-2020, 06:34 PM #9

BIOS version 4201 is the latest supported. I'm uncertain if Ryzen Master alters existing BIOS settings since I previously manually configured PPT, TDC, and EDC in RM. It seems they might already be set. Adjusting them in RM didn't seem to make a difference. I changed the curve and limited max OC to 200 in BIOS, as I decided not to rely on RM anymore. It did make a slight improvement, though not significantly.

G
Germaindu51
Member
52
07-13-2020, 09:24 PM
#10
I work through everything in the bios without using any software.
G
Germaindu51
07-13-2020, 09:24 PM #10

I work through everything in the bios without using any software.

Pages (2): 1 2 Next