F5F Stay Refreshed Hardware Desktop High-resolution undervolting outcomes (amazing!)

High-resolution undervolting outcomes (amazing!)

High-resolution undervolting outcomes (amazing!)

Pages (2): 1 2 Next
T
TheSonicRide
Junior Member
11
02-01-2022, 09:57 PM
#1
I also tried undervolting my CPU again and achieved some surprising outcomes. (My motherboard is an ASUS ROG Crosshair X670E Gene.) My curve optimizer showed -60 for core 0-6 and stable -30 for core 7. When I ran a CB R23 multi-core test, my score was just above stock levels, and the CPU only consumed slightly over 0.9V spread across all cores. (Typically it's around 1.1V, but this one can reach up to 1.15V.) It seems unusual that I can't apply tighter timings on my memory, which usually works well with other overclocking settings. Maybe the voltage isn't enough for it to function normally?
T
TheSonicRide
02-01-2022, 09:57 PM #1

I also tried undervolting my CPU again and achieved some surprising outcomes. (My motherboard is an ASUS ROG Crosshair X670E Gene.) My curve optimizer showed -60 for core 0-6 and stable -30 for core 7. When I ran a CB R23 multi-core test, my score was just above stock levels, and the CPU only consumed slightly over 0.9V spread across all cores. (Typically it's around 1.1V, but this one can reach up to 1.15V.) It seems unusual that I can't apply tighter timings on my memory, which usually works well with other overclocking settings. Maybe the voltage isn't enough for it to function normally?

P
Peento
Junior Member
3
02-01-2022, 11:11 PM
#2
I don’t notice anything unusual in this place. Overclocking the CPU didn’t change my RAM speeds.
P
Peento
02-01-2022, 11:11 PM #2

I don’t notice anything unusual in this place. Overclocking the CPU didn’t change my RAM speeds.

L
LunyGaming
Junior Member
10
02-08-2022, 06:08 AM
#3
It seems around -60 across nearly every core works well, while typical AM5 boards usually cap at -30.
L
LunyGaming
02-08-2022, 06:08 AM #3

It seems around -60 across nearly every core works well, while typical AM5 boards usually cap at -30.

N
Nordin2006
Junior Member
6
02-24-2022, 04:38 PM
#4
It seems -60 isn't really effective and the boundary is -40. Lowering settings doesn't guarantee they work.
N
Nordin2006
02-24-2022, 04:38 PM #4

It seems -60 isn't really effective and the boundary is -40. Lowering settings doesn't guarantee they work.

S
ShadyPineTrees
Junior Member
15
03-01-2022, 02:04 PM
#5
This won't let me upload at -20 kidding me!
S
ShadyPineTrees
03-01-2022, 02:04 PM #5

This won't let me upload at -20 kidding me!

S
SedentarySauS
Senior Member
411
03-01-2022, 10:45 PM
#6
Sounds like a fun bet on the future!
S
SedentarySauS
03-01-2022, 10:45 PM #6

Sounds like a fun bet on the future!

T
Thypolicon
Member
66
03-03-2022, 08:32 PM
#7
Yeah, I mean I can't complain as the difference in gaming compared to the 5950X I had before is really dramatic. Its those 1% low improvements that particularly changes things. I had enough problems with my 5950X locking up when idle on stock voltage, despite being stable with PBO on -20mV under load. For some reason that needs +5mV to be stable at lower clocks, I think because its an early stepping.
T
Thypolicon
03-03-2022, 08:32 PM #7

Yeah, I mean I can't complain as the difference in gaming compared to the 5950X I had before is really dramatic. Its those 1% low improvements that particularly changes things. I had enough problems with my 5950X locking up when idle on stock voltage, despite being stable with PBO on -20mV under load. For some reason that needs +5mV to be stable at lower clocks, I think because its an early stepping.

K
killer37890
Member
67
03-09-2022, 07:39 AM
#8
Absolutely, I see the distinction you're pointing out. Regarding the 5950x model, it's indeed defective if it fails to operate at the standard operating voltage.
K
killer37890
03-09-2022, 07:39 AM #8

Absolutely, I see the distinction you're pointing out. Regarding the 5950x model, it's indeed defective if it fails to operate at the standard operating voltage.

G
Gollum4545
Member
107
03-10-2022, 10:44 PM
#9
Possibly, but given I've had it since April 2021 and didn't notice the problem back then. It could be the ASUS board it was on before was compensating (ASUS being known for having their voltages on the high side), its deteriorated, or the Gigabyte board its on now is just ever so slightly too low at idle. Googling it seems not uncommon for early 5950X to have this issue, some were RMA and the replacement still did it. This is an OEM chip I bought in a bundle, which also makes it more complicated as I'd have to see what the company I bought it from had to say. They probably have a somewhat valid argument that it worked fine in the motherboard it was supplied with.
G
Gollum4545
03-10-2022, 10:44 PM #9

Possibly, but given I've had it since April 2021 and didn't notice the problem back then. It could be the ASUS board it was on before was compensating (ASUS being known for having their voltages on the high side), its deteriorated, or the Gigabyte board its on now is just ever so slightly too low at idle. Googling it seems not uncommon for early 5950X to have this issue, some were RMA and the replacement still did it. This is an OEM chip I bought in a bundle, which also makes it more complicated as I'd have to see what the company I bought it from had to say. They probably have a somewhat valid argument that it worked fine in the motherboard it was supplied with.

S
Spawn377
Member
215
03-11-2022, 04:28 AM
#10
Really sad that the powerful CPU got harmed because of issues with the motherboard BIOS... I’m hoping your repair goes smoothly! I also had to replace my first 7800X3D because I accidentally destroyed it after three days...
S
Spawn377
03-11-2022, 04:28 AM #10

Really sad that the powerful CPU got harmed because of issues with the motherboard BIOS... I’m hoping your repair goes smoothly! I also had to replace my first 7800X3D because I accidentally destroyed it after three days...

Pages (2): 1 2 Next