Not exactly sure. It is the one Ian recommended though. So I presume the cpu is not over heating.
I have just set the power level to 170 while leaving the determinism slider on AUTO.
See if that is stable.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
That image has several entries that are NOT in my bios menu.
Let's see how raising the cTDP limit to 170 and keeping the determinism on "Auto" works.
I will then look at "back flashing" to 2.40/2.42 if the performance is still not there.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
like I said in my previous post. do not worry about flashing the BIOS. I think some options just aren't exposed because you have a Naples first gen CPU.
you also have a slightly different board than Keith and I. we have the EPYCD8, you have the EPYCD8-2T. which the only difference is the inclusion of 2x 10GbE LAN ports on your board whereas we have 1GbE ports.
the different boards carry different BIOS version numbers. your 2.60 BIOS is equivilent to our 2.40.
BIOS flashing will be more trouble than it's worth for you and I would not recommend it, as you're unlikely to get the extra features exposed if they aren't already, and you're already on the latest BIOS.
I'm not sure about that Ian. Look at the file size for his 2.30 BIOS version compared to his 2.60.
A full MB larger and I could easily assume that larger BIOS contains more features than the cut down version 2.60.
I'd also be curious whether his 2.60 BIOS has fan control.
Our 2.40 BIOS didn't and we didn't get it back until we persuaded Support to send us the L2.42 BIOS version which restored fan control we had in 2.10.
That BIOS was 1MB larger than our 2.40 also.
both of these packages are compressed. and the difference in file size is on the compressed package only and either due to different compression parameters or even due to the slightly different file name.
the uncompressed files (what actually gets written to the BIOS) are exactly the same size at 32MB
Of course the uncompressed file size is 32MB. It has to for it to be a viable BIOS image. You have to flash all cells in the chip with all 32MB of addresses.
However, that doesn't mean that addresses that contains a function name is empty or not. I have read enough in the forums to have a basic understanding of what the BIOS editors do with rewriting the blocks so that a function becomes active.
So I understand that a compressed image file can be smaller or larger depending on the dictionary lookup that compresses an empty function or one that has something in it.
So until somebody actually uses a BIOS hex editor to look at the BIOS image or goes ahead and flashes the image to have an actual look at the menus, we are just guessing if the field is there or not.
Let's see how raising the cTDP limit to 170 and keeping the determinism on "Auto" works.
>>The higher power limit has increased the MHz some on "auto".
The MHz is currently running between 2.5 and 2.6
I am going to try cTDP auto and determinism Performance next.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Keith Myers wrote: What is
)
Not exactly sure. It is the one Ian recommended though. So I presume the cpu is not over heating.
I have just set the power level to 170 while leaving the determinism slider on AUTO.
See if that is stable.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Ian&Steve C. wrote: it's in
)
That image has several entries that are NOT in my bios menu.
Let's see how raising the cTDP limit to 170 and keeping the determinism on "Auto" works.
I will then look at "back flashing" to 2.40/2.42 if the performance is still not there.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Tom M wrote: Ian&Steve C.
)
I would not worry about flashing your BIOS. it could be certain options being shown based on your first gen CPU being installed. maybe not available.
do you have the package power limit line?
might be beneficial for you to just take a picture of the screen so we can see what you see.
_________________________________________________________________________
Tom M wrote: Let's see how
)
The higher power limit has increased the MHz some on "auto".
There is no version of the bios listed as 2.40/2.42 on this website page: https://www.asrockrack.com/general/productdetail.asp?Model=EPYCD8-2T#Download
Maybe I am in the wrong page?
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Tom M wrote:Tom M
)
like I said in my previous post. do not worry about flashing the BIOS. I think some options just aren't exposed because you have a Naples first gen CPU.
you also have a slightly different board than Keith and I. we have the EPYCD8, you have the EPYCD8-2T. which the only difference is the inclusion of 2x 10GbE LAN ports on your board whereas we have 1GbE ports.
the different boards carry different BIOS version numbers. your 2.60 BIOS is equivilent to our 2.40.
BIOS flashing will be more trouble than it's worth for you and I would not recommend it, as you're unlikely to get the extra features exposed if they aren't already, and you're already on the latest BIOS.
_________________________________________________________________________
I'm not sure about that Ian.
)
I'm not sure about that Ian. Look at the file size for his 2.30 BIOS version compared to his 2.60.
A full MB larger and I could easily assume that larger BIOS contains more features than the cut down version 2.60.
I'd also be curious whether his 2.60 BIOS has fan control.
Our 2.40 BIOS didn't and we didn't get it back until we persuaded Support to send us the L2.42 BIOS version which restored fan control we had in 2.10.
That BIOS was 1MB larger than our 2.40 also.
Keith Myers wrote:I'm not
)
both of these packages are compressed. and the difference in file size is on the compressed package only and either due to different compression parameters or even due to the slightly different file name.
the uncompressed files (what actually gets written to the BIOS) are exactly the same size at 32MB
_________________________________________________________________________
Of course the uncompressed
)
Of course the uncompressed file size is 32MB. It has to for it to be a viable BIOS image. You have to flash all cells in the chip with all 32MB of addresses.
However, that doesn't mean that addresses that contains a function name is empty or not. I have read enough in the forums to have a basic understanding of what the BIOS editors do with rewriting the blocks so that a function becomes active.
So I understand that a compressed image file can be smaller or larger depending on the dictionary lookup that compresses an empty function or one that has something in it.
So until somebody actually uses a BIOS hex editor to look at the BIOS image or goes ahead and flashes the image to have an actual look at the menus, we are just guessing if the field is there or not.
Tom M wrote: Let's see how
)
The MHz is currently running between 2.5 and 2.6
I am going to try cTDP auto and determinism Performance next.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
from what I've read,
)
from what I've read, Determinism = Power/Performance, and the cTDP at it's max should be the best setting.
_________________________________________________________________________