Remember... you need to wait upwards of a week or so to see any movement in your RAC numbers.
Put it back to +80/+3000 on auto after I moved the cpu radiator to the front of the case.
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!
Is the rtx 3080 ti still the leader or has anything else turned out to combine good/great processing with low(er) cost?
Yeah, I am talking used too.
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!
And a pair of Rtx 3080 ti Founders Edition cards here.
Now the single RTX 3080 ti is running a memory transfer OC of +3000 with power limits from 350 to 366 watts (card maximum).
It is not throwing more than ~20 "marked invalids" a day.
Any attempt to push the Founders Edition cards to +3,000 for a memory transfer OC starts the card(s) throwing "computation errors" at 2-4 seconds of calculation.
Since the FE cards are living on an x570 MB with the slots running PCIE 4.0 It's got me wondering. I have run these cards as high as +0/+1500 with a card power limit of 400 watts. Its running PL 350 with a +0/+1100 while I ponder on ideas.
My next idea is to free up more CPU threads and see if that makes a difference. It seems to have made a processing speed difference (runs a little faster) on the single GPU system which runs Pcie 3.0
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!
Contrast and comparison of two models of RTX 3080 ti cards.
===edit===
My next idea is to free up more CPU threads and see if that makes a difference. It seems to have made a processing speed difference (runs a little faster) on the single GPU system which runs Pcie 3.0
When I got done juggling the non-monitor GPU is running +0/+3000 while the monitor GPU is running +0/+2000
Someplace above the 2000 memory OC on that card it starts throwing computation errors.
Now to see how my "marked invalids" go up/down/same.
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!
I would still like to see if I can crowd the FE cards to same production as my EVGA rtx 3080 ti. The would give my two gpu setup a 7M RAC.
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!
I would still like to see if I can crowd the FE cards to same production as my EVGA rtx 3080 ti. The would give my two gpu setup a 7M RAC.
I think for my pair in the slots they are in on the motherboard the answer is I cannot crowd them much.
I am not certain but someplace past a memory transfer OC of +1500 a card, possibly the same card/slot drops offline at a PL of 400.
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!
I think I may have solved a long running issue with my two GPU system.
I have not been able to run the basic ram OC without losing the 2nd GPU and throwing lots of computation errors. Which throws me into e@h "jail".
A fix that has allowed me to run my gpu on my Epyc system is being tried on my 2 GPU system.
I have set all the 16x (long) slots to fixed gen4.
On the Epyc system I couldn't get the gpu to be reliably recognized without setting the video card pcie slot to a fixed gen3.
I want to thank Ian&SteveC for this idea (several years ago).
This MIGHT allow success in more aggressive memory transfer OC on that system.
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!
I am looking at 3090 GPU's as a possible upgrade to.a smaller EVGA rtx 3080 ti.
The MB is unreliable if I try to run two GPU's so I am limited to upgrading the gpu.
I have been told the rtx 3090 has overheating problems so an aftermarket solution for the backside of the card is necessary.
Except for the Founders Edition.
For the purposes of this discussion assume the MB cannot be fixed or upgraded.
What is the best next step? A rtx 3090 fe, a rtx 4070 or sit on my hands, etc.
Thank you for your Advice.
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!
GWGeorge007
)
Put it back to +80/+3000 on auto after I moved the cpu radiator to the front of the case.
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!
Bang for the buck read. Is
)
Bang for the buck read.
Is the rtx 3080 ti still the leader or has anything else turned out to combine good/great processing with low(er) cost?
Yeah, I am talking used too.
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!
faster for less money? no not
)
faster for less money? no not really. there's faster for more money, and slower for less money.
_________________________________________________________________________
Contrast and comparison of
)
Contrast and comparison of two models of RTX 3080 ti cards.
I have a "smaller" evga RTX 3080 ti card here.
And a pair of Rtx 3080 ti Founders Edition cards here.
Now the single RTX 3080 ti is running a memory transfer OC of +3000 with power limits from 350 to 366 watts (card maximum).
It is not throwing more than ~20 "marked invalids" a day.
Any attempt to push the Founders Edition cards to +3,000 for a memory transfer OC starts the card(s) throwing "computation errors" at 2-4 seconds of calculation.
Since the FE cards are living on an x570 MB with the slots running PCIE 4.0 It's got me wondering. I have run these cards as high as +0/+1500 with a card power limit of 400 watts. Its running PL 350 with a +0/+1100 while I ponder on ideas.
My next idea is to free up more CPU threads and see if that makes a difference. It seems to have made a processing speed difference (runs a little faster) on the single GPU system which runs Pcie 3.0
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: Contrast and
)
When I got done juggling the non-monitor GPU is running +0/+3000 while the monitor GPU is running +0/+2000
Someplace above the 2000 memory OC on that card it starts throwing computation errors.
Now to see how my "marked invalids" go up/down/same.
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!
I would still like to see if
)
I would still like to see if I can crowd the FE cards to same production as my EVGA rtx 3080 ti. The would give my two gpu setup a 7M RAC.
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!
Thinking of all you people
)
Thinking of all you people that like to game and crunch and are wondering why your RAC is reducing...
it turns out your Nvidia driver could be to blame!!
https://www.techspot.com/news/97835-nvidia-acknowledges-cpu-usage-spikes-geforce-driver-53118.html
Tom M wrote: I would still
)
I think for my pair in the slots they are in on the motherboard the answer is I cannot crowd them much.
I am not certain but someplace past a memory transfer OC of +1500 a card, possibly the same card/slot drops offline at a PL of 400.
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!
I think I may have solved a
)
I think I may have solved a long running issue with my two GPU system.
I have not been able to run the basic ram OC without losing the 2nd GPU and throwing lots of computation errors. Which throws me into e@h "jail".
A fix that has allowed me to run my gpu on my Epyc system is being tried on my 2 GPU system.
I have set all the 16x (long) slots to fixed gen4.
On the Epyc system I couldn't get the gpu to be reliably recognized without setting the video card pcie slot to a fixed gen3.
I want to thank Ian&SteveC for this idea (several years ago).
This MIGHT allow success in more aggressive memory transfer OC on that system.
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!
Best choice for crunching for
)
Best choice for crunching for rtx 3090?
I am looking at 3090 GPU's as a possible upgrade to.a smaller EVGA rtx 3080 ti.
The MB is unreliable if I try to run two GPU's so I am limited to upgrading the gpu.
I have been told the rtx 3090 has overheating problems so an aftermarket solution for the backside of the card is necessary.
Except for the Founders Edition.
For the purposes of this discussion assume the MB cannot be fixed or upgraded.
What is the best next step? A rtx 3090 fe, a rtx 4070 or sit on my hands, etc.
Thank you for your Advice.
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!