Question 6700 XT is not delivering expected performance ?

Feb 26, 2023
9
0
10
Hello peeps, recently i bought the Sapphire RX 6700 XT Nitro+ graphic card. The installation of it was fairly simple the drivers were installed properly and so on.
Nevertheless, i came upon the conclusion that the GPU is underperforming in games like Overwatch(which rlly was a buffling example). I will link at the end of this post a few videos where the same specs with mine are being used but for inexplicable reasons the FPS are higher.

My rig:
Mobo= MSI B550 Tomahawk
GPU= Sapphire RX 6700 XT Nitro+
CPU= Amd Ryzen 5 5600x
Ram= G. Skill DDR4 (Part number F4-3200C 16-8GVKB)x2
PSU= Corsair Gold 750M (which is 750W)

Truth be told after installing drivers i went to switch my C drive( where windows are installed) from MBR-->GPT as well as UEFI on bios instead of legacy. Problems did appear when these changes took place in form of a slower boot and a slighltly slower navigation in windows etc, some folders opened a bit slower. Not somethign significant though. THe aforementioned changes were made in order to activate sam(...) therefore BAR+4G were enabled on bios settings.

Dunno what possible solutions could be. Have tried uninstalling drivers via AMD driver uninstall program the amdcleanuputility.exe but to no avail. I am thinking of resetting windows BUT ONLY the option where i don't lose game/series folders 🙁 .

PS. I have a 2part cable leading to GPU x6 +x4 instead of one x12.


Video reference :
View: https://www.youtube.com/watch?v=bYO_KBg67qU
 
How much difference in fps is occurring. What are the expectations versus what is actually being measured/reported.

What tools or sources are you using to obtain the fps values?

Use Task Manager, Resource Monitor, and Process Explorer to learn more about what the system doing or trying to do.

Use all three tools but only one tool at a time.

Process Explorer (Microsoft, free):

https://learn.microsoft.com/en-us/sysinternals/downloads/process-explorer

Determine what resources are being used, to what extent ( % ), and what is using any given resource.

Observe performance at different times: gaming - non gaming, online - offline, application open - application closed.

Hardware specs may be the same: however, different configurations may be the root cause.

Consider drive space: could be identical drives with respect to make, model, and capacity. How full the drive is could be a factor.
 
How much difference in fps is occurring. What are the expectations versus what is actually being measured/reported.

What tools or sources are you using to obtain the fps values?

Use Task Manager, Resource Monitor, and Process Explorer to learn more about what the system doing or trying to do.

Use all three tools but only one tool at a time.

Process Explorer (Microsoft, free):

https://learn.microsoft.com/en-us/sysinternals/downloads/process-explorer

Determine what resources are being used, to what extent ( % ), and what is using any given resource.

Observe performance at different times: gaming - non gaming, online - offline, application open - application closed.

Hardware specs may be the same: however, different configurations may be the root cause.

Consider drive space: could be identical drives with respect to make, model, and capacity. How full the drive is could be a factor.

To answer NerdyVixen, no GPU is brand new with appropriate bios etc.
Ralston, while playing overwatch ultra +1440p resolution, these are the results: (btw drivers have at least 50GB free space)

GPU usage was fluctuating from 70~98% (where it was dark 70, heavy action 80+ which i would consider normal right?)
CPU usage ~ never climbed above 40. (is that normal?)
RAM being used was around 9GB
Temperatures were not significant though i have to admit GPU Junction temperature was 70C while gaming(is it okay?)
 
Regarding that Corsair Gold 750M (which is 750W) PSU: how old, history of heavy use for gaming, video editing, or even bit mining?

GPU (now) =

https://www.sapphiretech.com/en/consumer/radeon-rx-6700-10g-gddr6

What was the original GPU?

The RX-6700 specs recommend a 650 watt PSU. Even though the existing PSU is 750 watts, it may not be able to keep up with the overall power demands of the system.

All worked well (correct ?) with the original GPU but now the new GPU crossed some threshold wattage demand that the PSU can not longer reliably provide. Becomes more apparent during "heavy action".

I do not know if the GPU usage is normal per se. Depends very much on all of things the system is trying to do at any given time. If usage was 50-60% before then 70-98% is "not normal" without some reasons or expectations otherwise for a higher %.

Temperatures: Again I do not know what would be normal or expected per se.

However, if the temperatures are outside of the posted/documented specs and stated temperature ranges then 70C would not be okay.
 
Regarding that Corsair Gold 750M (which is 750W) PSU: how old, history of heavy use for gaming, video editing, or even bit mining?

GPU (now) =

https://www.sapphiretech.com/en/consumer/radeon-rx-6700-10g-gddr6

What was the original GPU?

The RX-6700 specs recommend a 650 watt PSU. Even though the existing PSU is 750 watts, it may not be able to keep up with the overall power demands of the system.

All worked well (correct ?) with the original GPU but now the new GPU crossed some threshold wattage demand that the PSU can not longer reliably provide. Becomes more apparent during "heavy action".

I do not know if the GPU usage is normal per se. Depends very much on all of things the system is trying to do at any given time. If usage was 50-60% before then 70-98% is "not normal" without some reasons or expectations otherwise for a higher %.

Temperatures: Again I do not know what would be normal or expected per se.

However, if the temperatures are outside of the posted/documented specs and stated temperature ranges then 70C would not be okay.


PSU is fairly new, 1.5 years purchase. the usage of it was strictly for gaming, no part of my whole rig was used for mining.

FOR THE LOVE OF GOD, just saw that i did not tag proper name of my GPU on my original post. it's this one https://www.sapphiretech.com/en/consumer/nitro-radeon-rx-6700-xt-12g-gddr6

no i would say that i did not come across any particular proplem with the wattage. while gaming ~80+% GPU usage is fine by me cause it's normal, and on idle it's also on fine levels. As it is brand name model i can't compare before and after, though my before GPU (which was nvidia) was also on ~80% usage.

Temperature seems again normal on idle-vlc etc. On heavy games although the core gpu temp stays around 60-69 the juncture can scale to 70. In order to preserve the longevity of my card was thinking of using Undervolt but i am not sure about the exact values.
 
The first step is to establish what the performance factors are and what quantifies those performance factors.

FPS, temperature, CPU usage, GPU usage, etc..

The second step is to then determine what values (or range of values) are applicable to each measured factor.

The third step is to measure the values.

Establish the current baseline for the build. As things are now. FPS (for example) now being some value or within a small range of values.

There will likely be some variations + or - results. All should be and remain within manufacturer specs.

If game play is not affected then leave well enough alone.

If game play is being affected then you can make changes (e.g., undervolting) in controlled increments.

What ever is done needs to be done methodically and only one change at a time.

Find the FPS that works for you and is safe (longevity) for the system. Who knows what all else others may be doing to get higher values.....
 
First off, I would get another drive, even if it is an external one, to backup your games. I purposely do not install games, on a drive/partition run by the os, to avoid what you are wanting to avoid. Is XMP turned on, for your ram?

Yes XMP is enabled, tested it both on bios and on task manager, by testing i mean checked it's enabling. To be honest non-OC my rig is getting scores of 11.583 on 3DMark which is, from what i've seen, normal given the fact that i am not either OC, UV.
Propably i panicked needlessly.
 
The first step is to establish what the performance factors are and what quantifies those performance factors.

FPS, temperature, CPU usage, GPU usage, etc..

The second step is to then determine what values (or range of values) are applicable to each measured factor.

The third step is to measure the values.

Establish the current baseline for the build. As things are now. FPS (for example) now being some value or within a small range of values.

There will likely be some variations + or - results. All should be and remain within manufacturer specs.

If game play is not affected then leave well enough alone.

If game play is being affected then you can make changes (e.g., undervolting) in controlled increments.

What ever is done needs to be done methodically and only one change at a time.

Find the FPS that works for you and is safe (longevity) for the system. Who knows what all else others may be doing to get higher values.....

I wanted to undervolt in order to ONLY decrease the juncture temperature which i believe the more we go into warmer months the higher it will become. Nevertheless, if leaving settings on default (and of course getting the wanted fps) do not harm in any way my GPU, because i am totally muggle on software tweaking, i am propably leaning towards NOT UV. You see in heavy AAA single player games i believe UV will be detrimental to the whole experience because i will sacrifice FPS-performance in order to run my GPU on lower Voltages. If by any chance this isn't the case, I would love to be corrected by a more expert person.
 
What is this? The card takes a 6 and 8 pin connector.
It also has a BIOS switch for high performance.
For the performance comparison to a YouTube video one game patch can change the performance in a game.

I meant that i use two seperate cables a 6 and 8 rather than one (cause i read that 2 are better performance wise). The bios switch unfortunately i haven't located. Is it on the graphics card or can it be found through amd or trixx software?
 
I would leave things alone and wait until the warmer months to actually see what happens.

And, in the meantime, keep some logs regarding system temperatures, %'s, FPS, etc. while gaming and not gaming.

Plus you can use the time to do more reading and get a better sense of what all is involved.

For example:

https://www.howtogeek.com/805771/what-is-undervolting-gpu-or-cpu/

"Should You Undervolt Your Computer?
There are two parts to answering the question of whether you should be undervolting your PC. The first is the risks involved. Broadly speaking, the risk that you’ll damage your computer through undervolting is effectively zero. It can be distressingly easy to kill a CPU or GPU by putting too many volts through it, but the opposite should be harmless.

You do, however, run the risk of having an unstable system if the voltage is too low, and that can lead to data corruption. So, as always, it’s a good idea to back up anything that matters before attempting to modify your computer.

As long as you know how to reset your motherboard settings to default if it doesn’t boot (check the manual) or you know how to boot into Safe Mode, there’s no real reason to avoid undervolting if you have the time and patience for it."

= = = =

You can easily find other similar links and there may be additional comments and suggestions here.

Becomes a matter of trade-offs and how much risk you may be willing to take.

I am in the "If it ain't broke, don't fix it" club.
 
I would leave things alone and wait until the warmer months to actually see what happens.

And, in the meantime, keep some logs regarding system temperatures, %'s, FPS, etc. while gaming and not gaming.

Plus you can use the time to do more reading and get a better sense of what all is involved.

For example:

https://www.howtogeek.com/805771/what-is-undervolting-gpu-or-cpu/

"Should You Undervolt Your Computer?
There are two parts to answering the question of whether you should be undervolting your PC. The first is the risks involved. Broadly speaking, the risk that you’ll damage your computer through undervolting is effectively zero. It can be distressingly easy to kill a CPU or GPU by putting too many volts through it, but the opposite should be harmless.

You do, however, run the risk of having an unstable system if the voltage is too low, and that can lead to data corruption. So, as always, it’s a good idea to back up anything that matters before attempting to modify your computer.

As long as you know how to reset your motherboard settings to default if it doesn’t boot (check the manual) or you know how to boot into Safe Mode, there’s no real reason to avoid undervolting if you have the time and patience for it."

= = = =

You can easily find other similar links and there may be additional comments and suggestions here.

Becomes a matter of trade-offs and how much risk you may be willing to take.

I am in the "If it ain't broke, don't fix it" club.
Yeah i am of the same club too cause i am rlly afraid of the repercusions. Will re-visit on summer though ONLY if temperatures go crazy, promise.

Thanks for the interest and your replies.
 
Then you posted the wrong link for the video card.
No, the manufacturer I would guess mistakenly while posts that bios switch is present, it really is not. Many reddit posts refer to the same problem

View: https://www.reddit.com/r/Amd/comments/mgxzlo/sapphire_rx6700_xt_cards_pulse_maybe_nitro/

View: https://www.reddit.com/r/AMDHelp/comments/wcjw16/sapphire_nitro_rx_6700_xt_missing_bios_switch/


To be exact i see this https://ibb.co/g41qKdk but no pin in order to move it through the positions.