nVidia seems to have repeated ram fiasco with the 980ti DISPROVEN! Request closing thread

[Phixsator]

It turns out its just windows desktop that eats a lot of ram causing it to slow down. The actual ram is there at full speed. I made an error in the way that i tested it, as pointed out by Layte(thanks man) and as such this is NOT an issue.

I kindly ask a moderator to close this thread as there is no point to it any longer. Sorry for the confusion guys.

ProoF:Test has been rerun through iGPU

Old outdated and disproven

Spoiler

I took a crack at the memory benchmark used to discover the 970 ram issue..To my big surprise yet again, the last 500mb seems gimped..I dont know what to think of this.. This is with my spanking new 980 ti..


New run has been doneo with gpuz running as well.

Spoiler

EDIT: changed the title to reflect what card this is concerning.

 

[Phixsator]

Are you using another card to drive the display whilst running that test?

No my only display is connected to the 980ti.. ill redo the the tests with some a proper gpuz window next to it for validation reasons.

 

[Phixsator]

i have updated the OP with a new run..

Well that completely invalidates your test then.

please explain why that is? cause if there is any way this is not true i would very much like to confirm that

 

[Phixsator]

Layte claims the test is invalid due to the way my screen is setup.. so until the tests validity has been proven or disproven i think everyone should hold their horses.. no need to throw mud just yet

 

[Phixsator]

I'm going to be honest looking at the numbers and the fact it's 0.5gb again I can see this being true.

I have no issues with being disproven .. which is why i ask people to calm down. Now i just need to figure out how to redo it through another card..hmm

 

[Phixsator]

Can't you use the on board graphics of your CPU?

Yeah i just figured i would give this a try

At 3440x1440 you'll have a chunk of VRAM in use by aero/dwm which will be slow to read.

Indeed what the issue is.. running it again through iGPU resolves the issue. Ill better update the OP and ask for a close of the thread.. theres no point to it anymore.. Apparently i made a big error in my testing method.. Ohh well. Thanks to Layte for pointing it out!

Spoiler