Search results

Doesn't work for everything, NWN for instance is stuck at 24hz, the turbo button was a way to fix it.

You are wrong and there is no sources available past present or future that you could provide to deem otherwise. The Tensor cores are not exposed at all to any of the graphical API's, being only exposed to CUDA with specific limitations to its utility in the sdk. Volta's BVH calculations are...

incorrect, both are 40nm Kepler was also a uarch+shrink from fermi Turing is build on 16nm+++ ala 14nm I think you're 12 months behind, 1.1 is a revisional update to the already available DXR spec. No, actually, they can't. The tensor cores are purely 16fp matrix multiplication Cards...

had a 3 platter wd20efrx fail at the 1TB boundary, corrupting several large data files have a 2 platter wd20efrx thats got weak sectors in the last 300GB, and the drives smart is not doing its job and reallocating them. also have a 2006 3 platter 1TB black and a 2011 2 platter 1TB black and...

The Rampage II boards that I previously said cannot boot Turing graphics cards, actually can! However you need an older graphics board to do it, as the issue is Turing relies on ACPI 2.0 features (yes, that setting in the bios that is disabled by default) Switch that on and you can boot from...

Garbage, you clearly don't know what you're on about, raising all associated clocks is better for performance overall.

nonsense i have 48GB's across 6 sticks at 1.25v on a x5690 (21x200 @4.2 / DDR3 1600) and 24GB's across 6 sticks at 1.26v(fluctuates to 1.28v) on a 5670 (19x200 @ 3.8 / DDR3 1600 either you have dud xeons or don't know what you're doing.

Retpoline restores most of the lost performance.

going by user reports and whether 3dmark results exist. a 3dmark result exists with a 2080 on the X58A-UD3R but the revision is not specified in the result.

threads already exist tracking x58 compatibility with turing. Partpicker doesn't have any background knowledge on what is and isn't actually functional, its just listed based on the pci-e capability. It looks like you are in luck and there are 3 working Gigabyte boards known so far, but...

Turing is only known to work in the following x58 motherboards so far The first 8 are reportedly bootable, the rest it is not sure if a uefi loader or second card to boot off is required. Rampage III Extreme Sabertooth x58 P6T P6t Deluxe P6t Deluxe v2 P6T7 WS SUPERCOMPUTER P6X58D-E P6X58D...

if you overclock properly you don't need to disable c states.

lol, these snobs whining about ugly coolers and then say they'll use an out dated coolermaster unit that will oxidise and discolour.

Silver Arrow http://thermalright.com/product/silver-arrow-ib-e-extreme/

http://valid.x86.fr/bauhrk

You appear to have set foot into the wrong thread.

and this is where ignorance reigns. You can significantly affect the life of solid caps with inconsistent regulation of voltage elsewhere, a problem common to gigabytes ever shitty vrm and phase regulation.

Gigabyte boards are junk, always. the voltages on them are nearly always out of spec.

just disable meltdown protection, our cpu's can't do it well, and its not even a real concern for end users.

look for a setting called "memory configuration protect" and turn it off.

my x5690 is 4.2 stable at 1.25v (1.267 set in bios)

SATA2 caps out at 260MB/s per port after transmission overheads.

I don't particularly care where he was banned from, the point is AMD has a flaw in its rendering pipeline and has not fixed it in the 5 years since it was first identified. They were quick enough to fix the AF flaw in the 5xxx series cards though after it was identified. As for driver vsync...

Scali responds to Richard Huddy accusations regarding The Witcher 3 in 2010. In other words, it has been known since 2010 that AMD's rendering pipeline chokes on its own design, and AMD has done nothing to rectify the situation except add a hack which for all intents and purposes should have...

Sorry to necro this, but im disgusted that these companies are allowed to get away with this b.s. This permittal of selling stuff with vague wording is disgraceful, a product marketted as having 'up to 3Gb/s' performance, when the chip inside is only SATA 1 compliant is knowingly misleading of...

I hope you compared HDAO on the Radeon vs the Geforce it will look different. - HDAO is inherently a Radeon optimised technique, while HBAO is Nvidia's own preferred method. Infact, what you claimed better actually appears inherently worse in HDAO vs HBAO. The grass appears to just stop...

I hope you compared HDAO on the Radeon vs the Geforce it will look different. - HDAO is inherently a Radeon optimised technique, while HBAO is Nvidia's own preferred method. Infact, what you claimed better actually appears inherently worse in HDAO vs HBAO. The grass appears to just stop...

Average boost speed.

The answer, on a technicality of DirectX, is no.

skype automatically sets itself to run at startup as well, so your post makes no sense

so what, i'll just hack the wlm2009 files so it appears like a version of skype :D

the disappearing sound is not a driver issue, the Windows Audio service is failing to start due to service time out. There is a hotfix available for this, linked in that very thread you mentioned on the creative forums.

@ TroyX those are DanielK's conclusions, not the posters, and i support danielk's findings having tested it against better onboard audio like AD's smartmax would never compare to realtek's junk.......

Finally a response

im soon to follow, just waiting for the 4GB cards to become more available.

im really starting to think these drives have an issue with intel I/O controller hubs.

Has been occuring since Release 250 its a Bios issue in certain EVGA 275 cards

revision marking is compiled into the kernel driver (nvklddm.sys) it can be read back using some of the gpu utilities, such as nvidia inspector, or gpu caps viewer or by hex viewing the nvklddm file (the file in the c:\nvidia\{drivers} folder requires expanding with expand.exe before you can...

301.25 is 2 commits newer than 301.24

these are older than 301.24/25