Discussion in 'Videocards - NVIDIA GeForce' started by alanm, Jan 23, 2015.
I felt this was pretty appropriate :infinity:
Think this thread generates a lot of new users to the forum, and site traffic in general. I don`t see any good that can come of a thread of this nature tbh.
Meh, I'll ask someone. This is just plain foolish and baiting.
Coming from the guy posting daft pictures and saying his last post was 3 posts ago
Once the claims of the shills was proven false, every forum should have closed to deleted every one of these threads. The longer they stay open, the more undo harm is done to NVidia and the forum community.
Sometimes you simply can't resist And the post thingie... Incredibly bored at work waiting for Windows updates to finally finish on two machines.
I blame the North Koreans.
Another official explanation is coming soon from Nvidia by PcPer...
Previous Official Statement was here from ManualG :
I just find it amusing that if they listed this as a footnote in a whitepaper someplace, absolutely no one would cared. Kinda have to give them props a bit for being open about the technical information regarding it. A lot of tech companies probably would have just ignored it.
Anyway, yeah this is like inflategate of the tech world.
This is such a hot topic at the moment that it's inevitable other threads on the subject will be opened as soon as this one is closed.
For the time being it's been decided to keep this open, but please keep any discussion civil and within the forum rules of course!
Please bear in mind Hilbert is hoping to get a statement from Nvidia, so hopefully that will shed more light on the matter and put this to rest one way or the other.
lol I don't know what is more amusing, Nvidia got caught doing this or everyone pretending like its okay
They're trying to corner the market on VRAM
fixed that for ya :banana:
How is it not OK? There is literally no discernible difference in performance that cannot be attributed to variance or the fact that there is a major hardware difference between the 980 and 970. It also appears for all intents and purposes to utilize the .5GB partition and bandwidth. So what exactly is the issue that makes this not OK? Like what is the issue here?
In the mean time my $800 Swift monitor that I've owned 3 of has a major manufacturing defect, one that's replicated across hundreds of Swifts (as shown on the ROG forums) and there isn't a single tech-wide investigation or peep out of ASUS about that.
So i can kill my self right now when i see this ;>
Nai's Benchmark, edited by VultureX
Device: GeForce GTX 970 (4.00 GB)
Memory Bus Width (bits): 256
Peak Theoretical DRAM Bandwidth (GB/s): 224.320000
Allocating Memory . . .
Chunk Size: 64 MiByte
Allocated 48 Chunks
Allocated 3072 MiByte
Msi GTX 970 Gaming G4 with new Bios and drivers ;f
Yup, you and all the Nvidia fanboys :question:
And when i was just about to get a SLI of 970 ..well guess ill have to wait until new refresh cards from the red team :blunt:
even if reading 40 pages is a bit much to ask , you could have at least read the post right above yours.....
Yeah, idk about this thread.
Anything involving that "benchmark" can be completely ignored. It reports the bandwidth for the CUDA virtual memory pool and not the VRAM itself. The "CUDA Virtual Memory Pool" includes VRAM, DRAM, caches and pagefile (if present). Due to this behavior, it's a useless "benchmark" and should have never been used to begin with.