Rumors of dual-GM200 Graphics Card Instigated

Published by

Click here to post a comment for Rumors of dual-GM200 Graphics Card Instigated on our message forum
https://forums.guru3d.com/data/avatars/m/206/206905.jpg
No problems i had with my 690 actually smoother than 2x 680s in SLI because of the "hardware based frame rendering". I would still be using my 690 if it wasnt for the 2gb vram. It also ran very smooth and is no different to this Titan X for smoothness, i see no difference. I also ran quad gpu's back in the Nvidia 295 days both dual pcb and single PCB. The single PCB were much smoother than the dual PCB and ran much cooler. Dual PCB had voltage unlock but was pointless running on an air setup wayback in 2009/10. Again no problems apart from the 1 or 2 games out of maybe 200 games, no issues for me. I would definitely go a dual card again, would not even need to consider it.
https://forums.guru3d.com/data/avatars/m/175/175902.jpg
I would definitely go a dual card again, would not even need to consider it.
i would definitively too... if i had empty PCIe slot lol but well... i had to chose: -buy expensive hardware for gaming, and lost money having fun OR -buy ultra expensive hardware for work and earn money with it (and now i event don't pay the ultra expensive stuff 🙂 ), and btw i can even play with it, not as a gaming machine but far than enough 😛c1:
https://forums.guru3d.com/data/avatars/m/124/124168.jpg
Got no problem buying a dual gpu again, though not a dual gm200 as I already have that in 2 980ti.
https://forums.guru3d.com/data/avatars/m/206/206905.jpg
690. Same price as 2 680's. It had more overclocking headroom because they were downclocked to hit 300w and they were the highest binned 680 chips. Pretty sure Veteran got his clocked near classified 680's at one point. Nvidia also claimed the PLX chip on board resulted in less jitter than doing 2 separate 680's. As for your second point, they can't do that. That's the entire point. There is an upper reticule limit on chip manufacturing, it's roughly ~600mm2. The 980Ti is already at that limit. The only way to get more cores is a smaller density on the process (16nm) or just make two of them and stick them together. Bus width and ram is essentially pointless. The Fury X has 1TB of bandwidth and it barely performs better if at all at 4K. It's also doing that on 4GB of ram. 6GB @ ~300GB/s is fine for now. As for SLI, I tend to agree. In some games it's perfectly fine, in others it's not fine. I loved my 690, especially after they fixed all the initial weirdness with multiple monitors, but I'll probably never go SLI again.
Yeap got mine running to 1372mhz (Agent-A01 helped also with I2C commands) fully stable which i posted videos of in Arma 3 fully maxxed out. I couldnt get it to budge past 1372mhz though. This was the highest clocked 690 that i seen anywhere on all the main enthusiast sites without exotic cooling. I had mine on custom water of course. Still a good card the 690 performance wise, at 1200p or under its still one of the best. [youtube]Ht-ppOK-58c[/youtube] Also here is hardware based frame metering review, check out the stutter in the graph on the AMD CFX.....LoLs http://www.computerbase.de/2012-05/test-nvidia-geforce-gtx-690/8/
https://forums.guru3d.com/data/avatars/m/105/105985.jpg
Dual cards so so f-ing stupid. Always slower and more expensive than 2 single gpus, and usually with less overclocking headroom. Also they sell such small volumes that I can't imagine it being very profitable either. They should make a single card with double the shader cores, double the ram, double the bus width and a decent stock OC and then we'll talk. SLI scaling and sli compatibility will always be a problem and one I have always been unwilling to overlook.
you forgot that is also stifles technology screw those frankin cards