New Posts  All Forums:

Posts by ChronoBodi

It's still a sidegrade... just look at the massive 300+ clock rates it takes for the GTX 980 to beat the 780 Ti. Granted, the Maxwell cores ARE faster, but still doesn't make up for the more cores the 780 Ti has, even in slower Kepler form. Roughly, it's like this, (thank you OccamRazor!) GTX980@1450mhz = TitanBlack@1250mhz = GTX780Ti@1300mhz = Titan@1350mhz = GTX 780@1450mhz the 900 series are upgrades to everybody except those with GTX 780 and higher.
that is a 5820k though, six cores are easier to deal with heat wise than eight cores, but, voltage is the same idea no matter the core count?
because the old way of doing it (*80 being the absolute top end card) and introducing the big daddy first is expensive and yield issues are plenty when you do big dies on any node.It makes more money for them to pass off a midrange die (GK104, GM204, etc) as top-end for people that doesn't know any better besides us, as it is cheaper to make those smaller dies on the same node than the big daddy models.This strategy all started with GTX 680 that was able to compete with...
I do have to admit i do want to buy a new GPU or two even if i don't need it... actually, maybe i do, SLI Titans can barely do 4K 60FPS on Crysis 3, actually, more like 45 fps with no AA. It's just sooooo slow nowadays, the GPU refresh cycle. Makes me fear the upcoming stagnation if node shrinks are getting this hard to do.
Yea yea, will the chip degrade that much faster at those voltages? Mind you, this is $900 chip at Microcenter i paid for, so understandably i don't want to fry it lol. I think i can do 4.2-4.3 if i put it on 1.23v at least, but it seems soooo much higher than 1.185v.Actually, i remember running a i7 3770k at 1.35v for 4.5 ghz 24/7 and i7 3930k at 1.31v for 4.3 ghz. What was the stock volts on those chips?
I would not recommend it yet... Yes this post is a while back, but let me explain why and you should be aware so that you don't meet my fate, luckily nothing important was lost.So, first issue: if you change the PCI-E slot the m.2 is in from GEN 3 to GEN 2, the card does not show up in windows, but does show up in Disk Management, unformatted. And, if you update the BIOs, Windows 7 will think something is wrong with the m.2 card and do a CHKDISK on it at boot, it seems...
so i've been on 1.185v the whole time, because i'm somewhat chicken of higher volts, how many people around here says its safe to do 24/7 on Vcores of like 1.2v to 1.25v?
how there can be a GM200 on just 28nm? The amount of cores they can put in is limited by 28nm, and as i recall it was easily 80-100% performance boost when you compare the flagship Fermi to the flagship Kepler. I do not believe they will do a 28nm GM200, the die size required for that just kills it there, the yields would suck, the die size would be over 600mm, i'm just saying. UNLESS, Nvidia is WILLING to do GM200 on 28nm and just go ahead with the biggest die ever...
There ARE 20nm chips, but its ALL tied up in Apple SoCs like the A8 in the Iphone 6. lame. The question is when Nvidia gets access to 20nm, i'm curious.
waiiiiitttt, the Maxwells still have SLI bridges? I concede that AMD did do something cool by not needing them anymore for Crossfire, it's one less thing to worry about.
New Posts  All Forums: