vmem :
this is truly a beautiful advancement.
now all we need is true implementation. Because G-sync literally requires a hardware component in the monitor, we need a wide range of monitors to implement this, from your everyday cheap LCD panel to the color accurate IPS panels.
that aside, I know it's too much to hope for, but can Nvidia not lock down these nice things to their own GPUs? c'mon, open it up to AMD and (god forbid) intel's iGPUs
Let me know when AMD's Mantle is open. They've said it COULD run on other hardware, but lets be real here, open at this point means AMD allowing NV to make AMD gpus...LOL. How open is that? I don't see that happening either. If it was open wouldn't it work with all AMD gpus? You have to be running GCN or bust (making it just like needing NV 650+ to use G-sync), so NV would have to make GCN capable cpus, as in toss out kepler, maxwell etc. That isn't open. Also we have no idea what AMD would charge for this OPEN Mantle. Making Gsync open at this point would mean AMD making Kepler cards. We do not know there is any other way to do this, so claiming NV could even make it OPEN might be ridiculous as it may require things only Kepler currently does. We can say the same about Mantle needing GCN. Clearly it even requires special hardware just to work with an Kepler card. If they have to make a specific monitor card for each one it's even less easy to migrate to another vendor like AMD. We don't know about that yet either.
You first have to prove you can get something running on your OWN stuff before bothering with others and being open. I doubt they are even at a point where they can say someone else can run it. It can't be that easy or they would already have it running on 100 monitors instead of ONE launch monitor right? It wouldn't have needed a special NV card to work with the monitor if you could just roll this crap out willy-nilly right? It may never work with AMD without AMD making NV based GPU cards, and that will likely never happen. It would take AMD a good few years to put out a NV based chip and I'd venture to guess if they are that desperate they'd be bankrupt before fielding an NV based chip (they are near bankrupt now).
With BF4 being the first mantle game and it having so many problems already, I don't think NV will consider ever using mantle, and they can bleed for ages watching it first. I'm guessing Jen Hsun would rather bleed a billion or two before using anything AMD. With 2.7B in cash you can take the time to respond with your own R&D/Tech that matches Mantle rather than adopt it (nothing stopping that, NV certainly has funding to beat any AMD innovation). Until then lower prices if Mantle becomes important until you answer with your own tech. Lower pricing to Break-Even until you have an answer, and those lower prices will cut AMD margin to nothing or losses during the mean time. Lets see NV's options are adopt your enemy's tech legitimizing it, or break even for 1-3yrs to match their tech while keeping their margins at ZERO or losses while killing their tech and keeping them broke. I'll take door #2 every single time and so will NV and anyone else. You don't adopt your enemy's stuff if you have all the cash and no debt (the exact opposite of them). Instead of adopting their stuff, you just stall until you catch them, then go back to killing them and raising prices (see Intel...LOL - Note Haswell has went up $20 across the board pretty much).
The first 4 Tegras were NV bleeding until they could field their own ARM SOC cpu and pair it with their own GPU from the desktop and a modem; which up to now Qcom has had on everyone else basically owning the phone market due to this advantage (but that is over shortly). The first REAL Tegra is Tegra6. This will be the first one to make money also, though T5 might make some first (it's only missing the IN-HOUSE cpu part but the modem was most important I think for now as it gets more phone sales-we'll see how T4i does there)
The first 4 were just delay tactics until this point of bringing it all together. Shield was just prepping some great games in time for T5/T6 and getting gamepad/streaming working better. Next xmas I would be running from AMD/Intel Stock as they both have many people entering their world of CPU's (T6 and all it's enemies). TSMC just said 20nm is entering Volume production and 16nm is entering before end of 2014 ahead of schedule (16nm entering Risk production right now). The real party happens then as SOCS will be better than xbox360/ps3/wiiu and 1080p consoles will be a tough sell by then.
http://www.digitimes.com/news/a20131212PD211.html
Apparently they are 1Q ahead.
GT6 on PS3 shows old console's power envelopes still have enough power to make a great game for TV (even though car prices are absolutely stupid, the game itself looks pretty impressive). Until consoles hit 10mil+ sold nobody will concentrate on them. Next Gen will get crappy ports just like PC's until they sell 10mil units or more (except for Sony/MS titles, and a few EA ones I guess). It's smarter to aim at 100mil each of xbox360/ps3 than aim at 2-3mil of each next gen hoping one day they reach 100mil each. Like it or not, most will aim at last gen consoles for another year or more (and the rest will aim at mobile or PC - both 360/ps3 have ~14% aiming at them, while next gen both have less-see GDC 2013 surveys). My theory is, by the time next gen sales are enough to shoot at (next xmas when they have 10mil+ each?), there will already be a tablet/phone in your hand that does a job at least as good as xbox360/ps3 with next xmas android games. At that point a $400/500 console (with $60-70 games, and a VERY small catalog for both xbox1/ps4 even next xmas) is a tough sell vs what you already have in your hand with $0-20 games and a massive android library+PC gpu streaming to TV (NV will roll this out to any tegra device, it's a no brainer) and nothing hiding behind PAY WALL (like netflix etc all behind XboxLIVE $50 yearly subs! You shouldn't have to pay MS/Sony $50-60 to see netflix). The best experience will be a combo of your PC+mobile (mobile/android games out of the house, PC+streaming gpu+android inside house).
By next xmas consoles will be lucky to have 10-12mil each in the market. Very small compared to the numbers of 20nm products that will be sold by next xmas and shortly after as 20nm ramps fully everywhere. I don't see how consoles survive this + 1/2 dozen consoles coming (mojo, shield, gamestick, gamepop, wikipad, ouya etc etc). It took ONE day to sell 1 million of each new console, but another 20 days to sell the 2nd million of each. If the trend continues you'll be lucky to hit 3mil by xmas and need a few months after that to hit 4mil and I'll be shocked to see it not slow further all next year just like wiiu did (xmas/black friday don't come every month of the year). Wiiu had no new consoles coming after it (alone for a whole year basically), where xbox1/ps4 have a 1/2 dozen+ 20nm tablets, phones, PC GPU streaming, steamboxes etc along with no games catalog for another year or more. There is a tough road ahead for them. New models have better hardware obviously (both blow away wiiu/ps3/xbox360), but no games will be aiming at them for a while so extra power is no help. Black Friday showed last gen still sells very well even today.
Facebook
Twitter
Instagram