I might as well weigh in on this since I tweeted this as a discussion topic:
Mantle:
- Unproven
- Requires no new hardware if you already have a GCN (7000+) card
- Requires developer support (but we know that at least some is coming)
- Unknown performance improvement over non-Mantle
- Takes time to implement, limited compatibility with existing hardware and fragments market, which will be a major barrier
- No support for existing titles unless ported in after the fact
- Potentially makes Linux gaming more feasible
G-Sync:
- Requires Kepler GPU (600 series or better)
- Requires new hardware on monitor side of things (but not that expensive to implement compared to existing scaler hardware)
- Does not require developer support & supports pretty much your entire existing library of games.
- Proven performance. Holy crap does it ever make a big difference. I'd rather play 50FPS with G-Sync than 80FPS with tearing from what I saw. Seriously..
- Requires hardware manufacturer support (at least some is coming, but not from A+ players like Samsung. I saw Philips, ASUS, BenQ, and Viewsonic on the list...)-
I'll discuss this more on the WAN Show this week, but let me put it this way. To all of the folks saying any of the following about G-Sync:
- a useless features
- can't convince me that lower framerates can look better
- snake oil
You just don't understand the technology here fully. This is ABSOLUTELY game-changing. It makes a HUGE difference. Enormous. Earth-shattering. Not open for debate. End of discussion. Unless you've tried it, (I have, you haven't at this point unless you're another journalist that was at the event) then it's best to say nothing before you make yourself look like a fool.
The challenges for G-Sync have to do with adoption by monitor makers, NOT to do with whether it's a GREAT technology that ideally EVERY gamer would have. For me a key factor here is going to be the availability of monitors with G-Sync that don't suck. If you can't buy an IPS panel with G-Sync, you're going to have a hard time selling someone like me who games very little compared to doing real work on his PC on it.
Mantle has the capability to dramatically improve performance in the traditional sense, but doesn't address the same issues that every card has the way the G-Sync does. Mantle's success rides on the implementations we see from devs like DICE with BF4. If it's awesome, then it will drive more developers to embrace it. If it sucks and it turns out to be really hard to work with, it won't.
So here's the summary in my mind. They're both potentially incredible. The only difference is right now we KNOW that G-Sync is incredible, and Mantle has the POTENTIAL to be incredible. We still need more information about BOTH of them before I'd be willing to drop real money on either of them, though.
I live in Australia. We basically mirror what the USA do, but with higher prices all round.. Since our population is much lower; competition is not as cutthroat.