Maxwell wins in performance per watt because it was designed with power efficiency in mind and is much newer than GCN, even GCN 1.2.
No one is denying NV the win in performance per watt, we're just wondering why that's the only metric that matters. It's a little bit troubling when a $300 card that's 2.5 years old rivals the performance of a 1 year old $600 card, even though said $600 card is designed only for the task the $300 is matching it in. That's what's going on in Quantum Break, Hitman, Ashes of the Singularity and Killer Instinct, remember?
I think the argument that the haters make goes something along the lines of: "blah blah, not the best bang for the buck, blah blah no Async Compute, blah blah NVIDIA sells overpriced crap."
Quantum Break is a broken, poorly coded POS console port. AoTS seems to be just a technicality away from being an AMD-sponsored tech demo and is hardly a big seller, and Killer Instinct also looks like a poorly written, broken game.
I haven't followed Hitman.
I think it will be a long time before this "supposed" AMD edge in DX12 really matters to gamers, and it's not even clear that by the time DX12 is popular and widespread that AMD will have a clear advantage.
Quantum Break is a broken, poorly coded POS console port.
Most games are console ports these days, especially AAA titles. You say it like it's a shock. lol
Better get used to it man, all the AAA Xbone games that gets put on PC will be DX12.
Basically it's like GimpWorks vs NV, but AMD doesn't have to get involved or sponsor them at all. It comes gimped automatically for NV GPUs because they are bad at DX12.
Yes, I do. Your reasoning for why it will be a soft launch while AMD's Polaris will be a hard launch makes absolutely zero sense.
Right, sorry
You place too much faith in TSMC's ability to fill the wafers for both the P100 and any consumer GPUs basically at the same time, esp with GP100 being so gigantic.
It's OK. Even if what you say is true and DX12 runs worse on Maxwell than on GCN, I am sure the 1080 (TI?) cards that will replace the 980 Tis I'm using now (and will be selling shortly) will make it all better
Most games are console ports these days, especially AAA titles. You say it like it's a shock. lol
Better get used to it man, all the AAA Xbone games that gets put on PC will be DX12.
Basically it's like GimpWorks vs NV, but AMD doesn't have to get involved or sponsor them at all. It comes gimped automatically for NV GPUs because they are bad at DX12.
So broken stuff is ok since you are a fan of a single IHV? Wow. Kinda goes against the meme that the ADF used to push. Not surprising in any way on this forum. Kinda seems like the tess thing.....AMD wasn't good at it so it shouldn't be used so much..........
I suspect karma will raise it's head sooner or later. It always does for either company in the GPU business. In the mean time I will enjoy what I can and not sweat "fairness" "milking the customer" etc.
Kind of odd to lock yourself into the 1080ti without knowing the performance of vega.It's OK. Even if what you say is true and DX12 runs worse on Maxwell than on GCN, I am sure the 1080 (TI?) cards that will replace the 980 Tis I'm using now (and will be selling shortly) will make it all better
Some of us saw this coming a few years ago. We were called all sorts of names when we said NV's uarch was weak for DX12. Reality will hit home in due course. Now, the only thing that we should be concerned with, is whether Pascal is any good at it.
Great, so your 980 Tis degrade to 390/X level, and you just happily buy another $700 card from the same company?It's OK. Even if what you say is true and DX12 runs worse on Maxwell than on GCN, I am sure the 1080 (TI?) cards that will replace the 980 Tis I'm using now (and will be selling shortly) will make it all better
Great, so your 980 Tis degrade to 390/X level, and you just happily buy another $700 card from the same company?
No wonder NV makes so much money with loyal buyers like this.
All Maxwell has left is its perf/watt, and I don't see why that's really important when we're putting these things inside 30 pound ATX towers. It's not like Maxwell cards produce no heat and are absolutely silent either (I had a 980, and it put out quite a bit of heat, and certainly wasn't silent). Is that extra 30 - 100W really that important?
The assumption in your hypothetical was that this happened, and if it did, giving NV another $700 would make it all better.No proof that this degradation will happen across a broad range of games.
I don't question why you bought a 980 Ti, I don't deny that right now the Fury X completely sucks, I question why you'd be so eager to buy another mythical "future" NV flagship that you're sure will fix all their DX12 woes, even though your last ones crashed and burned a year after they debuted. That was what you said you'd do.They make so much money because at the time of purchase they offered me cards that offered the best-out-of-the-box performance on the market and overclock really well. I genuinely don't understand why somebody would sacrifice performance in today's titles for yet-to-be-released mythical "future" titles.
Yet, AMD leads in performance amongst DX12 titles, the arena we're discussing right now, the reason some are saying that GCN is overall the superior arch. Maybe NV is still ahead in perf/w in DX12 titles (250W vs 275W), but it's behind in performance, and not by an insignificant margin. In Ashes, for example, the Fury X is ahead of the 980 Ti by 23%. (HardOCP, 980 Ti DX11 vs Fury X DX12) Considering the around 10% overclocking headroom on the Fury X, I don't see a definitive winner on either side there, even with the 980 Ti max overclocked. If other DX12 games display similar behavior to this, NV may not be the performance leader forever.Perf/watt leadership at a given wattage -> performance leadership. My overclocked 980 Tis utterly obliterate anything I could have bought from AMD, stock or overclocked.
No proof that this degradation will happen across a broad range of games.
They make so much money because at the time of purchase they offered me cards that offered the best-out-of-the-box performance on the market and overclock really well. I genuinely don't understand why somebody would sacrifice performance in today's titles for yet-to-be-released mythical "future" titles.
Perf/watt leadership at a given wattage -> performance leadership. My overclocked 980 Tis utterly obliterate anything I could have bought from AMD, stock or overclocked.
So the good games are judged by how your 980Ti's perform? If they slaughter the other offerings they are done right? If it's close their poorly coded? If it looses it's a broken console port?
So the good games are judged by how your 980Ti's perform? If they slaughter the other offerings they are done right? If it's close their poorly coded? If it looses it's a broken console port?
Best post on this forum since a while. Must be sticky :thumbsup:It all comes down to perception. When people bought 980Tis they bought what was, at the time the top consumer graphics card available. Their expectation is that it will continue to be the top graphics card. Any data that doesn't confirm their expectation, will automatically be suspect in their opinion.
Other people bought AMD cards because their expectation was that since previous AMD cards had continued to improve relative to their competition, their purchase would continue to improve in relative performance. Any data that tends to show that expectation will automatically be perceived as valid since it matches their expectations.
The reality as I see it is that we're dealing with two different mindsets for two different perceptions of value; short term and long term. The people who prefer short term value want the best card for a short period of time (1-2 years) after which they will replace their card, while the people who prefer long term value want their cards to last 2-4 years or longer.
Neither of these is right or wrong since they are personal preferences, however what happens is that in discussing the value of different cards by different vendors since people aren't talking about the data relative to the same baseline expectation people end up talking past each other instead of with each other.
Just my $0.02 discounted for inflation.
980 Ti looks like it's winning and doing so at a higher performance point while using significantly less efficient GDDR5 memory versus the expensive and boutique HBM1. Proper comp is Fury X and it looks like the 980 Ti is ripping that one to shreds.