So it says that the Black has all the computational power enabled and that it's double precision math is pretty awesome. Does this mean that this card will be a lot better (along the lines of the 290x) for mining or no? I'm not going to buy one but I'm just trying to figure it all out... Like why did Nvidia create a card with DP math being better unless it made a true difference elsewhere? I thought the 780 TI didnt have all the computational strength and look at how beast mode that card is. Just trying to see if this will "level the mining field of play" sort of thing.

I'm only using my 780 Ti  because I'm trying to save up for another AMD R9-280x. ($220 cheaper) and the AMD card hashes better. As soon as I get my second card, my Ti is getting a well deserved retirement from mining and is going back to gaming. If you are serious about mining, forget Nvidia bro, go with AMD.

With all due respect... neither of you answered my questions nor seemed to read my post at all. I said explicitly I am not buying one but am rather just trying to figure out what the performance is vs. Its predecessors since the computational power is fully unlocked now. I already mine with 3 x 280x and am well aware of how crappy nvidia is vs amd in terms of mining. I asked if since the computational power is fully unlocked and that its double precision is so high if it will indeed perform phenomenally better than any previous card at mining. I have 2 x gtx 780 sc and get 600 hash each. I dont mine on them but rather only game. I only see what each update of cudaminer brings in terms of mining rate to my cards.

houkom
With all due respect... neither of you answered my questions nor seemed to read my post at all. I said explicitly I am not buying one but am rather just trying to figure out what the performance is vs. Its predecessors since the computational power is fully unlocked now. I already mine with 3 x 280x and am well aware of how crappy nvidia is vs amd in terms of mining. I asked if since the computational power is fully unlocked and that its double precision is so high if it will indeed perform phenomenally better than any previous card at mining. I have 2 x gtx 780 sc and get 600 hash each. I dont mine on them but rather only game. I only see what each update of cudaminer brings in terms of mining rate to my cards.

I did read your post, and my answer reflected my viewpoint on the question: the Titan Black will never be worth buying purely for its mining potential, and at that point, there's really no need for a discussion of its theoretical mining power.

To answer your question directly, though, I imagine it will see very similar performance to a 780 Ti (as the GK110 is fully unlocked on both of them), as the only main difference is the amount of VRAM. The 780 Ti KPE under water with plenty of juice will likely give the best hashrate from any NVIDIA card, followed by the 780 Ti Classified under water, and then the 780 Ti Classified on air. I'd put the Titan Black in between the 780 Ti Classified and the 780 Ti SC ACX in terms of mining potential, as the added VRAM will not affect Scrypt mining.

Edit: Forgot to address the DP math part. I really don't think this will have much of an impact on mining. This is probably aimed at video processing (or something else that I don't know about that needs DP math), and is likely something that is more of a marketing point than anything else when it comes to the average consumer.

_Nite_
I wasn't even aware a 750 TI was out, it has a single 6 pin plug too so that pegs my interest (same power consumption as an R9 270 non X), I wonder what hashrate it does.....
 
I'm thinking about getting one cause it would be a great PhysX card replacement for my GTX 460 as well.

265-300 kh/s depending on the OC. The attractive feature of the 750 Ti, imo, is that you can get it without the 6-pin, and still get those numbers with only the power from the PCIe slot, if I'm reading reviews correctly.
The EVGA 750 Tis with the ACX cooler have the 6-pin, but the "reference" model (still custom cooling, just the base model) from EVGA, and even the Twin Frozr 750 Ti from MSI, don't have the 6-pin. I guess it would be nice to have the 6-pin connector to play with overclocking the Maxwell card, but for mining purposes, if it can get that kind of hashrate at 60 watts, it's basically the best card out there for kh/s / W. 

_Nite_
Holy crap really? 265 - 300 KH/s with only 75w power usage from the pci-e slot would be insanely efficent......
 
I'm gonna grab one when I can, its replaceing my GTX 460 for sure then.

That might have been a review for the 750 Ti FTW, which would give them more room to work with an OC (with the 6-pin), but still impressive numbers given that it's not on a Maxwell-optimized kernel on cudaminer. I expect that the the 750 Ti reference will have at least 230-250 kh/s, if not more, especially once cudaminer gets a Maxwell-optimized kernel. 
Maxwell is delivering on NVIDIAs promise of efficiency ;D

For that low of a hashrate, I'm not sure if you can justify that cost for all of the overhead involved with each pcie slot (mobo, cpu, ram, psu, case, risers, flash drive, ethernet cord,etc) unless its hashing 300-350 solid. If it can, expect these to sell out everywhere. 

ROMORC
For that low of a hashrate, I'm not sure if you can justify that cost for all of the overhead involved with each pcie slot (mobo, cpu, ram, psu, case, risers, flash drive, ethernet cord,etc) unless its hashing 300-350 solid. If it can, expect these to sell out everywhere. 

The 270/270x is still a great option exactly for that reason, but for someone who has extremely high energy costs or can't get AMD cards, a 750 Ti rig could be a viable option even if they weren't solidly at or above 300 kh/s. The 750 Ti seems like a great option, to me, as a PhysX card or just a card to throw in the second slot of your main rig and mine with, or for a Steam box or mini-rig (small card with plenty of power). 


We seem to have derailed this thread a bit ;D

_Nite_
Equitum is stealing my idea now I see

Heh, imitation is the sincerest form of flattery :P
If my rig's airflow permits putting in another card (my 780 Ti Classy gets pretty hot under the skyn3t BIOS), I'm definitely stealing your idea instead of throwing the 750 Ti I plan to get in the family computer.

http://www.pcper.com/reviews/Graphics-Cards/Bitcoin-Currency-and-GPU-Mining-Performance-Comparison/Analysis-and-Conclusio
Why is AMD so much faster than NVIDIA?
Obviously from our testing the AMD architecture is much better suited for this mining algorithm than NVIDIA's GeForce cards, but why?  It really comes down to AMD's use of smaller and simpler shader processing units compared to the design that NVIDIA uses; AMD runs more SPs at lower frequencies (1600 SPs on an HD 5870 running at 850 MHz) while NVIDIA runs fewer SPs at higher frequencies (GTX 480 has 480 SPs running at 1400 MHz).  This gives AMD a better ALU output than NVIDIA: 

• AMD Radeon HD 6990: 3072 ALUs x 830 MHz = 2550 billion 32-bit instruction per second
• NVIDIA GTX 590: 1024 ALUs x 1214 MHz = 1243 billion 32-bit instruction per secondSource: Bitcoin Wiki

While in gaming this difference is raw compute power can be offset by other GPU technologies, for the raw mathematical power need for mining Bitcoins (and other ALU-bound work like password cracking) AMD definitely has better utilization of its processing power.
There is another difference between how AMD and NVIDIA operate on the SHA256 keys that affect performance.  From the wiki:
...another difference favoring Bitcoin mining on AMD GPUs instead of Nvidia's is that the mining algorithm is based on SHA-256, which makes heavy use of the 32-bit integer right rotate operation. This operation can be implemented as a single hardware instruction on AMD GPUs, but requires three separate hardware instructions to be emulated on Nvidia GPUs (2 shifts + 1 add). This alone gives AMD another 1.7x performance advantage (~1900 instructions instead of ~3250 to execute the SHA-256 compression function).

Equitum

I did read your post, and my answer reflected my viewpoint on the question: the Titan Black will never be worth buying purely for its mining potential, and at that point, there's really no need for a discussion of its theoretical mining power.

To answer your question directly, though, I imagine it will see very similar performance to a 780 Ti (as the GK110 is fully unlocked on both of them), as the only main difference is the amount of VRAM. The 780 Ti KPE under water with plenty of juice will likely give the best hashrate from any NVIDIA card, followed by the 780 Ti Classified under water, and then the 780 Ti Classified on air. I'd put the Titan Black in between the 780 Ti Classified and the 780 Ti SC ACX in terms of mining potential, as the added VRAM will not affect Scrypt mining.

Edit: Forgot to address the DP math part. I really don't think this will have much of an impact on mining. This is probably aimed at video processing (or something else that I don't know about that needs DP math), and is likely something that is more of a marketing point than anything else when it comes to the average consumer.

I do appreciate you tackling my question head on. Thats what i was hoping for. I know 100% hands down that it is honestly not worth it nor will it ever be (more than likely) but i cant help but still want to know what numbers will look like. Switching back to the AMD side, it STILL makes no sense to buy single cards anymore. Some of the R9 270 (non X) will do 400 hash where a single R9 280x will pull 735 hash (my cards). You could buy 2 x 270's and have them push higher output with less energy than you would with a single 280x. This can easily be said with almost 4 x 270 vs 1 x 290x at current prices. I was mostly trying to figure out what makes AMD cards "so much better" for MINING than Nvidia and I think KillerKanadian fixed that. I was hoping that fully unlocked compute power + more DP math would help but i guess that isnt the answer based on what KillerKanadian pointed out above :( *Edit* as well as _Nite_