Nvidia Maxwell Speculation

[minerva330]

Apparently, the specs for the 750Ti were leaked over the last couple of weeks. http://videocardz.com/nvidia/geforce-700/geforce-gtx-750-ti

The scuttlebut is its suppose to be released around mid-feb, and the response has been pretty lackluster.

Does anyone know why Nividia would release such a low performing card to kick off their new architecture?

The I reason I think is that they want to test Maxwell on a larger fabrication process, the 28nm before scaling down to the 22nm, in order to work out any bugs.

Lastly, the leap from Fermi to Kepler was about 30% does anyone foresee that being the case with Kepler to Maxwell?

I do not think we will see such a huge leap maybe 20%. I think Nvidia is starting to hit a wall. I also think that until they make a breakthrough dual-GPU cards are where the biggest performance gains will be, like the upcoming GTX 790 for example.

 

[Mousemonkey]

Apparently, the specs for the 750Ti were leaked over the last couple of weeks. http://videocardz.com/nvidia/geforce-700/geforce-gtx-750-ti

The scuttlebut is its suppose to be released around mid-feb, and the response has been pretty lackluster.

Does anyone know why Nividia would release such a low performing card to kick off their new architecture?

The I reason I think is that they want to test Maxwell on a larger fabrication process, the 28nm before scaling down to the 22nm, in order to work out any bugs.

Lastly, the leap from Fermi to Kepler was about 30% does anyone foresee that being the case with Kepler to Maxwell?

I do not think we will see such a huge leap maybe 20%. I think Nvidia is starting to hit a wall. I also think that until they make a breakthrough dual-GPU cards are where the biggest performance gains will be, like the upcoming GTX 790 for example.

This was the reason given by Nvidia, pretty much the same as AMD did with the 4770 a couple years ago.

 

[minerva330]

Yeah it definitely seems like the norm, which is unfortunate for the consumer. Those of us who can only afford the mid-high to high-low end cards kind of get shafted.

I am curious to see what they would even price the 750ti at? The price to performance ratio looks pretty bad on paper but who knows until proper benchmarks are released

 

[Mousemonkey]

Yeah it definitely seems like the norm, which is unfortunate for the consumer. Those of us who can only afford the mid-high to high-low end cards kind of get shafted.

I am curious to see what they would even price the 750ti at? The price to performance ratio looks pretty bad on paper but who knows until proper benchmarks are released

Why is it unfortunate for the consumer when a manufacture wants to make sure that a new process works OK? Surely it would be unfortunate to release a new node that doesn't work as they thought it would?

 

[minerva330]

I completely agree but why can't that be done in house/closed beta?

When a company develops a new drug they don't release it to the general population and see how it goes.

 

[Mousemonkey]

Because it costs money to to have the GPU's made? AMD have done the exact same thing, get over it! It's not compulsory to buy one of these 750's you know?

 

[minerva330]

I realize that. I created this thread to vent and speculate, as the title suggest. Its okay if we have different views.

I understand their reasoning for doing so but in my view if more money was dedicated to research and development (c'mon they can test Maxwell on a 28nm fab process in house) rather than flooding the market with sub-par cards they would be less incremental gains and more substantial ones. I also realize that that is would probably generate less revenue for them, therefore it will never happen. Like I said, just venting

 

[Mousemonkey]

But why would they waste money? That makes no sense.

 

[minerva330]

I don't see as a waste but rather less revenue in short-term. On the flip side how much money do you think they will actually net from the 750ti, granted they have the Fab down but they still have to manufacturer it. Just take that money and reinvest in R&D.

 

[Mousemonkey]

I don't see as a waste but rather less revenue in short-term. On the flip side how much money do you think they will actually net from the 750ti, granted they have the Fab down but they still have to manufacturer it. Just take that money and reinvest in R&D.

I'm guessing you don't run a company, corporation or business?

 

[minerva330]

Nope I am an academic....all the way on the other side of the spectrum

 

[Mousemonkey]

Nope I am an academic....all the way on the other side of the spectrum

In that case it should be child's play to understand why they are doing a "tester" card.

 

[minerva330]

I don't want to focus on such benign things. The major point was that I think that overall, while the strategy makes sense from a business angle it water-downs the market and hurts the consumer. I know businesses are there to make money and more power to them, people buy it up. It my view you can only make a difference with your wallet, and that's what I'll be doing

 

[minerva330]

Well that's why I come here, for insight. Thanks for the discussion it definitely helped.

 

[Mousemonkey]

I don't want to focus on such benign things. The major point was that I think that overall, while the strategy makes sense from a business angle it water-downs the market and hurts the consumer. I know businesses are there to make money and more power to them, people buy it up. It my view you can only make a difference with your wallet, and that's what I'll be doing

An academic who doesn't want to focus on the concept of testing things out? The future of mankind is not looking promising.

 

[chrisso]

Actually, the performance per £ or $ has increased a lot in recent years.
A gtx 460 1gig settled around the £145 mark, and a 650ti is only £90.
It is just faster too, obviously.
Tho I doubt it will last as long as the 460, longevity has gone out the window.

 

[minerva330]

I don't want to focus on such benign things. The major point was that I think that overall, while the strategy makes sense from a business angle it water-downs the market and hurts the consumer. I know businesses are there to make money and more power to them, people buy it up. It my view you can only make a difference with your wallet, and that's what I'll be doing

An academic who doesn't want to focus on the concept of testing things out? The future of mankind is not looking promising.

I think you misunderstood the purpose behind my comments. I am not advocating the completely lack of testing but rather a more focused way of going about it. I also understand it is not practical from business perspective but was merely voicing my opinion.

Honestly, I come to TH to learn and I was hoping this thread would initiate a respectful back and fourth. I am disappointed.

I am relatively new to computer hardware, and I am definitely not a businessman, I am Biologist by training, so this all new. Regardless, thanks for the discussion...it was insightful.

 

[minerva330]

Actually, the performance per £ or $ has increased a lot in recent years.
A gtx 460 1gig settled around the £145 mark, and a 650ti is only £90.
It is just faster too, obviously.
Tho I doubt it will last as long as the 460, longevity has gone out the window.

That's true but do you think they can keep up that pace?