GeForce GTX260 with New PCB Design

Status
Not open for further replies.

joebob2000

Distinguished
Sep 20, 2006
788
0
18,980
[citation][nom]razor512[/nom]I bet each video card only cost about $15 to make and by the time it flows through the companies greed, it ends up costing the consumer $500[/citation]

Yes, the boards cost $15 in parts to make. Never mind that the first board they made cost them $80 million, it's the cost to make one *today* that matters, right? In related news, would you like to be the person to buy the $80 million GeForce GTX360 when it comes out?

In case you were completely oblivious to everything around you (on this site especially) there isn't a whole lot of margin on graphics cards since there is stiff competition both within the market (AMD vs Nvidia) and external pressure (Xbox, PS3, Wii) for the consumer's money.
 

vertigo_2000

Distinguished
Feb 26, 2007
370
0
18,780
$15 per card might cover the cost of the parts themselves, but like most companies, you have a lot of other expenses that are essentially passed on to the customer. Heating and maintenance of any buildings owned by the company, interest on any accounts payable, salary and benefits of employees, and the big one for nVidia - R&D of new products --- these are all just some examples of how a $15 graphics card becomes %400-$500 to the consumer.
 
Vertigo and joebob are correct. But let's keep in mind. The article says nVidia's cost would be REDUCED by $10 to $15. They didn't say that's the actual cost to manufacture.

The interesting part will be to see how much this potentially reduces the price for the consumer. I'd love to get into a new GTX 260 or better, but frankly my 8800GTS 512MB does fine for everything I play, and I can't see spending the money they want for the new GTX line.

Either way, this is potentially good news for us consumers.
 

A Stoner

Distinguished
Jan 19, 2009
326
105
18,960
We have successfully hamstrung our device such that it will be rare that a unit will overclock very much over stock and even if it does overclock, it will be totally unstable due to power fluctuations. Our chips are top notch, but the PCB powering the chip is basically cardboard with a little lead free solder. So, to our enthusiast base, who are the only target audience for this product, sorry, to everyone else, who would not buy this product at a $15 discount anyways, we invite you to buy the product.

Hmm, I see a flaw in this logic, but it is nVidia's product.
 

traviso

Distinguished
May 1, 2008
18
0
18,510
I am somewhat excited about this new redesign, but considering I have a 65nm 8800GTS 512mb, this is a decent but only mildly tempting upgrade. I'm really holding out for 45nm before I upgrade my video card.

Albeit I have read in numerous reviews that Nvidia's 55nm designs are equally efficient on heat & power usage as ATI's 45nm chips. Despite, I don't want that extra bit of improvement to really sweeten the deal.
 

hairycat101

Distinguished
Jul 7, 2007
895
0
18,980
[citation][nom]A Stoner[/nom]We have successfully hamstrung our device such that it will be rare that a unit will overclock very much over stock and even if it does overclock, it will be totally unstable due to power fluctuations. Our chips are top notch, but the PCB powering the chip is basically cardboard with a little lead free solder. So, to our enthusiast base, who are the only target audience for this product, sorry, to everyone else, who would not buy this product at a $15 discount anyways, we invite you to buy the product. Hmm, I see a flaw in this logic, but it is nVidia's product.[/citation]

+ 1

2 phase to single phase. This doesn't sound like a good idea to me.
 

A Stoner

Distinguished
Jan 19, 2009
326
105
18,960
[citation][nom]hairycat101[/nom]+ 12 phase to single phase. This doesn't sound like a good idea to me.[/citation]
Exactly right here. And while the normal video card buyer is not going to care much about 2 phase or 1 phase, the people who buy the high end cards actually do care about these things.

I want prices to come down, but price drops for enthusiast level products should come from improved parts, improved packaging, and better processes, not cost cutting that degrades the actual end product. Maybe the real reason for the change is that they are now moving it down the scale, but I do not see $15 moving it down far enough to reach mainstream card buyers, which would be below $150 according to most charts I see.
 

roofus

Distinguished
Jul 4, 2008
1,392
0
19,290
i could see them pulling something like that with the rebranded 9800GTX that was the 8800GTS that is going to be the 2** but not the 260. glad i already got mine sheesh.
 

JonnyDough

Distinguished
Feb 24, 2007
2,235
3
19,865
As recently noted in an article either on here or on Anandtech, just because something is fewer phases (they were talking about a motherboard) does not mean that the electricity feeding into the component is more erratic and less stable. Please stop thinking that more phases = better. That's like saying that just because a highway has more lanes means it's a better road.
 

bounty

Distinguished
Mar 23, 2006
389
0
18,780
More phases = uses more power I think? So maybe the new card will be more power effecient. Also most video cards now days don't OC well anyways, at least not like the extra 40% you get on an E5200 etc. Usually the memory doesn't go anywhere and the rest of the card can get 5-20% that doesn't matter since it's castrated by the memory.
 

A Stoner

Distinguished
Jan 19, 2009
326
105
18,960
[citation][nom]jonnydough[/nom]As recently noted in an article either on here or on Anandtech, just because something is fewer phases (they were talking about a motherboard) does not mean that the electricity feeding into the component is more erratic and less stable. Please stop thinking that more phases = better. That's like saying that just because a highway has more lanes means it's a better road.[/citation]

Got a link? I looked for the review but did not find it. I would like to see what there is.

You are correct about roads, but with all other details being equal, more lanes does equate to more traffic bandwidth, although maybe not a linear scaling. I think that phases though are individual steps down in voltage, and a single large jump from 12V +/-.2V down to 1.125V might tend to have more opportunity for large variations in the final 1.125V where an 8 phase step down process would tend to have a more stable final 1.125V. I am not into electronics, so I have no real clue here, just going off the information I get from these websites and manufacturers.
 

jwl3

Distinguished
Mar 15, 2008
341
0
18,780
razor512 if you're a teenager your ignorance excusable. Otherwise, you're just a moron. There's such a thing called variable cost and fixed cost. Variable cost entails the increased cost from producing ONE MORE graphics board. This is the cost of the materials and the electricity and labor spent on it. This is probably $30 bucks. Then there is the FIXED Cost which is far more significant typically. This is the overhead cost - things like the cost of the machinery to manufacture the goods, the cost of the warehouse, the cost of the factory, the research and development costs, the marketing costs, the employees' salaries for both hardware and software compatibility. Costs like these NEED to be allocated to each board - otherwise, you have to sell the first GTX 260 at $8 million dollars and the others at $30. That's insanity.

So yeah, it's corporate "greed" that drives them to charge $200 a card and not $30. Keep hating corporations you liberal hack. Were it not for corporations and their investments, 80% of the conveniences we enjoy (medicinal drugs, electric and computer goods, heating, air conditioning, cars) would not exist.
 

JonnyDough

Distinguished
Feb 24, 2007
2,235
3
19,865
[citation][nom]A Stoner[/nom]Got a link? I looked for the review but did not find it. I would like to see what there is.You are correct about roads, but with all other details being equal, more lanes does equate to more traffic bandwidth, although maybe not a linear scaling. I think that phases though are individual steps down in voltage, and a single large jump from 12V +/-.2V down to 1.125V might tend to have more opportunity for large variations in the final 1.125V where an 8 phase step down process would tend to have a more stable final 1.125V. I am not into electronics, so I have no real clue here, just going off the information I get from these websites and manufacturers.[/citation]

More bandwidth on one path does not = better performance though. Even if you fill a roadway with cars, if you only have one toll booth open at the end off ramp there will be a bottleneck. Phase change has nothing to do with lanes, it has to do with conditioning of the current. It was just an analogy to point out the pointlessness of the idea that more = better. Less often = more efficient and simpler. It's why I don't do as much decorating as the women I've had in my life. I moved a lot growing up, and after awhile you learn to just minimize your workload and moving costs by having less worthless crap around.
 

superstition

Distinguished
Feb 27, 2009
150
0
18,710
Anandtech did an article about Nvidia motherboards that would blow up with AMD Phenom 125W chips because they didn't have enough power phases. The cheaper motherboards had fewer phases. And - gasp - were the ones that blew up. So, it does seem that better power regulation = less likely to blow up.
 

superstition

Distinguished
Feb 27, 2009
150
0
18,710
"The vast majority of the 780G boards have a three-phase or four-phase PWM circuitry design. These designs are completely acceptable for the 45W, 65W, 89W, and 95W TDP rated processors; however, drop in a 125W TDP processor such as the Phenom 9850e or 6400+ X2 and you are asking for trouble. Trouble is exactly we found, as each board we tested eventually succumbed to the greater power requirements of these 125W TDP processors."

http://www.anandtech.com/mb/showdoc.aspx?i=3279&p=2
 
Status
Not open for further replies.