Prices of graphics cards
- Thread starter Reverend
- Start date
I suspect the answer is as much "tradition" as much as anything. Y'know, good old fashioned "that's what we've always done" understanding across a broad group of stakeholders that has a tremendous amount of inertia behind it. Suggested changes to which are met with fear and distrust, and must be justified by studies and even then will be met with resistance and disquiet.
I mean, look at SLI and the move to roughly $1,000 video "solution". Clearly the initial response from ATI and their folks was "Right! There ain't no viable market at that pricepoint! Harrumph, Harrumph!!" So, y'know, as much as I'm not a huge fan of SLI, I do give NV tons of credit for taking the risk, fighting thru the resistance, and being right in the end, dragging everyone else along behind them.
I mean, look at SLI and the move to roughly $1,000 video "solution". Clearly the initial response from ATI and their folks was "Right! There ain't no viable market at that pricepoint! Harrumph, Harrumph!!" So, y'know, as much as I'm not a huge fan of SLI, I do give NV tons of credit for taking the risk, fighting thru the resistance, and being right in the end, dragging everyone else along behind them.
I don't think NV took much risk (the risk would be on SLI-purchasers) -- I don't think NV's SLI tech involved much costs for NV.geo said:
But this is going OT. How do they set, for example, the sub-100 and over-400 prices for video cards for the respective "market segments"?
Going further, who was the first to set such price "market segmentations"?
I tend to think that the customers set the price point, so the AIB and mobo manufacturers, based on what they think their customers will pay --hence the multiple stakeholders. Look at ATI's experience in IGP this year. What they quickly found out is it doesn't matter if you build a better mousetrap --the mouse-catchers have their price for mousetraps and that's the end of it; hit it and sell, or not.
geo said:
Credit for what? All they did was rehash the old concept of using two cards and marketed it as the best thing since sliced bread with cherry picked benchmarks. Then they used it as an excuse to increase prices because they found there are people willing to pay practically anything for the best, and even though that market niche is very small they also get to use it as PR for having the "fastest solution." There really wasn't any risk involved. Even if it had failed they would have still sold plenty of single cards since the NV40 didn't have any problems like the chip before it.
Nvidia's actions have been bad for the consumer in every way imaginable short of better performance in a limited number of games. So for consumers, getting two cards increases your performance by 10-60% best case while nvidia gets double the profits. Why do you think their revenue is so high?
Due to exchange rates and having to pay more in the UK we don't really have a 99, 199,299,399,499,599 price point scenario but I guess it's just a multiplication of those values from the USA,
Here's an interesting fact, in November 2004 I bought a Asus 6800 Gamer edition at 350/1000 for £270. In November 2005 I bought the XFX 6800GS XXX 480/1100 for £175 !
Here's an interesting fact, in November 2004 I bought a Asus 6800 Gamer edition at 350/1000 for £270. In November 2005 I bought the XFX 6800GS XXX 480/1100 for £175 !
_xxx_ said:
Jen said earlier this year they'd had 3 years developement at that point (i.e. summer '02, and isn't that something to conjure with?). So, yeah, there was some risk if it turned out there wasn't a viable market at $1,000. The brilliance is the initial design doesn't require much cooperation from the other stakeholders initially. . .that had to be a big part of the strategy for "getting away with it" at first.
But that requires another thread in depth. It only fits here tangentially, because certainly the creation of a new pricepoint was a key part of what happened.
I would think that prices for video cards are like prices for pretty much everything else, a combination of product cost (how much the product costs to make/ship/advertise), public perception (how much the public is prepared to pay), and circumstances (like needing to get an edge on the competition, investors pressure for higher margin, competition's product being late...).
Well duh! But that really doesn't mean too muchCorwin_B said:
However, as I have said before I do not believe it seems that ATI and Nvidia are competing on price at the top end really, I think they are just charging the most they both can and sell out the inventory, that is not evil. The only problem is if they are constraining the inventory, but if they just cannot meet demand then the high prices are a natural rationing process that will continue for the forseeable future.
I think they get to a price by a combination of things. I think the first thing that each does though is compare their card to the competition, see where that card is priced, then adjust the overall level of costs for the card. They probably arrive at the base GPU/VPU price from that comparison, and then create the card using components that will allow the manufacturer to achieve that price point yet still make adequate margins. NVIDIA does a lot of balancing on their GPU costs so they don't bankrupt their manufacturers, as well as insure that they make good margins on their cards.
In the case of the 6600 series, NVIDIA really is sitting in a great position with that product. It has been out for more than a year, development costs are pretty much already paid for by how much they have sold to this point, and the cost of manufacturing the GPU itself has gone down quite a bit over the past year. We also see the prices on components go down, and DDR-2 memory is a great example. Definitely not the same stuff as was released with the FX 5800 or the 9800 Pro 256, but dirt cheap for good densities.
As for the high end, I would say that the base cost for a high end card such as the X1800 XT or 7800 GTX 512 is quite a bit higher than you would expect. A manufacturers cost on such a board would probably be around $450 once components and man-hours are accounted for. The manufacturer then needs to make some margin off of the card, and that card then goes to the wholesalers, which then take another chunk, and then onto the retailer (who also needs their cut). Admitedly, the 7800 GTX 512 priced at $749 is somewhat crazy, but we live in a world of supply and demand. NVIDIA priced the card so that everyone could have good margins at $649, but with the price at $749 everyone is getting a bit fatter. Still, how many video cards does Newegg sell vs. how many total 7800 512's they have pushed through? While margins are nice, the rest of their vid card business simply buries it. If there were no demand for the 512's, then they would be sitting on shelves and the retailer would lower the price to get their money moving.
So, while it is probably quite complex to take all of this into consideration, having a competitor and having already established price points, it makes it a lot easier. Now all these guys have to do is make sure their chip R&D doesn't go crazy, and they can find good enough and cheap enough components to make sure everyone in the supply chain has good margins.
In the case of the 6600 series, NVIDIA really is sitting in a great position with that product. It has been out for more than a year, development costs are pretty much already paid for by how much they have sold to this point, and the cost of manufacturing the GPU itself has gone down quite a bit over the past year. We also see the prices on components go down, and DDR-2 memory is a great example. Definitely not the same stuff as was released with the FX 5800 or the 9800 Pro 256, but dirt cheap for good densities.
As for the high end, I would say that the base cost for a high end card such as the X1800 XT or 7800 GTX 512 is quite a bit higher than you would expect. A manufacturers cost on such a board would probably be around $450 once components and man-hours are accounted for. The manufacturer then needs to make some margin off of the card, and that card then goes to the wholesalers, which then take another chunk, and then onto the retailer (who also needs their cut). Admitedly, the 7800 GTX 512 priced at $749 is somewhat crazy, but we live in a world of supply and demand. NVIDIA priced the card so that everyone could have good margins at $649, but with the price at $749 everyone is getting a bit fatter. Still, how many video cards does Newegg sell vs. how many total 7800 512's they have pushed through? While margins are nice, the rest of their vid card business simply buries it. If there were no demand for the 512's, then they would be sitting on shelves and the retailer would lower the price to get their money moving.
So, while it is probably quite complex to take all of this into consideration, having a competitor and having already established price points, it makes it a lot easier. Now all these guys have to do is make sure their chip R&D doesn't go crazy, and they can find good enough and cheap enough components to make sure everyone in the supply chain has good margins.
They designed a simple connector built into the PCB of all their cards, they added another PCIe port sharing the bandwidth with the primary port and they added support in the chipset. None of this was particularly costly and although it did take time, especially in driver development, it was only a small project that yielded big results._xxx_ said:
