it also does same thing with 8800 so ?
Printable View
Maybe I should have made myself clearer.
It was BECAUSE they made a ton of money (profit) up until the HD4800-series release that they were able to reimburse customers. Their board partners were able to look like white knights riding in for the rescue. I remember well the posts on here and other forums at the time; EVGA, XFX and others were made to look like heroes and positive press followed. It was a brilliant move and it won a lot of people's support thereafter. Without that initial profit, none of that would have happened.
Considering AMD's financial situation, if they get into a situation like Nvidia did, do you really think they would be able to offer the same if they priced their cards with very little margin for real profit? No. Which is why they have to hedge their bets and make as much money as possible before Nvidia hits back. That way they can make the most out of their lead and put some of that profit back into development of future products.
The survey shows that for DX10 cards, it is 3.88% for GTX260, 1.47% for the GTX280, 0.78% for the GTX285, 0.64% for the GTX275, 0.58% for the GTX295. Amazingly, the 4800 series went up by 0.93% in the past month for the Steam survey.
Either way, it is 11.33% for the 4800-series vs. the 7.35% for the GTX 200 series so roughly a 3:2 ratio with regards to the Steam survey of 4800's sold vs the GTX200's
Nice work Zerazax. 11.3% HD4800 & 7.3% GTX200. Seems like atm with this launch, ATi > Nvidia. Like the 4800 & 9700 days. Like the Geforce, 3, Ti, 6000, & 8800 days where Nvidia > Ati. How long before HD5000 series has a fair share of the pie?
No. I am saying that ATI needs to protect themselves in case Nvidia is able to beat their price / performance ratio. It is better for them as a company to give themselves more room to maneuver price-wise when Nvidia comes out with their new cards.
Let's put it this way. Yes, it simplifies the way the supply chain works and assumes profit but it's for clarity's sake:
ATI releases a $299 card and they make $15/unit.
Nvidia then releases their next gen card and prices one of the SKUs at $299. That SKU is able to beat the ATI card in the price / performance category.
ATI can now cut the price of their card by at most $15 before they start loosing money on every card sold.
Now, supplement that $299 with say $349 and suddenly ATI has much more wiggle room when it comes to matching Nvidia's next move.
Hmm, 349 US$ for HD 5870 and 249 US$ for HD 5850 sounds plausible and quite reasonable, if it does have 1600 SP and outperforms HD 4870X2 (HD 5870) or GTX 285 (HD 5850).
I am going to have to disagree as I just posted something similar on the guru3d forums I'm just going to copy and paste it.
--------------------------------------
I'm going to try to explain this the way I am about 90% sure AMD does it's business.
When AMD bought the ATI graphics division, they changed how they do business. Instead of introducing new models and pricing them compared to current models (which over a period of years lead to 800 dollar video cards). AMD I think is smart enough to realize that this kind of model of business will eventually burn itself out.
The majority of consumers aren't going to spend that much on a single piece of equipment (enthusiasts are small number of people), having such expensive hardware makes PC gaming to be a niche market.
AMD can't survive on only catering to a niche market, neither can PC gaming. PC gaming would have probably died if not for AMD buying ATI and changing the dynamic of how they do business.
AMD changed by instead of pricing a new product based on previous products (such as companys own lineup and that of it's competitor), they select pricing tiers and then design cards that will give the most performance per dollar/watt/temp.
Each design cycle is designed around producing a new product at the 199, 249, 299 price points (or 499/549 for enthusiast) and as time goes on or competition occurs the prices will drop lower. I don't work for AMD but I forsee that the launch price of 199-299 is going to be what AMD designs their GPUs for, for the next few years if not more.
They aim to also provide more value at even lower price points in order to provide every customer with the ability to get into PC gaming without having to break the bank. When going against the consoles, PCs have several advantages.
While a console may cost 300-400 dollars (more on launch) you need that full price (and more for controllers cables, etc) to play the games. PC advantage is that most people ALREADY have a computer, so you can drop in 100-175 dollar video card maybe spend 30 bucks on some more ram and you can have a better experience than on a console for the most part.
This is AMD's goal, at least I think it is, as in order for PC gaming (thus video card/CPU companies) to survive is to overwhelm the console systems. So that consoles fade away, but at least PC gaming can be fully cemented as an industry.
But hey, that is just my humble opinion. AMD is going to release their 5870 at the 299 price point and the 5850 at either 199 or 249 depending on specs.
----------------------
AMD is dictating the market by leading and designing products for a lower price point, which allows them to reduce price in response to competition. Just like last round, AMD's costs are far lower than Nvidia, no matter how much more powerful Nvidia's GPUs are - they are bigger and more expensive to produce (including PCB).
AMD has positioned themselves to be able to be more flexible and to dictate market price and style. Nvidia is sort of caught in a poorer place excepting design cycles around larger and larger more powerful GPUs trying to "out do" ATI by going bigger and better.
AMD came along and changed the tempo and fubared Nvidia's mojo. Until probably GT300's successor or later, the trend will probably continue for how AMD leads the market tempo. AMD/Nvidia will have cards priced and performance nearly equal - but AMD's costs will be lower until Nvidia can design a smaller GPU from the ground up.
With any product launch i m sure ATI has done their homework on their target market, and if the performance is as high as some have speculated, I don't think the price is necessarily indicative of performance, being first to market and right before christmas season, ATI could be looking at grabbing nvidia card holders looking to upgrade as well, and with an attractive price ratio could get them to switch brands and commit that money. @ 40nm, assuming the same RnD costs as the 4870, and given the recession ATI might be willing to take a 10-20% profit margin hit, but make up for it with volume... I m sure the resident economist has factored the most efficient theoretical sales path.. But I guess it all depends on the performance level of the card..
Will be interesting none the less, and interesting to see what Nvidia responds with.
The problem with that logic is this: what happens if initial costs on a more complex product means that they can't stick to that pricing structure?
In addition, we are talking actual costs to the consumer here as well. ATI and Nvidia love quoting SRPs to reviewers and consumers. How many times do we actually see those SRPs stick at the retail level? Very rarely, if ever for a new release.
One way or another, I am anxious to see what becomes of pricing in this round.
For sure, and it came from ATI... No doubt! :)
That is the problem that Nvidia has with their larger GPUs as compared to AMD. Forcing them to sell at a lower profit margin due to higher costs. You can either stay at cost and lose market share, or take the profit hit and keep market share (which was what Nvidia did).
That is why performance per mm2 is as important as performance per $$.
Btw, AMD's is making a pretty healthy margin at 299 price point of a x870 card, the 299 is just the starting point for new products. As competition and yields get higher they reduce the price to increase market share further. They aren't stonewall stupid, they know some people wait to get a card to it reaches a certain price point, such as 200 for a x870 type card.
AMD's position is market penetration over low penetration with high margins. As long as they can keep some advantage in die size vs performance they can keep this. Right now I personally think that AMD/ATI has this advantage getting GDDR5 and 55/40nm to the market first. Until either Nvidia leapfrogs ahead (hard to do that in manufacturing tech) or AMD screws up with a new process/technology AMD will prolly have this lead. If there is Nvidia can do is to not have any screw ups and see if they can beat out AMD in performance per mm2, then it won't matter if your chips cost a little more to produce as you will be able to charge more and have more leverage to dictate market prices/tempo.
So why aren't they gaining market share? Even with Nvidias supposedly low margins they're still making money (before the warranty charge). Thus far there are no tangible gains in marketshare or profits. A lot of goodwill though so that could eventually affect the bottom line.
I don't think people should underestimate what market share can do for a company. Look at MS - even though Apple and Google have nipped at them, it still rules 90% of computers in the world, and thus, they have immense clout in the tech industry.
And in the case of PC game developers, having market share can dictate whether companies decide to work with your hardware or not. Lets say that Nvidia had 90% of the market share - well first, most if not all companies would code their games to be optimized for Nvidia hardware. Also, things like PhysX and CUDA wouldn't be controversial - companies would say that 90% of their people can run it, so go ahead and integrate it.
Now because Nvidia and ATI are more closely locked than even just 1.5 years ago, companies are more reluctant to code their games in favor of one company or another. And they are very slow to adopt standards that have specific hardware requirements to it (hence things like Havok are being used as alternatives and OpenCL sometime to follow).
I don't have the time to post what one of the AMD guys (I think it was Dave Baumann) posted on B3D a while back but basically in summary, he stated that there was a range of 7 ways developers can code their games. On each extreme, the game company could code optimized entirely for one hardware developer and not the other (One can see how Lost Planet was heavily Nvidia favored, and Call of Juarez was heavily ATI favored with a huge drop in performance when switching to the unoptimized hardware). You also have cases of being slightly more optimized for one or the other (Crysis and Assassin's Creed for Nvidia and ATI respectively). Then you have games that are coded with both in mind and tend to be balanced with respect to hardware (as in equivalent hardware won't see a sudden drop in performance going from one card to the other).
And believe me, that's a HUGE benefit to the company that can get people on their side. It's no surprise that Nvidia's most recent time of dominance (the 8800 series) coincided with the height of TWIMTBP. Then you have Crysis, which is a huge marketing tool for Nvidia - because Nvidia cards tend to do a little better in Crysis, and Crysis is graphically intense, people say - oh hey, Nvidia must be faster even if the cards were equivalent elsewhere. This is just the nature of the business - say if DiRT2 is the graphical standard and comes out and favors the next gen ATI cards, people will say "Oh hey the ATI card is faster!" even if it isn't in other titles.
Yes, AMD can use the profits, but it bears notice that losing market share for short term profits can hurt them long term.
edit: Another example. The PS3 vs. XBOX 360 (I won't count the Wii here since it isn't in the same market both hardware and target-wise.) The XBOX 360 gained a lot of momentum from price cuts and having a market share to the point where developers were less inclined to make PS3-only exclusive games. In fact, many titles that were once Sony's mainstays are now also available on the 360. And now some developers have openly stated that they're abandoning the PS3.
And all this despite the serious disadvantages the 360 did have (pay to play online, major hardware failures, losing end of the HD war, etc.)...
Market share did move towards ATI's favor a bit since the 4800's release (I don't remember the #'s, it was done by that dude that reports on the GPU industry) - granted, a lot of the data is clouded out there since people need to differentiate discrete from integrated and total #'s etc. If the discrete notebook market has it true that ATI is now > 50%, then their development has certainly trickled down to the notebook market.
As for Nvidia making money? They're still beasts at marketing and PR and though they were probably outsold this last generation, their prices still gave them profits (albeit probably smaller profit margins) and with lots of cash from the better years saved up, they weren't in any real trouble
Good points...