Bilow,
I am honestly impressed. A well drafted post that honestly discusses your intepretation of the facts. And I can't say you are wrong. We just have far different interpretaions of the facts.
Your interpretation leans heavily to the fact that there are many other vendors out there who will be producing DDR chips, that DDR will sell at the price of sand (or at least maintain an enormous price gap), and that Intel, out of fear of losing marketshare, will indeed push their darndest to make sure that the 845 DDR chipset will indeed be a success. This I agree with. Intel, if it does put a DDR 845 out there will not be doing so to make a failure. Intel is not Rambus, Rambus is just a component in their systems.
However, my perspective is a bit different from yours. It is that an RDRAM system will provide more value than a similarly priced DDR system. It is not about DDR vs. RDRAM, it is about the system and I simply don't think the price/performance of DDR will provide the consumer a greater value, and this value gap will grow even greater as processor speed increases and newer software to take advantage of that processor speed is on the market.
I know I am thoroughly over-educated. Law degree as well as an MBA degree from Duke (freshly minted) that concentrated in marketing and strategy. One thing I learned from my two years at Duke is that price does not create value. Value is much more than just price. If price were the only thing that mattered AMD would have tossed Intel long ago, generics would dominate the marketplace, and we may have all, perhaps, just perhaps, have chosen to drive Yugos. But it seems we have not all made these choices. Price is just one aspect of value.
So, on this basis, my perspective is a bit different from yours. It doesn't matter if there are 100 vendors putting out DDR (certainly AMD and VIA have already taken a few swings at the DDR ball with a couple of foul tips at best, but they may come through in the clutch, maybe Ali will do it for them) it comes down to the product. DDR has yet to be standardized across vendors, just one strike against its value in systems.
But, indeed, if DDR remains priced below cost, and if DDR systems can be made for say $75 to $100 less than equivalent RDRAM systems, and IFfffff the so far mostly lacking complete remedy stability problems can be remedied on an up-to-Intel quality basis, gaining a similar reputation among businesses and consumers, then yes, DDR systems may have a strong position in the low-end of the market. This scenario is certainly possible and we shall see if DDR does gain this role. But these systems HAVE TO PROVIDE BETTER VALUE and as I stated value is much more than just price.
But that is a difference in perspective. The evidence I see is that (1) the price disparity is small enough between an RDRAM and DDR system that the purchase decision on price will probably not be much affected and that the decision will therefore come down to performance, quality, reputation and service and (2) this gap will only widen in an RDRAM systems favor as performance of CPUs continue to go up and systems become more complex. Almost inevitability.
But I do appreciate your contrary opinion. I don't agree with it, but it does add to the discussion. Thank you. This is the sort of fruitful discussion I enjoy and benefit from and I hope you and others do as well.
Tinker |
