Thanks again for a well thought out reply. I do have a few comments though.
First you said that until recently MU had been a technology laggard. I would take issue with this because true technology laggards don't last long in this business. What I would agree with is that until recently MU concentrated on using advanced technology differently than their competitors. They applied advanced technology to older chips to achieve further cost reductions making them the low cost producer of older generation chips while their competitors applied advanced technology to move on to the next generation. Thus MU was still reducing costs at 4MB while competitors were moving to 16MB, and MU was still reducing costs at 16MB while competitors were moving to 64MB. Now competitors are moving to 128MB and 256MB while MU is still working on cutting costs at 64MB. What is different this time is that MU is apparently also working on some advanced products such as RDRAM while the competitors are not forgetting to continue to reduce costs at 64MB. Thus both are adopting parts of the other's strategies.
I would agree with you however that a misunderstanding of MU's strategies caused many people on this thread to consider them to be a technology laggard, and to prematurely write them off as a result. And further I would add that the failure of their gloomy forecasts to come true, combined with a failure to adjust their perception, no doubt led to the vehement posturing.
I further agree that supply and demand will never be in balance, at least not for long. The industry will constantly flux from undersupply to oversupply, but hey, that's what makes it fun. <G> Furthermore with inventories at CPU makers dramatically reduced, the cycles should happen even faster than in the past. Now a small surge in demand could cause a shortage, and a small reduction in end user demand could cause a glut. But that just makes for even more fun. <G>
Your take on capex is much more in line with mine than Skeeter's view. Yes equipment sales are increasing, but from a very low level. Still my equipment stocks are doing quite nicely. <G> But I agree that equipment sales are still low enough that demand will eventually catch up with supply, probably towards the end of this year. Part of the clue lies in RDRAM. RDRAM requires more silicon, so a widespread switch to RDRAM will decrease the supply (or rate of increase) of DRAM helping to shift the balance towards shortage.
I also agree that we won't see any significant increase in dollars spent on DRAM per box by the box makers. To the extent that CPU prices fall, it will get passed on in lower box prices, not transferred to other components.
On the whole I largely agree with your analysis. We both agree that the DRAM makers will recover, and that the recovery will be faster than the semi industry as a whole. I see the recovery a bit sooner than you do, and I am more convinced than you are that MU will be a beneficiary, but basically we aren't too far apart.
Carl
|
