It is highly normal that the ASPs should decline overall for the ASICs FOR A SPECIFIC PRODUCT LINE. but the good item which shows up in your work is that the ASP has gone back up due to new product introduction and new functionality.
As the volume goes up in any one product, the vendor expects and gets volume price discounts. But at the same time, the buyer of silicon gets much the same volume price discounts, so even if the ASP goes down, the margin may well go up from simply volume discounts as well as from seondary cost savings, such as amotization of test equipment, amortization of labor to produce the chip, better streamlined proceedures for manufacturing, etc.. I would suspect also that the total overall margins for silicon got even better, as the chip sector has been in a long slump and I would expect that Qualcomm has gone out and renegotiated all their long term silicon contracts to take advantage of multi-year higher volume commitments with their vendors. It would show great management if they had gone out now and locked into pricing for 3-4 years out.
It is a good sign that the ASP jumped back up as the 1x volume went back up and I expect that it will jump once again when the newer ZIF chips get out there. This is exactly what you would expect. If you doubt it, check out the ASPs of Intel over time. Almost exactly the same. It is no wonder they introduce a new Pentium chip every 4 months, and restep the old ones about every 9 months.
This is why the retained cash is so very important for Qualcomm as it funds more R&D to come up with new innovations to regain the ASP price. Wihtout it, the average chip price would go down and after a long time, the total return on chips would be alot lower. This is also why I was so adamant a couple of months ago about the people who wanted all that free cash turned out to investors. By using it to fund deeper research and to fund more vendors for silicon, they can drive the buying price lower and grow margins faster. they can also do cost savings chips which are done solely to reduce the die size and thus the cost of the silicon.
I think a much better way to look at it is volume and margins combined. ASPs mean almost nothing in the long run as long as your cost goes down commensurate and the volume goes alot higher. Look at someone like National Semiconductor. they sell a ton of OP-AMPS at $0.15, but they only cost them $0.015 cents. they also sell lots of very low cost components, but they sell ALOT of them.
As for the CSM, it may be that they went from a 4 channel chip to a 32 channel chip and the actual cost of the chip was not that much higher. this would reduce the number of chips sold (not the number of CDMA call channels sold) and not increase the revenue much. but the CSM chip is like 1/100th the amount of hte MSM chip for new installtions and probably more like 1/500th or 1/1000 for long term fixed systems which have handset replacements going on year after year. Lots of factors here and I have not seen the data. but I would tend to focus on the MSM chipset sales and pricing for the bulk of the business case your making.
Hope this is helpful. |
