I'm not a stock analyst, but if I were to invest in Intel I would be doing it for the long term. In such a case, it's important to put things in that perspective.
http://www.intel.com/pressroom/archive/releases/20050719corp.htm
The reason why the results this year appear so low is, because last year was a record year for Intel. According to them, last year's Q2 revenue was up 15 percent over Q2 2004, and net income up 16 percent.
http://www.intel.com/pressroom/archive/releases/20040713corp.htm
Looking back at 2004, we see the Q2 revenue was $8.05 billion, which puts this year's $8 billion revenue in line. The difference is that while the revenue is the same, the net income is down from $1.8 billion in Q2 2004 to $885 million Q2 2006.
Now obviously, it'd be nice for last year's result to be the beginning of a trend, but from this year's result, the 2005 year was obviously an anomaly. Still what does that mean for Intel in the long term?
First, with revenue inline with 2004 it's clear that Intel is still selling well, although not as well as in 2005. Second, it is critical to note that while they are selling well, they aren't making as much profit. This we can imply as being the result of all those pre-Conroe price cuts. It's wrong to assume that Pentium Ds didn't sell, they did. The price cuts ensured that. It also ensured small margins though, which is what we are now seeing.
However, in the long term things are cautious, but are looking up. With the arrival of Conroe, Intel now has a full product line. What I mean is that instead of being bottom heavy ($100-$300), there are decent parts that are worth the higher prices, notably the E6700 and X6800. This means that sales will see shifts toward higher margin parts which will help improve profits even if revenue is maintained. This is even more true in the higher margin server market where for the first time in a while Intel appears to have a credible product line. (This is why they worked overtime in launching Woodcrest in June, Montecito in July, and Tulsa in August).
The question goes back to the old Netburst chips of course. However, if the revenues of Q2 are an indication, if you price then low enough people will buy. With additional price cuts coming, sales will no doubt continue to be sustained. The caveat is of course the X2 3800+ and X2 3600+, but the marketing people have to earn their keep somehow. Now, sales of Netburst parts at even lower prices may look good on revenue, but won't do anything for profits. This is of course where strength of server sales and sustaining the mobile market is critical. High-end desktop is also important which is why Intel long ago announced that 3.2GHz Extreme Editions are coming in Q4 and Kentsfield in Q1 2007. Something new every quarter.
All in all, over the long term things aren't particularly grim. How Sharikou can say Intel will be bankrupt in 5 quarters when they're still pulling in net income is beyond me. Still, things certainly aren't all rosy for Intel, which is why I'm not buying. They'll definitely have to ensure penetration of high margin areas to ensure they don't just have sales, but also profits.
For reference the 2003 Q2 results are here:
http://www.intel.com/pressroom/archive/releases/20030715corp_a.htm
Revenue of $6.8 billion and net income of $896 million. Definitely indicates a peak in 2005 and kind of flat overall.