Naveen Selvaraj's  Instablog

Very often when Advanced Micro Devices(NYSE:AMD) puts up a poor show, the tendency is to come to the conclusion that Intel(Nasdaq: INTC) is gaining market share and thus doing well.(I have written about how Intel needs to widen its horizons here) As the graph below shows, AMD is quite insignificant by Intel's standards and so AMD doing badly does not mean a bright future for Intel. And for AMD, most of the problems lie within and were acquired(read ATI acquisition) in more bullish times.

Exhibit 1: Revenues- Intel Vs AMD

Source: Gridstone Research

Coming back to AMD, revenues have declined in the last three quarters but so has Intel's. AMD grew the topline better in 2007 due to the ATI acquisition (completed in Dec06 qtr) but since then has lagged behing Intel, but not by much.

Exhibit 2: Quarter Revenue Growth - Intel Vs AMD-YOY

Source: Gridstone Research

Prior to ATI acquisition, AMD was generating better gross margins than Intel. But post that, its been a downhill descent though AMD has done much better in the last two quarters. In Dec08, note how AMD's gross margins took a steep dip due to underutilization of its foundries. Such volatility could have been the reason behind AMD wanting to hive of its foundries as a seperate unit.

Exhibit 3: Gross Margins - Intel Vs AMD

Source: Gridstone Research

So, in terms of top-line and gross margins, AMD has done badly but still is at a comparable distance from Intel. The trouble starts at the operating income level.

Exhibit 4: Operating Income - Intel Vs AMD

Source: Gridstone Research

Rather than margins, I took the operating income figures to compare the two companies so that the extent of losses than AMD has piled up since that fateful ATI acquisition becomes clear. Besides write-offs, the expense overheads were a bit too much for AMD to handle( Note: InJun09, Intel shows a small negative operating income due to a one-off - the european Comission fine)

As a result, AMD had decided some time back to spin-off foundry services into a seperate unit and compete with Intel and Nvidia(Nasdaq: NVDA) based on its R&D strengths and focus on chip design rather than both design and foundry services. While the intention is good, I really don't think the problem lies just there - i.e cutting costs by using outsourced foundry services. The problem lies very much with and in what AMD touts as the area of focus going forward - R&D

Exhibit 5: Quarterly R&D Spending - Intel(RHS) Vs AMD(LHS)

(Note: Intel and AMD numbers are on different scales to aid better comparison)

Source: Gridstone Research

The chart above shows that AMD spends a much higher proportion of sales in R&D. In the latest quarter, Jun09, Intel's R&D spending was at ~15% of revenues while AMD's R&D spending was at ~35% of revenues. Prior to the ATI acquisition, it was at less than 20% of revenues. The graphic below compares the fical numbers and shows how the R&D expenses have increased significantly in the last three fiscals while revenue has been flat.

Exhibit 6: Fiscal R&D Spending - Intel(RHS) Vs AMD(LHS)

(Note: Intel and AMD numbers are on different scales to aid better comparison)

Source: Gridstone Research

While AMD did need time to digest the ATI acquisition and swing into cost-control mode, I feel that it has still not done enough. This might be true for all costs(not R&D costs alone) as the graphic below shows

Exhibit 7: AMD - Revenue And Headcount Growth -YOY

Source: Gridstone Research

AMD's headcount bloated up due to the ATI acquisition and has stayed bloated despite the restructuring initiatives undertaken over the last year. On the other hand, Intel has been displaying headcount reductions from the peak in Sep06 though revenues were increasing at a healthy pace in 2007 and first half of 2008.

Exhibit 8: Intel - Revenue And Headcount Growth - YOY

Source: Gridstone Research

Clearly, Intel has been well ahead of AMD in managing the cost-side and that has been the major differentiator between the two companies rather than just top-line growth. Post ATI-acquisition, AMD could have cut costs more aggressively while integrating the two companies. Rather, it was focussed on taking Intel head-on for market-share and revenue share and that proved its undoing.

While some semblance of stability has returned to demands for computing processors and the graphics segment putting up a encouraging show, AMD can deliver better results with more control of operating expenses. Clearly the current cost structure is not justified based on historical and competitive comprisons and a lot more needs to be done.

Exhibit 9: AMD - Segment Revenues and Growth Rates

Source: Gridstone Research

In the Jun09 earnings call, AMD mentions that it has set a target to reduce quarterly operating expenses to $500M( was $690M in Jun09) and this is one target they have to overshoot! Else, the next few quarters will only add to the string of disappointing news from AMD.

Disclosure: No Positions