Flat R&D

I’ve made this point before about R&D in the pharmaceutical industry but, apparently, it’s true in the computer industry as well: R&D does not increase with revenues but with inflation (if that). Consider the R&D budget for Dell:

Year Net revenue (billions) Spending on R&D and engineering (millions) RD&E as percent of net revenue
2005 $49.2 $463 0.9%
2004 $41.4 $464 1.1%
2003 $35.4 $455 1.3%
2002 $31.1 $452 1.5%
2001 $31.8 $482 1.5%

Source: Dell 10K filings

Net revenues have increased sharply; RD&E budgets have stayed flat or decreased in real terms; RD&E has dropped as a percent of net revenue.

I recognize that running this lean is a requirement in this day of short-sighted investors but this is just too lean.

When you read complaints that there aren’t enough kids studying science and engineering here, do you think that American kids are lazy or that they know that there’s no future in it?

UPDATE: TM Lutas takes me to task in the comments for using Dell, a “box assembler” as in example saying that such companies just have a different business model. Actually, he’s making my point for me. Dell’s identity as a box assembler has grown over time. Here are some of Dell’s innovations of the past:

1986: Dell (then PC Limited) offered the first 12Mhz 286

1988: Dell uses SRAM to create the zero-wait-state 386

1990: Dell introduces 486 with graphics linked directly to CPU bus

1990’s: Dell incorporates lithium ion into notebook for system that lasts the time of a coast-to-coast flight

1990’s: Dell incrporates fast-charge scheme into notebook; 90 percent of charge done in one hour

That kind of stuff’s being done in Taiwan now, not here. Dell is the top PC manufacturer in the world with about 18% market share and employs roughly 4,000 engineers. That’s not just a box assembler.

I think that more than anything else this transition over time reflects two facts.  First, manufacturing has increasingly gone overseas.  It’s a fact of life that engineering and manufacturing are intertwined and IMO it’s inevitable that engineering follows manufacturing.  Second, a lot of what electrical engineers are doing these days in the United States is creating what’s lovingly referred to as IP: intellectual property, i.e. they’re programming.  More and more devices are going over to highspeed emulators.  There’s a good reason for this:  IP is protected.  Our strong protection for intellectual property here is creating market distortions.  And the movement of manufacturing offshore and the market distortions are working synergistically to make engineering obsolete here.

2 comments… add one
  • Dell is a box assembler. Box assemblers have a peculiar R&D need. They need to spend enough to introduce x new models per year, no matter what. More models merely increase confusion and lower profits. Flat R&D may be a reality or it might not be across the general economy. Picking Dell as your representative company is simply not going to work because a company like IBM has a different business model and thus a different R&D curve. Which of the two is closer to the national reality is something I just don’t have the data to judge. Neither will most of your readers given what you’ve written.

  • That’s wrong, TM Lutas, or at least it was wrong until quite recently. Dell was a technology leader in the 1980’s and 1990’s—that’s how they gained their position. They weren’t the lowest cost supplier. They produced the first 12Mhz 286 machine; they produced the fastest 386’s; the list goes on and on. Dell employs more than 4,000 engineers. That’s not “a box assembler”. You can do that with no engineers.

    IBM is, basically, out of the small machine business (and the disk business as well) so that’s really an apples and oranges comparison.

    The two manufacturers with the bulk of the small machine business are HP and Dell. The situation is similar for both companies.

Leave a Comment