The Pentium M on the desktop: DFI’s 855GME-MGF

Their SAT scores are lower, too.

Prescott’s weaknesses have prompted an abrupt change in direction at Intel. Most fundamentally, the company’s faith in one of the corollaries of what used to be called Moore’s Law has apparently been shaken. Ever-higher clock speeds are no longer a given. The 4GHz version of the Pentium 4 was slated to appear before the end of 2004, but Intel axed those plans. The company has instead introduced a new model numbering system that deemphasizes clock speeds and focused its future developments on dual-core versions of its processors.

Meanwhile, silent computing and small form factors are on the rise in desktop systems, as consumers become more aware of PC features beyond raw performance. Against this backdrop, certain corners of the market have fixed their gaze firmly on a tantalizing alternative: the Pentium M processor, optimized to deliver solid performance combined with miserly power consumption, that has seen great success in the mobile market as part of the Centrino platform.

Now comes DFI with exactly what we’ve been asking for: a desktop motherboard for the Pentium M. The DFI 855GME-MGF transcends boundaries by offering Pentium M support in a microATX mobo with an AGP slot and some decent overclocking options. But can the Pentium M really go toe to toe on performance with the Pentium 4 and Athlon 64 when paired up with desktop hard drives and video cards? We’ve gathered up a Pentium M “Dothan” processor and an extensive array of competitors, ranging from an Athlon 64 3200+ at 2GHz to a Pentium 4 at 3.8GHz, in order to find out.

The motherboard
DFI’s ever-so-snappily named 855GME-MGF is an unassuming microATX-sized desktop motherboard with a more or less complete slate of slightly dated options, including AGP, PCI, and PCI-X slots, but no PCI Express. The board itself is wears a handsome black color and sports passive coolers for the north and south bridge chips.

You can see most of the important bits in the photo there, but you’ll have to squint. Let me drop the specs on you to save you the eyestrain.

The mobo is limited somewhat by the Pentium M processor and by Intel’s 855GME chipset, which is now being replaced in newer Centrino laptops by a mobile version of the 915 chipset with support for PCI Express, DDR2 memory, HD Audio, and a 533MHz front-side bus. The 855GME’s front-side bus officially tops out at 400MHz, and its single channel of DDR memory won’t reach past 333MHz without overclocking—not exactly the stuff of legend. The motherboard is pretty well appointed, though, with a Gigabit Ethernet chip and six-channel audio (although both are based on relatively inexpensive Realtek chips).

Overclocking the bejeezus out of the Pentium M
Fortunately, DFI has given the 855GME-MGF a handful of key overclocking options, including adjustments for the front-side bus speed (purportedly up to 1GHz, although the practical limits seem to be much lower than that) and the CPU multiplier. Yes, thanks to the Enhanced SpeedStep clock throttling mechanism, retail Pentium M processors come with unlocked lower multipliers. This fact interlocks nicely with the DFI mobo’s support for a 533MHz front-side bus and its clock lock for the AGP and PCI busses.

In testing, I was able to get my retail Pentium M 755 CPU, whose stock speed is 2GHz, running stable all day long at 2.4GHz on a 533MHz bus. In fact, the system was wholly stable on the first try, with very little drama. The CPU darn near ran at 2.53GHz, as well. I could get it to boot into Windows and run deep into 3DMark05 before crashing, but I never could achieve 100% stability at 2.53GHz. I suspect I might have had more success if the DFI board could push the CPU voltage one or two steps past its 1.34V limits. The DFI board also lacks voltage adjustments for system memory and the north bridge, so it’s not an ideal overclocker like the boards in DFI’s LANParty line.

In part due to the RAM voltage limitations, I decided to keep the DIMM speed at a 4:3 ratio to the PCI/AGP clock, yielding a (roughly) 178MHz memory clock, or 356MHz once you take DDR memory’s clock-doubling mojo into account. With more juice to the RAM and more relaxed timings, the memory might have tolerated a 5:3 ratio and its corresponding 221MHz clock speed. Without extra juice, the system wouldn’t POST with the RAM at 221MHz, even with relaxed 2.5-3-3-6 timings.

Nevertheless, the 855GME-MGF overclocked the Pentium M easily and efficiently to 2.4GHz while keeping the AGP/PCI bus speeds in spec and allowing me to drive the RAM faster than the chipset’s usual 333MHz limit. That’s pretty smooth, all things considered.

All of this overclocking activity was taking place, by the way, using DFI’s relatively dinky cooler for the Pentium M. This cooler is a custom job included in the box with the motherboard. The cooler screws into a plate that has four screw holes and goes under the motherboard. Unfortunately, this cooler is likely your only option. AOpen makes a competing desktop Pentium M board that will accept larger, more capable Socket 478-style coolers, but there’s not enough clearance around the CPU socket on the DFI board for such things. As it stands, DFI’s cooler has a low enough profile to fit into many a slimline case for home theater PCs, and the board’s built-in Intel graphics might allow one to get away without a video card, so long as the drab combination of minimal 3D acceleration and a VGA output is acceptable.

Of course, things get more interesting when we fill up that AGP slot with a fast graphics card and compare the Pentium M to some desktop processors, which is what we’re about to do.

Dothan exposed
Before we dive into the test results, let’s have a quick review of what makes the Pentium M unique. The quick-and-dirty line on the Pentium M is that it’s a Pentium III core mated to a Pentium 4 bus, and that’s not entirely inaccurate. However, the Pentium M is much more than just that.

Yes, it is based on the Pentium III, or more properly, the P6 core that started out in the Pentium Pro processor, which evolved into the Pentium II and then Pentium III. And the Pentium M does use essentially the same bus protocol as the Pentium 4, quad-pumped and everything. But the Pentium M has been extensively modified for better performance, higher clock speeds, and lower power consumption. In fact, the Pentium M’s main pipeline is somewhat longer than the 10 stages in the original P6 core, although Intel is coy on exactly how many stages are involved. The number is probably closer to the 12 stages in the Athlon 64 than to the 20 stages in the original Pentium 4 Netburst architecture or the 31 stages in the P4 Prescott. Other factors aside, longer pipelines generally mean higher clock speeds and lower clock-for-clock performance. As we’ll see, the Pentium M hits clock speeds similar to the Athlon 64 and delivers comparable performance at those speeds.

The Pentium M we’re playing with here is actually the second generation of Pentium M, code-named Dothan. (Our review of the original Pentium M “Banias” core is here.) Dothan is manufactured on Intel’s 90nm fab process, and it packs a healthy 2MB of L2 cache RAM onboard (along with the corresponding logic for prefetching data into the cache.) That’s in addition to a 64KB L1 cache evenly subdivided between data and instruction caches. Thanks to the die shrink, Dothan’s 140 million transistors are packed into a die that’s only 84mm², nearly the same size as the original Pentium M Banias core, which had only 1MB of L2 cache. Compare that, if you dare, to the P4 Prescott’s 122mm² die size, or the massive 192mm² die of the 130nm Athlon 64. The 90nm Athlon 64 “Winchester” also has an 84mm² die, but that chip has only 512K of L2 cache. I don’t have the exact numbers, but I believe 90nm Opterons with 1MB of L2 are expected to be about 100mm².

The impressive thing about the Pentium M is that the entire processor core was designed, massaged, and tweaked in order to cut down on the amount of power it required. Intel’s Israel-based design team used extensive statistical analysis in order to guide its decisions in making tradeoffs between performance and power consumption, and the Pentium M CPU is the result of that process. That’s not to say that the Pentium M is full of compromises that harm performance. To the contrary, some of the very best types of power optimizations are performance enhancements, because getting work done in fewer CPU cycles can save power. Also, the Pentium M team didn’t lean too aggressively toward saving power because the CPU is only a small part of overall system power consumption in a laptop, where things like the hard drive and LCD display can dominate the battery life equation. For these reasons, the Pentium M may very well make good sense as a desktop processor, even when raw performance is one of the user’s primary concerns.

Intel has produced some very informative papers on the Pentium M’s design, and I can’t go into too much depth about such things here, but I would encourage you to read them if you would like more info. There’s one on power savings and another on microarchitecture and performance. I will give you the highlights, though, of some of the changes made to increase the Pentium M’s performance and power efficiency. Among them:

• Dynamic clock gating — Dynamic clock gating is, essentially, the ability to turn off unused portions of a chip and turn them back on as needed. Doing so requires extra logic inside the chip. Too much additional logic can diminsh the power-saving effects of clock gating, so the Pentium M’s clock gating is fine-grained, but not overly so. The Pentium M’s designers used some clever techniques in order to keep unneeded transistors inactive. For instance, the register files in the register renaming units are partitioned by data type, so that only the data width necessary is accessed. If the data being processed is in 32-bit integer form, there’s no need for 80-bit floating-point-sized registers to be active.
• Lower leakage transistors — In some cases, like in the L2 cache of the original Banias Pentium M, Intel used lower leakage transistors that required less power at the expense of speed. Doing so might increase cache latencies or limit peak clock speeds, but it can also save lots of power.
• A new branch prediction unit — The Pentium M’s branch prediction unit is based on the Pentium 4’s, but it’s significantly enhanced. Accurate branch prediction is crucial for performance in any modern CPU, but it’s especially crucial for power savings, because branch mispredictions amount to wasted energy. The Pentium M team added a loop detector to the branch prediction unit in order to enhance handling of program loops with lots of iterations, and they added an indirect branch predictor that better handles data-dependent indirect branches, as often found in object-oriented code. The branch prediction unit was further tweaked in the Dothan core, as well.
• Micro-ops fusion — Like most modern x86 processors, the P6 is a RISC-like core coupled to an x86 instruction decoder. This decoder translates x86 instructions into micro-ops, or instructions that execute on the RISC-like core. Sometimes, x86 instructions decode into multiple micro-ops and execute in a way that’s not entirely efficient. For instance, the store instruction becomes two micro-ops, one that calculates the address and another that writes data to that address. The Pentium M’s decoder fuses these into one micro-op and keeps them largely united as they’re processed. Only at the execution level, when necessary, are they decoupled.

  Intel claims micro-ops fusion cuts micro-ops by over 10% in Banias, leading to performance gains of 5% for integer code and 9% for floating-point. The additional logic for micro-ops fusion does consume more power, but Intel says the additional performance offsets this effect—an instruction sequence requires less energy to complete. The Dothan core apparently fuses even more instructions, although we don’t yet have any details on which or how many.

• A dedicated stack engine — This logic, situated near the instruction decoders, manages the updating of the hardware stack pointer register, again cutting down on the number of micro-ops that must be executed. Intel says this more efficient internal housekeeping cuts micro-ops by 5%.
• Enhanced SpeedStep — The last item on my list may be the most familiar to many of us. Intel’s Enhanced SpeedStep varies both clock speeds and CPU core voltages in order to conserve power when the CPU isn’t entirely busy. Enhanced SpeedStep has multiple “gears” and can step speeds up very quickly on demand in order to keep system performance snappy.

Like I said, those are just the highlights. In some ways, the Pentium M’s design philosophy is the polar opposite of the Netburst architecture in the Pentium 4, which dedicates lots and lots of transistors to enabling higher clock speeds without much regard for power consumption. In others, the Pentium M is very much a modern Intel processor that shares features with the Netburst designs. Those features include SSE2 instructions, by the way, but not SSE3.

Something that is not CPU architecture
My head hurts. Let’s look at some pictures.

Here’s a picture of our Dothan-based Pentium M 755 CPU. The flip-chip package leaves the chip exposed, with no metal cap to protect it. That gives us a nice view of the relatively small chip, which is situated on an organic package the same size as the one used for Socket 478 versions of the Pentium 4.

This is a side-by-side shot of the underbellies of a Socket 479 Pentium M and a Socket 478 Pentium 4. As you can see, they’re vastly similar, with only a pin or two of difference.

The socket on the DFI motherboard uses what looks like a standard laptop-style retention mechanism for the CPU. Instead of a pushing down a lever, one has to turn a screw in order to lock the Pentium M into place.

Finally, here’s a quick visual comparison of Pentium M and Pentium 4 coolers. DFI’s Pentium M cooler looks to be an all-aluminum affair that’s not very heavy, while the Pentium 4 cooler is a part-copper affair with a more complex design and quite a bit more weight.

Test notes
I should say up front that this isn’t an entirely fair fight. This DFI board may be a desktop unit, but it doesn’t have dual-channel memory or even DDR400 support. The rest of the systems are using dual-channel configs with RAM running at at least 400MHz. This discrepancy should be remedied by the next wave of Pentium M boards based on the new Alviso chipset.

That said, you’ll want to watch several comparisons closely. The Athlon 64 3200+ runs at 2GHz, just like the stock-clocked Pentium M. The Pentium M has quite a bit more cache than the 512K on the Athlon 64 3200+, but the Athlon 64 has the advantage of a much better memory subsystem. That comparison should be interesting. Similarly, the Athlon 64 4000+ has 1MB of L2 and runs at 2.4GHz, the same speed as our overclocked Pentium M. That should make for a nice matchup on a clock-for-clock basis. Finally, feel free to compare any flavor of Pentium 4 to the overclocked Pentium M. There’s little point in clock-for-clock comparisons between the Pentium M and Pentium 4, because the P4 has that radically long pipeline and is comfortable at much higher clock speeds. The real question, in any of these cases, is absolute performance.

Please note that several of our test CPUs are actually underclocked versions of other products. Specifically, the Pentium 4 model 540 and 550 entries are actually our Pentium 4 560 3.6GHz review sample running at 3.2 and 3.4GHz, respectively. Similarly, the 130nm version of the Athlon 64 3500+ is a down-clocked Athlon 64 3800+, and our Athlon 64 3200+ results were achieved by testing the 90nm Athlon 64 3500+ at 2.0GHz. For most intents and purposes, save perhaps for our power consumption tests, these underclocked processors should perform just like the real deals.

Our testing methods
As ever, we did our best to deliver clean benchmark numbers. Tests were run at least twice, and the results were averaged.

Our test systems were configured like so:

All tests on the Pentium 4 systems were run with Hyper-Threading enabled.

Thanks to OCZ for providing us with memory for our testing. If you’re looking to tweak out your system to the max and maybe overclock it a little, OCZ’s RAM is definitely worth considering.

Also, all of our test systems were powered by OCZ PowerStream power supply units. The PowerStream was one of our Editor’s Choice winners in our latest PSU round-up.

The test systems’ Windows desktops were set at 1152×864 in 32-bit color at an 85Hz screen refresh rate. Vertical refresh sync (vsync) was disabled for all tests.

We used the following versions of our test applications:

• Cachemem 2.65MMX
• SiSoft Sandra 2004 SP2.b
• Compiled binary of C Linpack port from Ace’s Hardware
• DOOM 3 1.1 with trdelta1 demo
• Far Cry 1.2 with tr3-pier demo
• Unreal Tournament 2004 v3323 with trdemo1
• 3DMark05 v110
• Sphinx 3.3
• LAME 3.96.1 (build from mitiok.cjb.net)
• Xmpeg 5.0.3 with DivX Video 5.21
• Cinebench 2003
• POV-Ray for Windows 3.6
• picCOLOR v4.0 build 479
• WorldBench 5.0

The tests and methods we employ are generally publicly available and reproducible. If you have questions about our methods, hit our forums to talk with us about them.

Memory performance
We generally kick off our tests results with synthetic memory performance. These tests aren’t necessarily an indicator of real-world performance, but they are tied to performance, especially in certain types of applications. Let’s see how the Pentium M, with its slower front-side bus and memory subsystem, measures up here.

The memory bandwidth picture isn’t pretty, as one might expect. However, Linpack shows us why the Pentium M may not mind so much.

What you’re seeing here is the effect of Dothan’s massive L2 cache, which is as big as the L3 cache on the Pentium 4 Extreme Edition. Even when we compute data matrices up to 2MB in size, the Pentium M barely slows down, because it barely has to go to main memory. I should point out that Linpack is also a measure of floating point math performance, and the Dothan’s peak performance at 2.4GHz matches that of the Athlon 64, which is good company to keep. The Dothan’s sustained performance across large matrix sizes, though, is even better. The Athlon 64 tends to slow down as it transitions from its 64KB L1 data cache into its L2 cache, as you can see on the graph.

Interesting stuff. Even though the Pentium M’s front-side bus is half the speed of the Pentium 4’s and its memory is running slower, its memory access latencies are similar. The same is true for the Pentium M at 2.4GHz on a 533MHz bus versus the Pentium 4 Extreme Edition on a 1066MHz bus. The Athlon 64’s integrated memory controller still can’t be touched, though.

Memory performance (continued)
Here’s a slightly indulgent look at memory access latencies in more detail. If the following intimidates you, just skip to the next page with the gaming results. Remember, though, to flip back here if the boss is looking over your shoulder.

I’ve colored the data series below according to how they correspond to different parts of the memory subsystem. Yellow is L1 cache, light orange is L2 cache, and orange is main memory. The red series, if present, represents L3 cache. Of course, caches sometimes overlap, so the colors are just an interesting visual guide.

The access latency graph on the previous page gave us a small slice of the info above, and it was largely representative. Notice how remarkably similar the numbers are for the Pentium 4 Extreme Edition at 3.46GHz and the Pentium M at 2.4GHz. The Pentium M, however, appears to have lower access latencies to its L2 cache, especially at larger block and step sizes.

Doom 3
We’ll begin the real-world tests with Doom 3. We tested performance by playing back a custom-recorded demo that should be fairly representative of most of the single-player gameplay in Doom 3.

Holy poop! The Pentium M blows away the Pentium 4 in Doom 3. At 2GHz, the Pentium M beats the P4 Extreme Edition 3.4GHz and nearly ties the Pentium 4 570J at 3.8GHz. The 2.4GHz Pentium M leaves the P4 in the dust and beats out the Athlon 64 3800+, which also runs at 2.4GHz. Very impressive.

Far Cry
Our Far Cry demo takes place on the Pier level, in one of those massive, open outdoor areas so common in this game. Vegetation is dense, and view distances can be very long.

Far Cry’s results show the Doom 3 scores weren’t a fluke. The Pentium M isn’t quite as stellar here, but it’s awfully darned good.

Unreal Tournament 2004
Our UT2004 demo shows yours truly putting the smack down on some bots in an Onslaught game.

The Pentium M at 2.4GHz whups up on the Pentium 4 in UT2004, too.

However, could the picture change during actual gameplay? Some folks from Intel suggested to us that we should consider testing gameplay performance with the FRAPS frame rate capture program instead of relying on an in-game benchmarking function. The suggestion makes some sense, because timedemo playback tools don’t always use every aspect of the game engine, such as physics, A.I., and user input routines.

I tried using FRAPS with a couple of games, including Doom 3 and Rome: Total War, but frame rate caps in those games prevented us from being able to show meaningful performance differences between different processors. UT2004, which is very much a CPU-bound game, was a different story. The results below are the averaged from five different 150-second gaming sessions played on the same Onslaught map as in our timedemo above, ONS-Torlan. I was playing against computer-controlled bots, so UT2004’s A.I. was working overtime.

The in-game results don’t differ too much from what we saw above. The Pentium M’s real-world gaming performance is very solid.

Before we move on, we tried one more thing with UT2004. We tested CPU performance using its software renderer, just to see what would happen.

What happened is that the Pentium M took the rest of the processors to the woodshed. If you happen to have a Centrino laptop with lousy 3D acceleration and you want to play UT2004, the software renderer might be a good option.

3DMark05

The Pentium M isn’t as at home with 3DMark05 as it is with current games. This benchmark is intended to simulate future games’ workloads, and in doing so, it’s very bandwidth intensive. With its slower bus and memory, the Pentium M can’t keep up here.

WorldBench overall performance
WorldBench uses scripting to step through a series of tasks in common Windows applications. Also like those benchmarks, WorldBench produces an overall score for comparison. More impressively, WorldBench spits out individual results for its component application tests, allowing us to compare performance in each. We’ll look at the overall score, and then we’ll show individual application results alongside the results from some of our own application tests.

The Pentium M is very competitive overall in WorldBench, which is a very good sign. It ties the Pentium 4 570J and Athlon 64 4000+. Only the pricey Athlon 64 FX-55 can surpass it.

Audio editing and encoding

LAME MP3 encoding
We used LAME to encode a 101MB 16-bit, 44KHz audio file into a very high-quality MP3. The exact command-line options we used were:

lame –alt-preset extreme file.wav file.mp3

MusicMatch Jukebox

The Pentium M is a monster in LAME, but not so much in MusicMatch Jukebox. Still, a respectable showing overall.

Video encoding and editing

XMPEG DivX video encoding
We used the default settings for the DivX codec to encode a 3000-frame sequence from a DVD-formatted MPEG2 source file.

Windows Media Encoder

Adobe Premiere

VideoWave Movie Creator

Video encoding can be very much bound by memory performance, so I expected the Pentium M to struggle here. Surprisingly, it wound up doing well in a couple of the tests, particularly in Premiere.

Image processing

Adobe Photoshop

ACDSee PowerPack

picCOLOR
We thank Dr. Reinert Muller with the FIBUS Institute for pointing us toward his picCOLOR benchmark. This image analysis and processing tool is partially multithreaded, and it shows us the results of a number of simple image manipulation calculations. We’re using a new build of picCOLOR this time out; it removes the video tests, which are highly dependent on the chipset and video card, from the calculation of the overall score.

Clock for clock, the Pentium M beats absolutely everything in Photoshop, and it performs respectably in the other image processing apps.

Multitasking and office applications

MS Office

WorldBench’s MS Office test runs multiple applications simultaneously and switches between them, just as most users tend to do. As a result, the Pentium 4 with Hyper-Threading really shines here, and the Hyper-Threading-deprived Pentium M does not.

Mozilla

For web browsing, the Pentium M is great. So is everything else, though, despite these numbers

Mozilla and Windows Media Encoder

Combine web browsing with video encoding, and the Pentium M remains competitive, even without Hyper-Threading.

Other applications

Sphinx speech recognition
Ricky Houghton first brought us the Sphinx benchmark through his association with speech recognition efforts at Carnegie Mellon University. Sphinx is a high-quality speech recognition routine. We use two different versions, built with two different compilers, in an attempt to ensure we’re getting the best possible performance.

Sphinx is extremely sensitive to memory subsystem performance, and so the Pentium M stumbles here.

WinZip

The Pentium M comes back into contention in WinZip compression.

Nero

The Nero test is all about the disk controller, and the 6300ESB south bridge’s controller is comparable to our Athlon 64 system’s, while the Pentium 4’s ICH6 I/O controller is quite a bit quicker.

Cinebench 2003
Cinebench is based on Maxon’s Cinema 4D modeling, rendering, and animation app. This revision of Cinebench measures performance in a number of ways, including 3D rendering, software shading, and OpenGL shading with and without hardware acceleration. Cinema 4D’s renderer is multithreaded, so it takes advantage of Hyper-Threading, as you can see in the results.

The Cinema 4D renderer doesn’t take well to the Pentium M. Then again, the Pentium M is just a little slower clock for clock than the Athlon 64 here.

The 2.4GHz Pentium M remains very competitive in Cinebench’s shading tests.

3ds max
We have used 3ds max in the past for CPU testing, but most of those tests have consisted of rendering only. WorldBench’s 3ds max tests replicate an entire modeling and animation work session, stressing the graphics card as well as the CPU and the rest of the system.

The combo of a 2.4GHz clock speed and 533MHz does wonders for the Pentium M in this mock 3D modeling session. The 2GHz chip on the 400MHz bus is quite a bit slower, but still not too far off pace.

POV-Ray
POV-Ray is the granddaddy of PC ray-tracing renderers, and it’s not multithreaded in the least, because it’s designed to be a cross-platform application. POV-Ray also relies heavily on x87 FPU instructions to do its work, and it contains only minor SSE optimizations.

This FPU torture test shows that the Pentium M’s x87 FPU is quite a bit more potent than the Pentium 4’s. At matching clock speeds, the Athlon 64’s very decent FPU is just a little bit faster.

Power consumption
There are some caveats about the voltage settings of the various processors in our power consumption test that I’ve explained in detail before. They still apply here. I should add an additional caveat: none of the processors are using a clock-throttling mechanism like SpeedStep or Cool’n’Quiet in the tests below. Unfortunately, the DFI motherboard that we tested doesn’t support SpeedStep. That means the Pentium M could be quite a bit more miserly at idle than you see here, as could the Athlon 64 with Cool’n’Quiet.

We measured the power consumption of our entire test systems, except for the monitor, at the wall outlet using a Watts Up PRO watt meter. The test rigs were all equipped with OCZ PowerStream 470W power supply units. The idle results were measured at the Windows desktop, and we used Cinebench’s rendering test to load up the CPUs. For P4s, we used the multithreaded version of the test to take advantage of Hyper-Threading.

Wow. True to its reputation, the Pentium M practically sips power, even when overclocked. The Pentium 4 pulls well over twice the juice under load. Even the 90nm Athlon 64 at 2GHz isn’t close.

Conclusions
DFI’s nifty little motherboard enables the Pentium M to come to the desktop, and for that reason, it’s a very compelling product. I would like to see some more overclocking options and support for SpeedStep, and I’d really love to get my hands on a desktop Pentium M board based on the new Alviso chipset. But I’m not gonna nitpick too much, because getting a Pentium M to the desktop on these terms is way better than not getting there at all. DFI is among the first to bust the market segmentation barrier, and they deserve mad props for it.

The performance that this motherboard enables speaks for itself, I’d say. As a desktop processor, the Pentium M fares very well. The stock Pentium M 755 at 2GHz rivals the lower speed grades of the Athlon 64 and Pentium 4 fairly consistently. Overclocked to 2.4GHz on a 533MHz bus, though, the Pentium M gets downright scary, shadowing the performance of the Athlon 64 4000+ through many of our tests, including games. The Pentium 4, even in its most extreme editions, often can’t keep up. Games and similar apps with lots of “branchy” code—and perhaps quite a bit of x87 floating-point math—apparently befuddle the Pentium 4. The Pentium M, on the other hand, slices through them with ease.

This processor’s low power requirements and lesser heat production aren’t insignificant, either. Not when it performs like this. Quiet computing is here to stay, small form factor systems are still growing in popularity, and home theater PCs are all the rage. Meanwhile, building a high-end enthusiast PC has become something more of a chore than in the past because of the need to attend very carefully to power and cooling requirements. The numbers and types of power connectors in a typical new system are growing, and the new BTX boxes weigh more than an English Mastiff. The 855GME-MGF’s single, simple ATX power connector and tiny CPU cooler are rare, welcome relief from such trends.

Unfortunately, that relief doesn’t come cheap. The 855GME-MGF is currently selling for about $239 at online merchants, which is an awful lot to ask for a microATX mobo with a modest set of mid-range components on it. Even worse, our Pentium M 755 cost us $435 when we bought it last month, and prices don’t seem to have budged since then. By comparison, the Athlon 64 3500+ costs roughly $279 right now, and the Pentium 4 is running about the same. The Pentium M 755 only becomes a good deal if it will overclock to 2.4GHz and challenge the higher end desktop CPUs in performance, but overclocking is never a sure thing.

For the right application, though, a system based on this mobo and a Pentium M processor would certainly be appealing. Let’s hope DFI enjoys enough success with this board that they put a new version, with PCI Express and dual-channel memory, on the fast track. I’d like to see a DFI LANParty board for the Pentium M with a full suite of overclocking options. And bring on the Pentium M small form factor boxes. Market segmentation be damned! The mobile desktop revolution has begun.