Shuttle’s XPC SB81P mini-barebones system

And we’ve complained every step of the way.

We decried the SV24’s limited selection of CPUs. Shuttle then produced a Pentium 4-based XPC. We fussed over the early XPCs’ blurry built-in video, lack of an AGP slot, and weak overclocking options. Shuttle responded with better integrated graphics and an AGP slot. We groused about noise and heat. Shuttle introduced a new generation of XPCs that packed an innovative heat-pipe cooler and speed-controlled fans. We gathered our breath and whinged about the lack of AMD-based XPCs. Shuttle answered by unleashing the nForce2-based SN41G2, an instant classic.

Our reply? We reiterated our desire for more overclocking options. The XPC company listened, producing the overclocking-friendly SN45G and a string of new mini-barebones rigs with performance BIOSes. We thanked them by campaigning for beefier power supplies and less noise. The 250W SilentX PSU was born.

“Great,” we said, “but can you do something about that ugly CD-ROM drive bay?”

That led us, by my recollection, right up to the introduction of the XPC SB75S, a gorgeous pearlescent white number with a stealthed optical drive bay and whisper-silent operation. Naturally, we complained that the eject button for the optical drive wouldn’t initiate a disc insert. We whined about the SB75S’s missing reset button. We lamented the fact we couldn’t run a two-slot GeForce card in the box. And, for good measure, we reminded everyone that Shuttle’s XPCs had audio quality no better than your average motherboard and were not big enough to house a RAID array.

Shuttle’s designers, ever vigilant, apparently weren’t daunted. They’ve taken the opportunity, with Intel introducing a range of new standards and technologies alongside the 915/925X Express chipsets, to redesign the XPC from a clean sheet of paper. The result is Shuttle’s new P-series chassis, the first incarnation of which is the XPC SB81P. This new XPC sits inside of a handsome, black, BTX-inspired case that’s slightly larger in every dimension than prior XPCs. With innovative three-zone cooling, Intel High Definition Audio, room for a RAID array, and a PCI Express X16 slot capable of housing a double-wide graphics card, the SB81P aims to stifle our complaints once and for all. Will it succeed? Keep reading to find out.

As you can see from the picture, the SB81P looks great. The optical, floppy, and front I/O ports are all stealthed, and the case comes in any color you want, so long as it’s black.

The specs
This first P-series XPC is true to its roots in that it includes built-in graphics, so that this mini-barebones box can become a fully functioning computer with only the addition of a CPU, memory, and drives. Also, although it’s based on the newfangled 915G chipset with PCI Express, the SB81P includes a couple of noteworthy nods to practicality. Here are the full specs.

You may have spotted those two nods to practicality. The SB81P accepts DDR400 memory, not DDR2, and it has a standard old PCI slot instead of PCI Express X1. Given the relatively high prices of DDR2 memory and the utter nonexistence of PCI Express X1 expansion cards, those choices make sense.

Beyond that, the SB81P is packed full of nifty new things courtesy of Intel’s 915G chipset. The built-in Graphics Media Accelerator has four pipelines and DirectX 9-class pixel shaders, and the PCI-E X16 offers discrete graphics cards more bandwidth than the FCC. The ICH6R south bridge chip endows the SB81P with four ports of Serial ATA storage, complete with RAID levels 0 and 1, plus eight ports of 24-bit, 192KHz audio.

It’s not in our table there, but the SB81P ships with a darn-near beefy 350W power supply unit, no doubt necessary to power the fastest Intel processors and PCI Express X16 graphics cards.

The one other highlight I’d like to point out up front is the SB81P’s Gigabit Ethernet chip. This Broadcom Ethernet controller is a true PCI Express device, so it should feature higher throughput than most GigE built into motherboards these days.

A closer look
At 320mm x 210mm x 220m, Shuttle’s new P-series chassis is a little bigger in all directions than the familiar G-series. Here they are side by side:

The SB81P is still a very compact box, but it’s a little larger than your average toaster.

Shuttle’s new stealthed drive cover does an excellent job hiding that big, bad, beige DVD drive. Notice that the eject button resides just above the plastic flap covering the drive. That means the button will serve to tell the drive to retract a disc, not just to eject one. This is exactly what we asked for last time out. Perfect.

The optical drive does sit back into the bay a little bit, so you’ve got to tilt the disc a little to remove it from the tray. By now, this behavior is second nature to me, though, because many systems are like this.

The other two covers up front hide the external 3.5″ drive bay and an array of audio, USB, and Firewire ports. That 3.5″ bay may be more likely to host an internal drive than a floppy, since those slots at the very top of the front face are part of the SB81P’s built-in six-in-one flash card reader.

Around back, the SB81P has a healthy selection of ports, including no less than eight different audio ports, including both optical and coaxial SPDIF outputs. There’s a VGA port for the integrated graphics, plus the expected RJ-45 port for Gigabit Ethernet. Also, right between the Firewire and VGA ports is a little, recessed button that will allow folks to clear the CMOS without opening the case. This is a good idea carried over from the Zen XPC, and I used it several times in working with the SB81P. I’d like to see makers of full-size PC enclosures follow Shuttle’s lead here, because the clear CMOS button is a very handy feature.

The SB81P’s belly is vented to allow air intake, and the round indention you can see there sits right underneath the CPU socket, presumably to pull some heat away from beneath the processor.

Under the cover
Pull the cover, and you get a look at the P series’ almost completely tool-free guts. There are custom tool-free drive rails, tool-free fasteners, and clever tool-free latching mechanisms to hold everything together.

At the top of the case are two internal 3.5″ drive bays, complete with neatly pre-routed Serial ATA power and I/O connectors. These drives fit into drive rails that snap securely into place.

Drives living here, in one of the SB81P’s three cooling zones, are cooled by the two upper fans at the rear of the case.

Here’s a shot of the snap-in doohickey that secures the tray for external drives. Actually this fastener is a part of the 5.25″ drive rails. Again, no screws are required.

The drive cage pops out for easy accessibility. The lower 3.5″ bay will house either an external drive, like a floppy, or a 3.5″ internal hard drive, but they mount different. Shuttle supplies rails for the floppy, while hard drives just snap into the bottom of the tray via a tension-based mechanism. This is the only location where an ATA/100 drive can live in the SB81P. The motherboard’s single ATA port will serve both the 3.5″ drive and the optical drive, as master and slave devices.

Taming the blowtorch

The picture above shows the CPU cooling tunnel inside the front of the case. The SB81P deviates from the BTX spec by pulling air side-to-side through the case to cool the CPU. You can see, on the left, the plastic shroud that keeps intake air separate from the rest of the system. In the middle is Shuttle’s new heatpipe cooler, and on the right is the second fan, which pulls air through the fins of the cooler.

The shroud itself is just a snap-in plastic piece.

Once the shroud is removed, the cooler is easier to see.

Like Shuttle’s other heatpipe coolers, this puppy is designed to pull heat up away from the CPU and into fins more easily cooled by an exhaust fan. The water-filled heatpipes run from inside the copper slug just above the CPU, turn 90 degrees, and protrude up among the fins.

Into the guts

The CPU exhaust fan swings back to allow access to the LGA775 processor socket. This is one of the new Intel sockets with the pins on the motherboard, so it’s a little more fragile than your standard socket. You can’t really see if in this picture, but the fan has to swing out in order for the lever that holds the socket closed to swing open completely.

With the drives and CPU cooling tunnel removed, the SB81P is wide open for access. Notice the placement of the power supply right across the middle of the rear of the case. This fan takes care of the SB81P’s third cooling zone, the one for the motherboard, memory, and chipset. Shuttle was confident enough in this zone’s temperatures to stick with fan-free passive cooling on both the north and south bridge chips of the 915G chipset.

The XPC SB81P’s PCI Express X16 slot for graphics cards sits on the outside of the case, as the AGP slot did on the G-series XPCs, but it’s on the opposite side of the case, so that any mega-wide dual-slot graphics coolers will hang out over the adjacent PCI slot.

However, the SB81P doesn’t have the new BTX-style six-pin auxiliary power connector for graphics cards, nor does it have a pair of spare four-pin Molex coming off of different power rails. I was able to run a GeForce 6800GT by connecting a power adapter to a single four-pin Molex connector, and it didn’t give me any problems.

Shuttle has routed the cables inside the SB81P precisely where they need to be and secured them with ties and clips to keep clutter to a minimum. This is a big step forward compared to earlier XPCs, and it adds to the professional look of the SB81P.

Throttling problems
As I was testing the XPC SB81P, some of my benchmark numbers were coming out much, much lower than expected. I’d already confirmed that all the basic system settings were correct (like memory speed, CPU frequency, and the like), so I decided to do some testing to see if CPU thermal throttling might be the culprit. For those of you not familiar, the Pentium 4 includes an internal facility to cool itself down when it gets too hot by reducing the amount of work that it does. Without getting too technical, the CPU doesn’t exactly lower its own clock speed, but the effect is the same. Thermal throttling is a failsafe, and it shouldn’t be encountered in normal use because the CPU should never get so hot that it’s necessary.

I’ve found that thermal throttling is a startlingly common problem on the fastest new Intel processors, especially the Pentium 4 “Prescott” 3.6GHz and the P4 Extreme Edition 3.4GHz. In fact, I struggled mightily with thermal throttling problems in preparing my review of the new LGA775 processors when the CPUs were running on an open test bench with a massive Intel cooler. After a tortured conversation with an Intel thermal engineer (“Is the heatsink seated properly?” “Did you use enough thermal paste?”), I got the impression that Intel knows the Prescott chips, in particular, are running especially close to the edge of the thermal envelope. With some work, I was able largely to banish thermal throttling on my desktop test rigs, but in my experience, this sort of problem is not uncommon on the highest speed Intel processors these days.

In order to test whether throttling was happening in the XPC SB81P, I used the same basic methodology described in this article. The RightMark CPU Stability Test measures computational throughput over time. It can be configured with multiple threads to make full use of Hyper-Threading, so the processor gets a rigorous workout. (Hyper-Threading has been known to increase the Pentium 4’s heat output quite a bit.)

When everything is properly cooled and thermal throttling is not active, computational throughput is steady, as this screenshot shows:

However, in the XPC SB81P, RightMark’s output almost immediately began looking uneven:

Not only was the performance erratic, but it was much lower than normal, as well. Thermal throttling was going to town. The screenshots above were taken with a Pentium 4 Extreme Edition 3.4GHz processor, but I saw the same thing, or worse, with the Prescott 3.6GHz. Both CPUs were experiencing severe thermal problems with the default fan speed settings in the SB81P’s BIOS.

Naturally, I popped into the BIOS to see what I could do to fix it. Shuttle’s fan speed control BIOS menu looks like this:

Note the little table/diagram on the right of the screenshot above. The XPC SB81P uses Intel’s new Prescott fan speed control mechanism, with linear fan speed control rather than a simple multi-stage arrangement. This setup, at least in theory, should provide better cooling and better acoustics, with fewer distracting transitions from one fan speed grade to the next and back.

This arrangement isn’t entirely linear, though. Only a specified range of temperatures triggers a change in fan speed. Below the trigger temperature, the CPU fan should run at some constant, low speed (or, in cases where the cooling design is capable enough, the fan could stop.) Between the trigger temp and the maximum temperature, fan speeds should ramp up in a linear fashion. Here’s a simple example where the trigger temp is 30C and the max temp is 70C.

That’s pretty much as it should be. I believe it’s also acceptable for the system designer to have the CPU fan speed kick into high gear once the max temperature threshold is reached. In that case, the diagram would look like so:

The system would get noisy in a hurry, but only in cases of extreme heat.

However, in the case of the SB81P, Shuttle has set the default trigger temp for its two CPU cooling fans at 70C. The max temp is set—unalterable in the BIOS—at 80C. The SB81P’s fan speeds would look something like this:

Now, if you’re wanting to keep the system as quiet as possible until absolutely necessary, concentrating the dynamic range of fan speeds near the high end of the CPU’s recommended operating temperatures might make some sense. However, in this case, our 3.6GHz Prescott processor starts thermal throttling right at 70C—or the same spot where, by default, the SB81P only begins to ramp up its fan speed from the lowest level. In fact, during our throttling testing with the SB81P’s BIOS defaults, I could never audibly detect a fan speed increase from the SB81P. The CPU would throttle itself back, and thus cool itself off, before the CPU fans would start to provide any extra help.

One option for addressing this problem is to lower the minimum trigger temp in the BIOS so that fan speeds begin ramping up sooner. Turning it to 50C helped, but didn’t solve the problem. With the trigger temp set at 40C, I still saw notable performance dips in the RightMark CPU Stability Test. At 30C, things were better, but the SB81P was running pretty darn loud at that point—beyond what I’d consider acceptable.

You can see the problem with adjusting the lower temperature limit illustrated on the graph below. Say you set the temp tag to 50C. This is what you will get:

The CPU fans start ramping up at lower temperatures, but they still won’t reach their peak cooling potential before the CPU begins throttling. Bottom line: the 80C max temperature threshold is simply incorrect for the Pentium 4 3.6GHz. For comparison, the max temp threshold on the Abit AA8 motherboard is 65C. Monkeying with the trigger temp on the lower end of the spectrum isn’t the appropriate fix.

I took this information to Shuttle, and they were able to confirm the throttling problem with a Prescott 3.6GHz. They proposed to address it in two ways. First, they have promised a BIOS update, likely with a new max temp threshold around 70C. Second, they sent me a beta copy of their Windows-based XPC Tools utility that they’re preparing for eventual public release.

Using this nifty little program, I was able to create my own custom fan speed profiles. I set the trigger temp at 50C, the max temp at 68C, and chose the upper and lower fan speeds associated with each. I set the upper speed in the linear range for both fans well below the maximum possible speed, creating a “broken” line of possible fan speeds similar to the second graph example above. Once I’d properly tuned everything, the XPC SB81P was able to run the CPU Stability Test for long periods of time without throttling and without excessive noise. In the confines of Damage Labs, the system generally kept the fully loaded P4 3.6GHz CPU below 68C without kicking up to maximum fan speeds. Running 3DMark03 on a GeForce 6800GT alongside the CPU Stability Test didn’t tend to invoke max fan speeds, either.

So it seems Shuttle’s P-series chassis can indeed manage to power and cool a blowtorch-like Intel Prescott processor and a massive NVIDIA GeForce 6800GT GPU without sounding like a C-130. It truly is a wonder. The current SB81P BIOS, however, needs to be fixed.

BIOS options and overclocking
Shuttle offers a couple of different BIOS versions for its XPCs. The standard “consumer and OEM BIOS” is the idiot proof version, but you can also flash the SB81P to Shuttle’s “performance BIOS.” This BIOS is most definitely not devoid of tweaking and overclocking options.

Memory timings, bus and memory clock frequencies, and voltages can all be adjusted at will. The memory options are about what one would expect, with no major omissions.

Like the best BIOSes from Abit, the SB81P allows users to choose CPU clock frequencies in 1MHz increments by keying in the desired speed—much easier than scrolling through 255 possible options. Simpler options, like voltage settings, are multiple choice. For CPU voltage, the choices come in 0.0125V increments up to 1.5875V.

One of the most unique options in the SB81P BIOS is the LED brightness setting. This setting controls the intensity of the blue LED that illuminates the SB81P’s front power button. This LED isn’t especially blinding at 100%, but at lower settings, this XPC should make a better always-on citizen in a dorm room or bedroom.

This screen controls memory partitioned off for the 915G’s built-in graphics. Thanks to Intel’s DVMT, or Dynamic Video Memory Technology, it’s possible to have the graphics core allocate only a small amount of RAM for normal 2D desktop displays, and then grab more as needed for 3D graphics.

Unfortunately, for all the SB81P’s tweakability, it’s no overclocking titan. Shuttle apparently hasn’t overcome the dreaded “overclock lock” rumored to be incorporated into the 915G chipset. With the PCI and PCI Express clocks set to 33/100MHz and north bridge voltage set to 1.7V, I was able to get the SB81P to boot into Windows with a 215MHz bus with both a Radeon X600 XT and a GeForce 6800GT. At 220MHz, the system wouldn’t post with either card.

Asus, Abit, and Albatron all have full-sized motherboards that claim to bypass Intel’s clock lock, and I was able to get the Abit AA8 running with a 270MHz bus. There’s lot of overclocking potential in the new Intel chipsets; let’s hope Shuttle can uncork it via a BIOS update.

Test notes
Please note a couple of hardware differences between our test systems in the configs below. The AGP systems use a Radeon 9600 XT card, while the PCI Express ones use an X600 XT, which has a higher memory clock speed. Also, the Abit IC7-G is running with a Pentium 4 3.4E, because the P4 3.6GHz isn’t available for Socket 478. Just something to keep in mind.

Also, we tested the SB81P with its integrated graphics, as well. Those scores are noted in the graphs as “w/IGP”.

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 Intel systems were run with Hyper-Threading enabled, unless otherwise specified.

Thanks to Corsair 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, Corsair’s RAM is definitely worth considering.

The test systems’ Windows desktops were set at 1152×864 in 32-bit color at an 85Hz screen refresh rate, with exception of the SB81P with integrated graphics active, which was at 1024x768x32 at 85Hz. Vertical refresh sync (vsync) was disabled for all tests.

We used the following versions of our test applications:

• Cachemem 2.65MMX
• SiSoft Sandra 2004 (9.89)
• Sphinx 3.3
• LAME 3.96 (build from mitiok.cjb.net)
• Xmpeg 5.0.3 with DivX Video 5.11
• Comanche 4 demo
• Quake III Arena 1.31
• Splinter Cell 1.2
• Unreal Tournament 2004 3236
• Far Cry 1.1
• HD Tach 2.61
• RightMark 3DSound 1.01

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

The SB81G’s memory performance isn’t bad, but it is a little disappointing compared to the other Pentium 4 system with DDR400, the Abit IC7-G based on the Intel 875P chipset. As expected, the SB81P’s integrated graphics core absorbs a little memory bandwidth when it’s active, even in these essentially text-based benchmarks.

Unreal Tournament 2004

Far Cry

Quake III Arena

Comanche 4

Splinter Cell

The XPC’s gaming performance more or less meets the expectations set by its scores in our synthetic memory benchmarks. The XPC nearly keeps pace with the Intel 925X motherboard with DDR2 memory, so it’s no slouch.

However, with the Intel GMA900 graphics enabled, the SB81P takes a huge performance hit. Although Intel claims the GMA900 does DirectX 9 with Pixel Shader 2.0, their drivers still need some work. When we first reviewed the 915G chipset, the GMA900 crashed out of Far Cry. This time around, with newer drivers, the GMA900 was able to run the game, but we saw way too much visual corruption. Also, many of the pixel shader effects simply weren’t being applied properly. The GMA900 may be serviceable for more mundane tasks, but for real gaming, you need a separate graphics card.

Also, I noticed that the GMA900’s video output wasn’t as crisp as I’d have liked, especially at resolutions above 1024×768. The analog VGA signal coming out of the Shuttle box doesn’t appear to be as clean and those produced by GeForce and Radeon cards.

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 that needs the latest computer hardware to run at speeds close to real-time processing. We use two different versions, built with two different compilers, in an attempt to ensure we’re getting the best possible performance.

The Shuttle generally keeps pace with the pack in Sphinx.

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

MP3 encoding seems to be almost entirely CPU dependent. For all intents, the SB81P ties with the other Pentium 4 3.6GHz systems.

DivX video encoding
This new version of XMPEG includes a benchmark feature, so we’re reporting scores in frames per second now.

Would-be HTPC builders, take note. The Prescott P4 processor that drops into the SB81P is very speedy at video encoding tasks, and the Shuttle box accommodates it fairly well.

USB performance
Starting with USB performance, the rest of our tests focus on the I/O capabilities of the motherboard instead of the CPU, memory and graphics. Generally, these tests are not about absolute performance so much as they are about validating basic functionality. We’re looking to see whether the board achieves acceptable transfer rates with reasonable CPU utilization.

We used HD Tach to measure USB transfer rates to a Maxtor DiamondMax D740X hard drive in a USB 2.0 drive enclosure.

Oddly, the SB81P chews up quite a bit more CPU time during USB transfers than the Abit and Intel boards. Performance is a little lower, too. Of course, the other boards have faster memory, but I doubt that accounts for the whole of the CPU utilization difference we’re seeing here.

Firewire/1394 performance
We also used HD Tach to measure IEEE 1394 transfer rates to a Maxtor DiamondMax D740X hard drive in a 1394 drive enclosure.

Shuttle’s little black box handles Firewire about as well as the larger 925X-based motherboards.

ATA performance
ATA/100 performance was tested using HD Tach and a Seagate Barracuda V 120GB hard drive.

Nothing to see here. Move along.

SATA performance
For SATA performance, we used a Maxtor MaxLine III 250GB hard drive with support for the AHCI spec, including Native Command Queuing. However, NCQ’s benefits aren’t likely to show up in a simple test like HDTach. For a better look at NCQ’s benefits, go here.

Serial ATA presents no problem for the Shuttle box, either. It’s right in line with the full-sized mobos.

Ethernet performance
We evaluated Ethernet performance using the NTttcp tool from the Microsoft’s Windows DDK. We used the following command line options on the server machine:

ntttcps -m 4,0,192.168.1.25 -a

..and the same basic thing on each of our test systems acting as clients:

ntttcpr -m 4,0,192.168.1.25 -a

We used our Abit IC7-G-based system as the server for these tests. It has an Intel NIC in it that’s attached to the north bridge via Intel’s CSA connection, and it’s proven very fast in prior testing. The server and client talked to each other over a Cat 6 crossover cable.

Here’s a pleasant surprise. The SB81P’s built-in PCI Express Ethernet absolutely clobbers the PCI Ethernet built into the Abit AA8 and requires less CPU overhead than the GigE chip on the Intel motherboard. Network performance on this thing is stellar.

Audio performance
RightMark3D measures CPU utilization with a number of audio tasks. We can see how well Intel’s High Definition Audio implementation works.

Although the Shuttle box uses the same audio controller and codec as the other two boards, it needs a few more CPU cycles in order to handle the same tasks. The difference is especially apparent in the DirectSound3D hardware test. Still, we’re not talking about vast amounts of CPU overhead.

Subjectively, the SB81P’s Intel High Definition Audio definitely sounds much than prior XPCs. However, listening to this box’s audio with a pair of headphones is still not a sublime experience. I could hear buzzes and interference as I moved windows around on the screen, and there’s a regular clicking sound that happens about once per second. Granted, one might not notice such things in everyday use, especially with speakers. Still, the noise is there. Obviously, getting clean sound out of a small computer with a compact motherboard is no easy task.

Noise levels
I used my trusty digital sound level meter (Extech model 407727) to measure the noise created by the SB81P in operation. I held the meter 1″ from the front and rear of each system in order to measure sound levels. Because of the outstanding problems with the SB81P’s fan control settings in the BIOS, I measured the system using my own custom profile that has more appropriate fan speed settings for the Pentium 4 3.6GHz.

These results more or less track with my subjective impressions. During normal use, the SB81P is quieter than previous XPCs, but when it has to crank up all the fans in order to cool the P4 3.6GHz, it gets louder. (The other cubes are running much slower, much cooler CPUs.) Thanks to its linear fan control and dedicated CPU tunnel, though, the SB81P doesn’t seem especially loud when it’s running under load.

Now, if you want loudness, you can crank all three of this box’s fan to 100%. Then, this puppy produces 64.4 dB up front and 67.8 around back. However, you really shouldn’t see the fans kick up to 100%, even with heavy use.

Conclusions
During the course of this review, I’ve pointed out some substantial shortcomings of the XPC SB81P. The Pentium 4 thermal throttling problem is at the top of the list. Shuttle needs to release its XPC Tools to the public, and more importantly, the company must release a new BIOS that fixes this problem before the SB81P goes on sale in mid-August. I’d also like to see that BIOS update include a workaround for the “overclock lock” in the 915G chipset, because right now, the SB81P simply can’t be overclocked much at all. Beyond the BIOS, this box’s built-in audio and video could be better. The Intel GMA900 isn’t much use for gaming, despite its DirectX credentials, and its video signal output is a little sub-par. And the SB81P’s High Definition Audio may be a step up from previous XPCs, especially in terms of capabilities, but the signal-to-noise ratio won’t tickle an audiophile’s sensitive ears.

But I fully expect the thermal throttling problem to be rectified with a BIOS update, and I wouldn’t be shocked to see the overclocking barrier busted, as well. Shuttle has amassed a heckuva track record in refining its XPC products, and they were very responsive when I pointed out the throttling issues.

So I have my complaints. But you have to understand something about the XPC SB81P. This little box is one of the coolest pieces of gear to make its way into Damage Labs for a long, long time. Shuttle’s new P-series chassis is a massive stride forward for the XPC line; it erases most of my objections to the smaller XPC boxes without compromising Shuttle’s small form factor concept. The fit and finish are characteristicaly excellent, and the new tool-free interior makes installation a breeze. With PCI Express graphics and dual-channel DDR400 memory, performance is comparable with competing full-size ATX systems. And the thing manages to wrap up the Pentium 4 Prescott 3.6GHz processor in an attractive package that doesn’t sound like a wet-dry vac. Also, with apologies to the SB75S, this may be the most aesthetically pleasing XPC yet.

Wrap all those things together, and you get pure geek lust. To see one up close is to want one. Shuttle may be banking on the lust factor in setting the XPC SB81P’s list price at a healthy $399. Given the thing’s merits, though, I can’t really complain. Although Shuttle’s engineers may disagree with me about that.