Results
Memory performance is pretty good - in line with other Intel machines we've tested, and the raw SATA performance matches that of other boards too.
Everest Memory Read (MB/s) | 7510 |
Everest Memory Write (MB/s) | 6060 |
Everest Memory Latency (ns) | 64.4 |
SATA Hard Drive Performance (HDTach - MB/s) | 64.3 |
If you were buying this system today we've likely kitted it out with a setup that many users would consider, with the latest drivers. In this respect, if we were comparing to our previous P35 and X38 results,
here, we see that the SP35P2 has a performance issue with small file use in Photoshop Elements and small file compression. This is a significant gap, not just several seconds that could be attributed to driver differences. The large file performance is very good though, with both Xvid encoding and large file compression performing as well as or better than other solutions but not enough to outright determine the difference between Shuttle and component improvements.
After several days of use as a system the SP35P2 still felt nippy and even though the results indicate that it can be a bit laboursome with certain file sizes it didn't feel that way so much.
|
Photoshop Elements (Seconds) | 370.6 |
|
Large Compression MT (Seconds) | 124.5 |
Large Decompression MT (Seconds) | 15.7 |
Small Compression MT (Seconds) | 173.6 |
Small Decompression MT (Seconds) | 35.6 |
|
Audio Compression (Intel) (Seconds) | 88 |
Audio Decompression (Intel) (Seconds) | 28 |
Audio Compression (MS) (Seconds) | 96 |
Audio Decompression (MS) (Seconds) | 29 |
|
VirtualDub Xvid Rendering (Seconds) | 726.3 |
Armed with a 22" LCD that uses a popular resolution of 1680x1050, we tried to find the best rough settings for a few popular games. Even on Ultra High detail in
Enemy Territory: Quake Wars, we maxed out resolution and the system could still churn out plenty of FPS. Turning up the antialiasing and anisotropic filtering will add to the quality, but also put more weight on the performance of the graphics card rather than whole system.
The
World in Conflict demo is typically more intensive, but even with a maxed out resolution you'll find somewhere between High and Very High detail should be playable.
Crysis at 1280x800 with High-to-Very High in-game settings is surprisingly playable, although 1680x1050 does tax it quite a bit more. Personally I preferred the lower resolution with higher detail settings – even though it isn't as crisp, it still looks far better. After several hours of "system evaluation and stress testing" whilst playing
Crysis (it's a hard life, I know), I found that 1280x800 with a 50-50 split between High and Very High options (and only post processing on medium because I can't stand the overzealous motion blur) the game looked fantastic and was still completely playable.
|
Enemy Territory Quake Wars 1.2 (1680x1050 0AA 0AF Ultra High Detail) | 79.1 |
|
World in Conflict 1.005 (1680x1050 0AA 0AF High Detail Minimum FPS) | 20 |
World in Conflict 1.005 (1680x1050 0AA 0AF High Detail Average FPS) | 35 |
|
World in Conflict 1.005 (1680x1050 0AA 0AF Ultra High Detail Minimum FPS) | 13 |
World in Conflict 1.005 (1680x1050 0AA 0AF Ultra High Detail Average FPS) | 24 |
|
Crysis (1280x1024 0AA 0AF High Detail Average FPS) | 47.52 |
Crysis (1680x1050 0AA 0AF High Detail Average FPS) | 26.36 |
|
Conclusions and Value
While it's obviously a very capable gaming system, it does get quite noisy when everything is hot and loaded. With headphones on or your speakers turned up this isn't a huge problem; but with it sitting on the desk next to you, it becomes far more obvious than a case sitting under your desk.
You could build an alternative system like an
Abit IP35 Pro, an
Antec P182 and a
Corsair HX520 power supply for about £250, give or take a few pounds. Personal choice will vary the components, but in comparison the Shuttle SP35P2 is
£293 – that's £50 more, but the space saving and attention to detail in the design, as well as the integration of the SP35P2’s parts is second to none.
On the other hand, many people keep their cases between upgrades and sometimes even their PSUs too – so as a whole purchase, the Shuttle XPC concept gets less inviting. Unless, of course, you're specifically looking for a small form factor with desktop level performance and don't want to be constrained by the inability of a notebook's expansion – then the SP35P2 should be ideal.
Most of the people I know have had at least one Shuttle XPC at some point. We've all tried it, but it's an ideal that all of us grew out of and went back to ATX. My current case is a massive Antec P190, but does the SP35P2 make me pine for that small world after all? No, but it does a great job in trying – it looks fantastic, its build quality is superb and it's got an unmatched performance-to-size ratio.
Even though Shuttle has tried to maximise the SP35P2s flexibility as well, I feel this SFF is stuck in a no man's land between a completely portable notebook and a vastly more extensive and easily upgradeable ATX PC. If you're a LAN gaming nut, then the SP35P2 is a must buy... unless you have a car. Or perhaps the missus just wants one because she thinks they're cute? There's dozens of reasons why it's a good idea, but also a dozen reasons why it’s not – the SP35P2 Pro still hasn't made the XPC a definitive buy.
Final Thoughts
I still see the XPCs as excellent home theatre PCs and Shuttle has other models perfectly suited to that task – this isn’t one of them. If you're really, really wanting something small and powerful then definitely get the SP35P2 Pro without a shadow of a doubt, other SFFs just don't cut it. However it won't make the rest of us want one (again), because it's just far too noisy when loaded – most of you will happily stick with ATX.
- Build Quality
- x
- x
- x
- x
- x
- x
- x
- x
- x
- -
- 9/10
- Performance
- x
- x
- x
- x
- x
- x
- x
- -
- -
- -
- 7/10
What do these scores mean?
Want to comment? Please log in.