EUROCOM X3（蓝天P157SM）笔记本测评

Compared to our pre-production K56-3N2 unit, the final Eurocom model sports the same high quality construction and rubberized surfaces for a more luxury feel than the Clevo P150EM. Other minute upgrades include sharper-looking edges and corners and more stylized vent grilles that remind us of the slick Asus ROG notebooks. Certain areas that do not have the new rubberized surface treatment, such as the inner display bezel, side edges and belly, feel rigid nonetheless with little to negligible warping under applied pressure. It’s still a boxy chassis compared to the heavy stylized Maingear or Alienware notebooks, but that is to be expected from a barebones unit made for easy end-user maintenance and troubleshooting.

One minor complaint is the new rectangular speaker grilles that look rather lackluster in design. Both grilles are also protected by slits of glossy covering, which are thin enough to bend when pressed. Compared to sleeker speaker designs as found on many MSI-based barebones, the speakers here appear much more unremarkable. Fortunately, the wobbly display that we experienced on the K56-3N2 is much less severe on the Eurocom equivalent. The small hinges are very stiff with no noticeable screen vibrations whilst typing on a solid surface.

Keyboard

The differences between keyboards on Clevo barebones are minor as most are beveled with raised rectangular keys and have a similar layout and overall feel. We especially like that the left- and right-hand edges of the backlit keyboard are close to the sides of the notebook for additional space. The directional keys, for example, are thankfully the same size as the QWERTY keys, but the Enter key could be a bit too small for some. Travel also feels shallow for a DTR, though this is made up for by the solid yet quiet feedback of the keys.

Touchpad

Whereas the keyboard has remained relatively unchanged compared to the Racer 2.0, the lightly rubberized touchpad (10 cm x 6 cm) and click keys have been slightly redesigned with a less boxy look. The touchpad itself is smooth and responsive with no warping if pressed firmly. The click keys are similar, but with a shallower travel than what one may expect from a thick notebook. Also noteworthy is the auditory click, which is different in tone between the left and right keys. Feedback is satisfactory all around, though we would have liked a deeper travel for the keys in general. 

Color space reproduction nearly covers the AdobeRGB standard at about 95 percent. This is very similar to the display of the Eurocom Racer 2.0, rivals that of the similar ThinkPad W530 and is much wider than the gaming-centrc MSI GT70H, which makes the Eurocom a better choice for professional digital artists.

Further display analyses were performed with the CALMAN 5 software. When out-of-the-box and uncalibrated, the Eurocom was already able to show good color and gamma levels, though the overly cool color temperature threw off a bit of the display’s accuracy. Results post calibration with the X-Rite i1Display 2 lowered the color temperature to a much more reasonable 6580K for both a more accurate grayscale and saturated color reproduction.

Performance from the i7-4930MX CPU is similar to that of the desktop Ivy Bridge i7-3770K in CineBench benchmarks. Compared to mobile processors, the CPU is similar to the i7-3940XM, the fastest quad-core Ivy Bridge processor currently available. This is to be expected as both cores share the same Turbo Boost clock rates and L1 through L3 cache sizes. Despite the architectural improvements, however, the Haswell core in the X3 shows only small benchmark performance benefits when compared directly to the i7-3940XM. Multi-thread wPrime results in particular are slower for the i7-4930MX than the i7-3940MX (240 seconds vs. 219 seconds) while Super Pi and CineBench R10 results are only slightly better.

The X3 is capable of housing up to five internal drives: 2x 7mm or 9.5 mm SATA III drives, 2x mSATA SATA III drives and 1x ODD SATA II drive with RAID 0/1/5/10 support. This alone is impressive for a 15.6-inch barebones where up to two or three simultaneous drives are much more common. Due to space constraints, the second mSATA bay is hidden underneath the first HDD bay while the second HDD bay is hidden underneath the ODD SATA II drive. Note that an optical drive occupies the ODD bay by default and must be swapped out in order to insert a hard drive.

Our review model is equipped with two drives: a primary mSATA 120 GB Crucial CT120M500SSD3 and a secondary 7200 RPM 1 TB Hitachi Travelstar HGST HTS721010A9E630. Both drives return very fast benchmark numbers; the HDD in particular averages 106.7 MB/s transfer rates on HD Tune, which is one of the fastest 9.5 mm 1 TB drives we have tested.

Similarly, the Crucial SSD is near the top of the list in terms of sequential read speed alongside the high-end Lite-On LMT-128M3M and ADATA XM11. Sequential write speeds (137.5 MB/s) could be a bit better as many Samsung and Intel 320/520 Series of SSDs return slightly higher numbers in Crystal Disk Mark, though real-world performance will be subjectively similar. The ability to easily enable RAID 1 between the already superfast SSDs or HDDs will surely appeal to power users.

For more benchmarks and comparisons between HDD and SSD models, see our growing list here.

We utilize Prime95 and FurMark to stress the CPU and discrete GPU, respectively, with CPU-Z, GPU-Z and HWiNFO monitoring tools. With only Prime95 active, the CPU cores were observed to be operating at a steady 3.3 to 3.5 GHz range with no drops or sudden throttles. Note that this range is below the rated 3.7 GHz maximum Turbo Boost for all active cores, but still around 300 to 500 MHz faster than its base 3.0 GHz clock rate. Core temperatures also rose very quickly to over 85 degrees C within the first 10 seconds accompanied by an increasingly louder system fan.

With only FurMark active, the GTX 780M was observed to be operating at its 770.8 MHz core clock, but slowly dropped to a more stable 692.4 MHz. Memory clock remained steady at 1250 MHz throughout while GPU temperature hovered close to 80 degrees C.

With both Prime95 and FurMark active simultaneously for over an hour, the CPU clocked a bit lower at a steady 3.2 to 3.4 GHz while the GPU was able to operate at a steady 692.4 MHz throughout the entire duration. CPU and GPU temperatures reached a very hot but expected 94 and 87 degrees C maximum, respectively, by the end of the test. Neither processing cores were observed to throttle, though users should keep in mind that the high 94 degrees temperature is quite close to the 100 degrees thermal operating limit of the Haswell processor. A quick 3DMark 11 run immediately following the stress test returned no reductions in both final and CPU scores, which confirm no throttling issues for even intense everyday workloads.

If running on batteries, both CPU and GPU performances drop significantly, even when running on the High Performance profile. The CPU operates only up to 1.8 GHz for all cores while the GPU similarly maxes out at 405/400 MHz. A 3DMark 06 run on batteries returned much lower numbers with a final score of 14027 points and 3780 points for the CPU. These scores are essentially halved compared to an available AC supply, which returned 27795 points and 7119 points in the same test.

Like the Racer 2.0 and K56-3N2, the Eurocom X3 comes with a dense 8-cell 5200 mAh 76.96 Whr battery and no other available capacities. Performing the same Reader’s Test and Classic Test as on the K56-3N2 returned similar runtimes as expected, if not slightly lower overall. When under general WLAN use with the occasional videos at 150 cd/m2 brightness, users can expect a solid 3 hours of runtime before needing to shutdown. Enabling the dGPU can potentially knock battery life down by almost 2 hours if under heavy processing loads.

Average runtime on the X3 compares favorably against the Pitcairn-based Racer 2.0 and is more or less comparable to the similarly specced Clevo P150SM. The Nomad 15 and its high-end IVB and GTX 680M components managed to last about an hour longer under the same WLAN test. Generally speaking, we have yet to record significantly longer battery life advantages from mobile Haswell and high-end GTX 700M systems, so users should still expect similar runtimes compared to barebones with previous generation hardware.