Samsung conveniently provide the key hardware specifications on the box label. My NP900X4C contains the following:
- Processor: Intel i5-3317U CPU (1.7GHz with TurboBoost to 2.6GHz, 3MB cache)
- Chipset: Intel HM75
- Screen: 15.0 inch anti-glare 1600 x 900 WXGA+ LED Backlit
- Memory: 8GB DDR3-1600 1.5V dual channel RAM (2 x 4GB Samsung, 11-11-11-28)
- Graphics: Intel HD4000
- Storage: 128GB mSATA SATA 3 SSD (Sandisk U100)
- Optical Drive: None
- Wireless: WiDi compatible Intel 6235 802.11abgn (dual band + Bluetooth 4.0)
- UK 87 key island type backlit keyboard
- Elan touchpad (117mm x 76mm or 4.6″ x 3.0″)
- 1.3MP web camera (at display top) and microphone (on left side)
- 1 x USB 2.0 port; 2 x USB 3.0 ports; Samsung mini VGA port (needs dongle); micro HDMI port (maximum 1080p resolution); Samsung mini gigabit ethernet port (dongle supplied); combo audio port; SD card slot (SDHC / SDXC compatible)
- Battery: 7.4V 62Wh 8-cell
- 40W “slim” PSU
- Windows 7 Home Premium 64-bit
- Dimensions: 357 x 239 x 15mm (including rubber pads about 1.5mm thick) or 14.1″ x 9.4″ x 0.6″
- Actual weight: 1.65kg / 3.64 lbs
- Travel weight: 1.98kg / 4.37 lbs (with 40W PSU and 1m mains cable)
Versions of this notebook with a 17W i7 CPU and 256GB SSD are available in some markets. As supplied, only about half of the 128GB SSD is available to the user because, in addition to the pre-installed Windows and software, there is an 8GB hibernation partition and a 20GB recovery partition. A user upgrade to a 256GB mSATA SSD is technically feasible now that they are becoming available. In fact I have already done this and the impact is revealed in some of the benchmark results.
Performance and Benchmarks
The Samsung NP900X4C has adequate performance for everyday usage and moderately demanding work. The rapid bootup and the fast resume both enhance the user experience.
wPrime CPU Benchmark
The wPrime benchmark shows that the raw CPU performance lags behind many other computers. This is a result of the relatively slow i5-3317U CPU which, while nominally having a TurboBoost speed of 2.6GHz has a sustained speed of 2.4GHz (according to HWiNFO). There is a modest speed improvement when compared to the i5-2467M CPU which is mainly the result of the higher clock speed. I would prefer the i7 CPU with its extra speed but Samsung don’t currently offer it in the UK.
Samsung make some of the best SSDs. However, they chose to provide my NP900X4C with the Sandisk U100 mSATA SSD which has class-trailing overall performance. It has reasonably good sequential data throughput but, as I had discovered with this same SSD in my Samsung NP900X3B, chokes when given multiple datasets to process concurrently. I refer readers to my review of the Crucial m4 mSATA SSD for comparative performance data for the U100 and the m4. I would note that, to date, Samsung have not used the U100 SSD in the models of the NP900X4C sold with a 256GB SSD.
The PCMark Vantage benchmark measures overall system performance. I am giving two results for this benchmark: With the supplied 128GB Sandisk U100; and with the 256GB Crucial M4 which shifts the overall performance, according to this benchmark, from above average to the best in my house. A detailed examination of the breakdown of the PCMark Vantage results reveals, however, that components which should not, logically, depend on SSD performance are boosted by the change in SSD, so perhaps the extra capacity also influences this benchmark.
PCMark 7 is a newer benchmark to measure overall system performance. The change in SSD from the U100 to the m4 has much less impact on this benchmark score but pulls it ahead of the recently reviewed ThinkPad T430s or Acer Timeline Ultra M3. This benchmark supports the impression that a fast SSD has more impact on overall performance than a slightly faster CPU.
3DMark06 measures gaming graphics performance. While there are the newer and more demanding 3D benchmarks, the older 3DMark06 benchmark is more appropriate for ultrabooks with their less powerful graphics. The NP900X4C shows a significant improvement compared to the NP900X3B. Much of this can be attributed to the Intel Ivy Bridge architecture but having 8GB of 1600MHz RAM on board also helps.
Heat and Noise
The NP900X4C, with only one cooling fan, is unable to match the smaller NP900X3B (with two fans) in the quietness league table when under load. It is also noisier than my Lenovo T420s. Various factors appear to contribute to the fan noise: The first is that the airflow is more complicated. The two air inlets on the base of the computer are about 20% smaller than those on the NP900X3B requiring a more fan effort to suck the air in but it appears that air is also sucked in through one of the vents on the back. In addition, relatively conservative fan rules make the fan go to maximum speed once the CPU temperature exceeds 70°C which can occur even with an intensive single-threaded task. The highest CPU temperature I have observed is 78°C so there is plenty of headroom to raise the thermal settings (the CPU in the T420s can exceed 90°C under sustained load). Under light usage the fan noise is not noticeable and if there is a need to prioritise silence over performance then Fn+F11 will enable quiet mode which limits the CPU to 800MHz. The one benefit of the fan noise is that external temperatures do not get excessively hot in spite of the thin chassis and the temperatures after sustained CPU load are shown below.
The nominal battery capacity is 62Whr (7.4V, 8400mAh) and Samsung claim up to 10 hours running time. Although the notebookreview standard battery rundown test (Windows 7 Balanced power profile , wireless active and refreshing a web page every 60 seconds) normally uses 70% screen brightness I consider this inappropriate for a nominal 400 nit brightness display where the backlight consumes a substantial proportion of the power. After all, I find that 5/8 brightness is optimum for indoor use. A 70% brightness test wouldn’t provide a fair comparison with a notebook such as the Lenovo T430 which can give up to 15 hours under those test conditions but uses a display that is half as bright (200 nit). I therefore ran two tests, at 3/8 brightness and 4/8 brightness and these provided 9 hours 57 minutes and 8 hours 4 minutes respectively. Samsung’s claim of 10 hours is therefore realistic for working off-line or with a less bright display.
I investigated further the power consumption related to the backlight and created this graph which shows that maximum brightness adds about 4.5W to the power consumption which, considering the basic power consumption is around 5W, is nearly 100% increase. Not surprisingly, high brightness and long battery life do not go together. Samung provide a Battery Life Extender option which limits the maximum charge to 80% (ie up to about 8 hours away from a power socket). I hope this will substantially delay the date when a new battery is needed.