Introduction and Setup Impressions

GIGABYTE's BRIX Pro (using an Intel Iris Pro part) has made a big splash in the market, particularly as a Steam machine. Enthused by its success, GIGABYTE has introduced a set of mini-PCs under the BRIX Gaming moniker. The lineup consists of mini-PCs in a form factor similar to the BRIX Pro (which itself had the footprint of an Intel NUC kit). The current flagship in the BRIX Gaming line is the BXi5G-760, a machine featuring a Core i5-4200H Haswell CPU and a NVIDIA GTX 760 discrete GPU. Given the paper specifications of the GTX 760, the machine promises to be a gaming powerhouse in its size class.

Similar to the BRIX Pro, the BXi5G-760 comes barebones. Users have to bring in their own DDR3L SODIMM sticks as well as an mSATA or 2.5" drive for completing the build. We configured the review unit to end up with the following components.

GIGABYTE BXi5G-760 Specifications
Processor	Intel Haswell Core i5-4200H (2C/4T x 2.80 GHz (3.40 GHz Turbo), 22nm, 3MB L2, 47W)
Memory	2 x 4GB DDR3L-1866
Graphics	NVIDIA GTX 760 (NVIDIA GTX 870M) 6 GB GDDR5 135 MHz / 941 MHz (Turbo)
Disk Drive(s)	ADATA SX300 128 GB mSATA SSD + Spare 2.5" Drive Slot
Networking	1x Gigabit Ethernet, 1x1 802.11ac mPCIe
Audio	Capable of 5.1/7.1 digital output with HD audio bitstreaming (mini-HDMI / mini-DP 1.2)
Operating System	Retail unit is barebones, but we installed Windows 8.1 Pro x64
Pricing (As configured)	$800 (barebones) + $195 (DRAM + mSATA SSD)
Full Specifications	GB-BXi5G-760 Specifications

The BXi5G-760 kit doesn't come with any pre-installed OS, but we do have a USB key with Windows drivers. In addition to the main unit, the other components of the package include a 180 W (19.5V @ 9.23A) adapter, a US power cord, a splitter for the single audio jack, screws for installing a 2.5" drive, a mini-DP to DP cable and a mini-HDMI to HDMI cable.

The stand-out aspect of the BXi5G-760 compared to mini-PCs in a similar form factor include the presence of three video outputs (2x mini-HDMI and 1x mini-DP). The gallery below takes us around the hardware in the unit. In particular, the dimensions of the unit are compared to the BRIX Pro. The width turns out to be slightly more in order to accommodate the side fans for cooling the GPU.

We configured our unit with an ADATA SX300 128 GB mSATA SSD as a boot drive and put in two Corsair Vengeance 1866 MHz SODIMMs. All BIOS settings were left at default. The DRAM configured itself to run at 1866 MHz without any user inputs

In the table below, we have an overview of the various systems that we are comparing the BXi5G-760 against. Note that they may not belong to the same market segment. The relevant configuration details of the machines are provided so that readers have an understanding of why some benchmark numbers are skewed for or against the BXi5G-760 when we come to those sections.

Comparative PC Configurations
Aspect	GIGABYTE GB-BXi5G-760	Habey BIS-6922GIGABYTE GB-BXi7-4770R (2)Zotac ID89 PlusASRock Vision HT 321BASRock Vision 3D 252BGIGABYTE GB-BXi7-4770R (1)Zotac ZBOX EI750 CustomZotac ZBOX EI750 PlusASRock VisionX 420DIntel D54250WYKH
CPU	Intel Core i5-4200H	Intel Core i7-3720QM
GPU	NVIDIA GTX 760	Intel HD Graphics 4000
RAM	Corsair Vengeance CMSX8GX3M2B1866C10 10-10-10-32 @ 1866 MHz 2x4 GB	Super Talent W1333SB4GH 9-9-9-24 @ 1333 MHz 2x 4GB
Storage	ADATA XPG SX300 AX300S3-128GM-C (128 GB, PCIe Module mSATA 6Gb/s, 25nm, MLC)	Intel® SSD 330 Series (60 GB, SATA 6Gb/s, 25nm, MLC)
Wi-Fi	Realtek 8821AE Wireless LAN 802.11ac (1x1 802.11ac - 433 Mbps)	N/A
Price (in USD, when built)	$995	$1300

 

Performance Metrics - I

The BXi5G-760 was evaluated using our standard test suite for low power desktops / industrial PCs. We revamped our benchmark suite earlier this year after the publication of the Intel D54250WYK NUC review) We reran some of the new benchmarks on the older PCs also, but some of them couldn't be run on loaner samples. Therefore, the list of PCs in each graph might not be the same.

Futuremark PCMark 8

PCMark 8 provides various usage scenarios (home, creative and work) and offers ways to benchmark both baseline (CPU-only) as well as OpenCL accelerated (CPU + GPU) performance. We  benchmarked select PCs for the OpenCL accelerated performance in all three usage scenarios. These scores are heavily influenced by the CPU in the system. The i5-4200H is not as powerful as, say, the i7-4770R in the BRIX Pro or the ZBOX EI750.

Miscellaneous Futuremark Benchmarks

The 3D Mark 2013 results below may appear anomalous, but they are probably well served by the eDRAM in the Iris Pro parts. As we will see later, things get back into the expected mode when benchmarking actual games

3D Rendering - CINEBENCH R15

We have moved on from R11.5 to R15 for 3D rendering evaluation. CINEBENCH R15 provides three benchmark modes - OpenGL, single threaded and multi-threaded. Evaluation of select PCs in all three modes provided us the following results.

The surprising aspect here is that the single-threaded performance of the Core i5-4200H in the BXi5G-760 seems to lag the Core i5-4200M in the VisionX 420D.

In the OpenGL mode, the GPU kicks in and comfortably puts the BXi5G-760 ahead of the pack

Performance Metrics - II

In this section, we mainly look at benchmark modes in programs used on a day-to-day basis, i.e, application performance and not synthetic workloads.

x264 Benchmark

First off, we have some video encoding benchmarks courtesy of x264 HD Benchmark v5.0. This is simply a test of CPU performance. As expected, the Core i5-4200H can't match up to the i7 parts in the other mini PCs that we compare against.

7-Zip

7-Zip is a very effective and efficient compression program, often beating out OpenCL accelerated commercial programs in benchmarks even while using just the CPU power. 7-Zip has a benchmarking program that provides tons of details regarding the underlying CPU's efficiency. In this subsection, we are interested in the compression and decompression MIPS ratings when utilizing all the available threads.

TrueCrypt

As businesses (and even home consumers) become more security conscious, the importance of encryption can't be overstated. CPUs supporting the AES-NI instruction for accelerating the encryption and decryption processes have, till now, been the higher end SKUs. However, with Bay Trail, even the lowly Atom series has gained support for AES-NI. The Core i5-4200H in the BXi5G-760 does have AES-NI support. TrueCrypt, a popular open-source disk encryption program can take advantage of the AES-NI capabilities. The TrueCrypt internal benchmark provides some interesting cryptography-related numbers to ponder. In the graph below, we can get an idea of how fast a TrueCrypt volume would behave in the GIGABYTE BXi5G-760 and how it would compare with other select PCs. This is a purely CPU feature / clock speed based test.

Agisoft Photoscan

Agisoft PhotoScan is a commercial program that converts 2D images into 3D point maps, meshes and textures. The program designers sent us a command line version in order to evaluate the efficiency of various systems that go under our review scanner. The command line version has two benchmark modes, one using the CPU and the other using both the CPU and GPU (via OpenCL). The benchmark takes around 50 photographs and does four stages of computation:

• Stage 1: Align Photographs
• Stage 2: Build Point Cloud (capable of OpenCL acceleration)
• Stage 3: Build Mesh
• Stage 4: Build Textures

We record the time taken for each stage. Since various elements of the software are single threaded, others multithreaded, and some use GPUs, it is interesting to record the effects of CPU generations, speeds, number of cores, DRAM parameters and the GPU using this software.

Dolphin Emulator

Wrapping up our application benchmark numbers is the Dolphin Emulator benchmark mode results. This is again a test of the CPU capabilities, and the Core i5-4200H comes up short against the Core i7-4770R

Gaming Benchmarks

The gaming credentials of the GTX 760 are quite impressive, but it is simply impossible for such a GPU to be put in a mini-PC with the BRIX form factor. A detailed analysis of the various features of the GTX 760 reveal that the GPU is nothing but a rebranded GTX 870M. In order to sell itself in the desktop market, the mobile-targeted GK104 part has been rechristened as the GTX 760 targeting small form factor machines. The GTX 870M is pretty good, finding a place in high-end gaming laptops such as the Razer Blade. In this section, we will see how it performs when provided with a little bit of additional thermal headroom.

For the purpose of benchmarking, we chose five different games (Company of Heroes 2, Sleeping Dogs, Tomb Raider, Bioshock Infinite and DiRT Showdown) at three different quality levels. As someone focusing on HTPCs and multimedia aspects, I rarely get to process gaming benchmarks, even while evaluating GPUs. One of the aspects that I feared was spending lot of time in installing the same games again and again on different PCs under the review scanner. The solution was to go the Steam route. Unfortunately, Steam also likes to keep the game files updated. A quick online search revealed that Steam could make use of an external drive for storing the game executables and downloadable content. With the Steam drive on-the-go use-case being read-heavy, the Corsair Flash Voyager GS USB 3.0 128GB Flash Drive (with read speeds of up to 275 MBps) was ideal for use as a portable Steam drive.

Benchmark Numbers

Company of Heroes 2

Sleeping Dogs

Tomb Raider

Bioshock Infinite

DiRT Showdown

It goes without saying that the tussle is between the AMD Radeon equipped VisionX 420D and the BXi5G-760. The BXi5G-760 handily wins most benchmarks, with the Sleeping Dogs and DiRT Showdown games showing a much closer fight. In the big scheme of things, the BXi5G-760 manages to emerge in pole position when talking about the gaming capabilities.

Networking & Storage Performance

We have recently started devoting a separate section to analyze the storage and networking credentials of the units under review. On the storage side, one option would be repetition of our strenuous SSD review tests on the drive(s) in the PC. Fortunately, to avoid that overkill, PCMark 8 has a storage bench where certain common workloads such as loading games and document processing are replayed on the target drive. Results are presented in two forms, one being a benchmark number and the other, a bandwidth figure. We ran the PCMark 8 storage bench on selected PCs and the results are presented below.

The presence of a mSATA SSD greatly increases the usability of the unit if the workloads have major dependence on storage performance. Obviously, these scores depend on the drives used by the end user to complete the system configuration.

On the networking side, we restricted ourselves to the evaluation of the WLAN component. Our standard test router is the Netgear R7000 Nighthawk configured with both 2.4 GHz and 5 GHz networks. The router is placed approximately 20 ft. away, separated by a drywall (as in a typical US building). A wired client (Zotac ID89-Plus) is connected to the R7000 and serves as one endpoint for iPerf evaluation. The PC under test is made to connect to either the 5 GHz (preferred) or 2.4 GHz SSID and iPerf tests are conducted for both TCP and UDP transfers. It is ensured that the PC under test is the only wireless client for the Netgear R7000. We evaluate total throughput for up to 32 simultaneous TCP connections using iPerf and present the highest number in the graph below.

In the UDP case, we try to transfer data at the highest rate possible for which we get less than 1% packet loss.

The Wi-Fi module in both the BRIX Pro and the BXi5G-760 are the same, but the latter shows better benchmark numbers. This is possibly due to the difference in the arrangement of the antenna, as well as difference in the drivers that were used for evaluation. The major drawback for the BXi5G-760 is the presence of a 1x1 802.11ac card instead of a 2x2 that some of the other mini-PCs employ. This means that the WLAN rates are not as good as those of the 2x2-equipped VisionX 420D.

BXi5G-760 as a HTPC

Thanks to its discrete GPU, the BXi5G-760 has to tackle a much higher thermal workload with its cooling mechanism. To accommodate this, a few extra fans have been thrown in compared to the BRIX Pro. It goes without saying that the acoustic profile is not entirely suitable for most HTPC applications (unless the gaming aspect gets precedence). Despite this, the HTPC credentials deserve coverage. We didn't take the trouble to look into the refresh rates. NVIDIA has one of the best custom resolution / refresh rate creation tools in the industry, and even minor deviations from the expected refresh rates in one's particular setup can be worked upon for more accuracy. We concentrate on two aspects - network streaming efficiency and decoding / rendering benchmarks.

Network Streaming Efficiency

Evaluation of OTT playback efficiency was done by playing back our standard YouTube test stream and five minutes from our standard Netflix test title. Using HTML5, the YouTube stream plays back a 720p encoding, while Adobe Flash delivers a 1080p stream. The power consumption at the wall as well as the GPU usage while playing them on Mozilla Firefox are provided in the table below. Note that NVIDIA exposes GPU and VPU loads separately. This indicates whether any playback issues are due to the post-processing aspects (GPU) or the video decoding block's inability to handle the stream (VPU). Netflix streaming evaluation was done using the Windows 8.1 Netflix app. Manual stream selection is available (Ctrl-Alt-Shift-S) and debug information / statistics can also be viewed (Ctrl-Alt-Shift-D). Statistics collected for the YouTube streaming experiment were also collected here.

GIGABYTE BXi5G-760 Network Streaming Efficiency
OTT Stream	GPU Load (%)	VPU Load (%)	Power (W)
YouTube - HTML5	15.54	10.43	31.68 W
YouTube - Adobe Flash	12.71	26.26	30.25 W
Netflix - Windows 8.1 App	5.66	27.18	30.18 W

Decoding and Rendering Benchmarks

In order to evaluate local file playback, we concentrate on EVR-CP and madVR. We already know that EVR works quite well even with the Intel IGP for our test streams. Under madVR, we used the default settings for one of the passes. In addition, we also set up a 'madVR stress configuration' with the upscaling algorithms set to Jinc 3-tap with anti-ringing activated and the downscaling algorithm set to Lanczos 3-tap, again with anti-ringing activated. The decoder used is from the LAV filters integrated in MPC-HC v1.7.6.

An important point to note regarding the GPU loading is that the values reported by NVIDIA's drivers don't take the clock speed into consideration. For example, the 20% GPU loading reported for 1080i60 H.264 with EVR-CP is at a higher clock rate compared to the 45% reported for the 1080i60 MPEG-2 stream. The power consumption at the wall is the true metric of how much the system is stressed. The loading factors should be used only to determine if the stutters in playback are due to the hardware's incapability.

GIGABYTE BXi5G-760 - Decoding & Rendering Performance
Stream	EVR-CP			madVR - Default			madVR - Stress
	GPU Load (%)	VPU Load (%)	Power (W)	GPU Load (%)	VPU Load (%)	Power (W)	GPU Load (%)	VPU Load (%)	Power (W)
480i60 MPEG2	49.19	4.08	34.30	11.07	2.00	56.45	27.88	2.00	72.54
576i50 H264	46.44	10.18	33.14	11.17	5.00	57.61	29.61	5.00	77.39
720p60 H264	42.64	38.94	34.98	34.37	27.58	52.29	53.99	25.03	66.87
1080i60 MPEG2	44.93	26.29	48.41	24.05	17.72	63.12	37.56	17.10	75.90
1080i60 H264	20.69	29.53	54.99	24.99	27.14	65.36	38.58	27.12	77.97
1080i60 VC1	75.49	54.14	44.11	24.38	23.86	63.75	38.1	24.46	77.70
1080p60 H264	49.66	84.95	37.12	38.05	64.17	49.37	54.25	65.34	59.92
1080p24 H264	20.22	34.77	32.99	40.3	44.18	36.83	54.72	37.36	39.01
4Kp30 H264	22.73	74.73	45.02	57.33	57.95	76.37	92.48	57.75	96.85

The results show that the BXi5G-760 is the most powerful madVR-capable HTPC we have reviewed in its form factor. The only stream that it couldn't handle was the 4Kp30 H.264 clip when played on a 1080p display with the stressful madVR options.

In the gaming benchmarks, we found it tough to make a call between the AMD R9 270X-equipped VisionX 420D and the GTX 760-equipped BXi5G-760. In the HTPC benchmarks, though, the GIGABYTE model emerges a clear winner. However, if acoustics are a concern, the ASRock VisionX 420D may make a better choice.

Miscellaneous Aspects & Final Words

The power consumption at the wall was measured with the display being driven through one of the mini-HDMI ports. In the graphs below, we compare the idle and load power of the BXi5G-760 with other low power PCs evaluated before. For load power consumption, we ran Furmark 1.12.0 and Prime95 v27.9 together. The BXi5G-760 is not the most power efficient PC around, but the target market (gamers) don't need to care too much. The numbers are not beyond the realm of reason for the combination of hardware components in the machine.

Even though the steady state power consumption in the simultaneous CPU and GPU loading case was around 105 W, we did see an initial spike into the 150 W range before stabilization.

Thermal Performance

Given the size of the chassis, throttling under heavy load was always a possibility. Our fears turned out to be true, as the two graphs below show. The various clocks in the system as well as the temperatures with the unit are presented below. For the first 15 minutes, we have the unit at idle. This was followed by 30 minutes of pure CPU loading, followed by another 30 minutes of both CPU and GPU being loaded simultaneously. We found throttling in effect at that point. After this, the CPU load was removed, allowing the GPU to be loaded alone for another 30 minutes.

In the graphs on either side, we find that the idle clocks are around 800 MHz for the CPU cores and around 135 MHz for the GPU, with the temperatures between 40 and 50 C inside the unit. With the plain CPU loading for 30 minutes, we find there is no throttling and the CPU temperatures stabilising around 94 C with the core clocks around 2.5 GHz. Once the GPU kicks into gear, we can see the red dips in the temperature graph below. Since we are graphing DTS here, a value of 0 means that the system needs to throttle to bring temperatures back in the operating range. The core clocks go down ro around 2.1 - 2.3 GHz. Note that the GPU clocks were only very briefly at the advertised 941 MHz. Eventually, under sustained loading, they end up oscillating between 202.5 MHz and 405 MHz. When the Prime 95 loading is taken off, the CPU has more headroom and ramps up to 3.4 GHz frequently to help keep the GPU loaded. An advantage with the active thermal solution is that the temperatures return to idling points very quickly after removal of the load.

The safe temperature limit for the GTX 760 has been set at 97 C, and we found that the solution in the BX i5G-760 levelled out around 93 C under these loading conditions. The GPU does fine-grained management of the clock frequency (which is interpreted by Furmark as throttling) to manage the thermals.

Concluding Remarks

GIGABYTE has managed to cram impressive gaming components into a small chassis. If there were a metric for gaming performance per cubic-foot, the BXi5G-760 would win it hands down. The NVIDIA GTX 760 / GTX 870M is also a top notch choice for gaming in this form factor. Its HTPC credentials are also excellent. The pre-integration of the WLAN component (Realtek-based 1x1 802.11ac) makes things very simple for the end user to complete the build. The triple-display output is a welcome feature for gamers.

On the pricing side, the BXi5G-760 comes in at $800 without the DRAM or a boot drive. That is steep for a machine with limited expansion possibilities. GIGABYTE has also tied up with 'Unified Remote' to provide control of the PC using mobile devices. It would have been nice on GIGABYTE's part to supply a few codes for the full version of the Unified Remote mobile app. That said, this is a nice alternative to the MCE remotes that other vendors are supplying with their mini-PCs.

In conclusion, if one is looking for a highly portable non-DIY gaming mini-PC, it is hard to go wrong with the GIGABYTE BXi5G-760.