Small form-factor PCs have become a major growth segment in the PC market over the last decade. In particular, UCFF (ultra-compact form-factor) PCs have become a welcome and permanent fixture in the desktop PC market, all the while they've also seen a good bit of success in the embedded and industrial market segments.

Further segmenting the UCFF market is the level of performance desired, and by proxy the CPU that gets used. Intel's two CPU architectures, Core and Atom, serve to split the market into premium and entry-level devices. And, even with the relatively lower performance of Atom-based SoCs, their aggressive prices make them an attractive proposition for economical desktop PCs as well as industrial motherboards and systems. Atom-based SoCs are long-life products, with Gemini Lake being the most recent SoC family in that product line. Today, we're taking a look at two contrasting Gemini Lake UCFF PCs - the fanless ECS LIVA Z2 and the actively-cooled Intel NUC7PJYH.

Introduction

Intel's Apollo Lake SoCs introduced in 2016 were the first to use the Goldmont CPU microarchitecture. The Gemini Lake SoCs (introduced late last year) are an evolutionary upgrade, bringing in double the amount of on-die cache and providing better performance despite running at approximately the same frequency as their Apollo Lake counterparts. The integrated GPU is also slightly more powerful - both in terms of EUs as well as multimedia capabilities. Prior to the 14nm supply constraints issue, multiple vendors had introduced Gemini Lake-based systems in the market. Similar to our Apollo Lake experiments (reviewing an actively-cooled Arches Canyon NUC and a passively-cooled ECS LIVA ZN33), we got hold of a couple of Gemini Lake UCFF PCs for evaluation - the Intel June Canyon NUC (NUC7PJYH) and the ECS LIVA Z2.

A comparison of the Arches Canyon NUC against June Canyon, and the ECS LIVA Z2 against the ECS LIVA Z, shows the following updates:

• Usage of DDR4 SO-DIMM slots compared to the DDR3 ones in the Apollo Lake systems
• Standardization of at lease one HDMI 2.0 display output
• Replacement of the Apollo Lake SoC with a Gemini Lake one

June Canyon also makes use of a more advanced WLAN solution (AC 9462 vs. AC 3168 in Arches Canyon) that takes advantage of the integrated wireless MAC in the Gemini Lake SoC. However, the ECS LIVA Z2 still uses the older AC 3165. The form factor of the LIVA Z2 is quite different from the LIVA Z - It has a smaller footprint, but is thicker, and doesn't have the dual LAN capabilities of the older version.

The June Canyon NUC comes in multiple flavors, with our review sample being the highest-end configuration. Similarly, the LIVA Z2 comes with either the Celeron N4100 or the Pentium Silver N5000. Both versions come with Windows 10 Home pre-installed on an eMMC card. The two UCFF PCs come with a 65W (19V @ 3.42A) power adapter and a VESA mount.

Both machines integrate a dual-array microphone. This allows the end user to configure it as an always-listening machine (if needed), without the need to connect an external microphone. The other selling point is the availability of a HDMI 2.0 port with HDCP 2.2 support. 4Kp60 capability is present, allowing for specific digital signage use-cases. It also lends itself to usage as a HTPC capable of driving a 4K display.

Platform Analysis

The Gemini Lake SoCs support up to 6 PCIe 2.0 lanes, 8 USB 3.0 ports, and 2 SATA 3.0 ports. The distribution of the PCIe lanes in the two PCs is as below:

• June Canyon NUC7PJYH
  • PCI-E 2.0 x1 port #3 In Use @ x1 (Realtek RTS5229 PCI-E Card Reader)
  • PCI-E 2.0 x1 port #5 In Use @ x1 (Realtek RTL8168/8111 PCI-E Gigabit Ethernet Adapter)
• ECS LIVA Z2
  • PCI-E 2.0 x1 port #4 In Use @ x1 (Realtek RTL8168/8111 PCI-E Gigabit Ethernet Adapter)
  • PCI-E 2.0 x1 port #5 In Use @ x1 (Intel Dual Band Wireless-AC 3165 AC HMC WiFi Adapter)

Note that the usage of the integrated AC MAC in the NUC allows Intel to utilize one of the PCIe ports for a high-performance card reader.

In the table below, we have an overview of the various systems that are being considered today. 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 a particular system when we come to those sections.

Comparative PC Configurations
Aspect	Intel NUC7PJYH	Intel NUC7PJYHECS LIVA Z2Intel NUC6CAYH
CPU	Intel Pentium Silver J5005	Intel Pentium Silver J5005
GPU	Intel UHD Graphics 605	Intel UHD Graphics 605
RAM	Kingston HyperX KHX2400C14S4 DDR4 SODIMM 16-14-14-35 @ 2400 MHz 2x16 GB	Kingston HyperX KHX2400C14S4 DDR4 SODIMM 16-14-14-35 @ 2400 MHz 2x16 GB
Storage	Crucial BX300 CT480BX300SSD1 (480 GB; 2.5" SATA III; Micron 3D MLC)	Crucial BX300 CT480BX300SSD1 (480 GB; 2.5" SATA III; Micron 3D MLC)
Wi-Fi	Intel Dual Band Wireless-AC 9462 (1x1 802.11ac - 433 Mbps)	Intel Dual Band Wireless-AC 9462 (1x1 802.11ac - 433 Mbps)
Price (in USD, when built)	$170 (barebones) $518 (as configured, No OS)	$170 (barebones) $518 (as configured, No OS)

BAPCo SYSmark 2018

The Gemini Lake UCFF PCs were evaluated using our Fall 2018 test suite for small-form factor PCs. In the first section, we will be looking at SYSmark 2018.

BAPCo's SYSmark 2018 is an application-based benchmark that uses real-world applications to replay usage patterns of business users in the areas of productivity, creativity, and responsiveness. The 'Productivity Scenario' covers office-centric activities including word processing, spreadsheet usage, financial analysis, software development, application installation, file compression, and e-mail management. The 'Creativity Scenario' represents media-centric activities such as digital photo processing, AI and ML for face recognition in photos and videos for the purpose of content creation, etc. The 'Responsiveness Scenario' evaluates the ability of the system to react in a quick manner to user inputs in areas such as application and file launches, web browsing, and multi-tasking.

Scores are meant to be compared against a reference desktop (the SYSmark 2018 calibration system, a Dell Optiplex 5050 tower with a Core i3-7100 and 4GB of DDR4-2133 memory to go with a 128GB M.2 SATA III SSD). The calibration system scores 1000 in each of the scenarios. A score of, say, 2000, would imply that the system under test is twice as fast as the reference system.

SYSmark 2018 also adds energy measurement to the mix. A high score in the SYSmark benchmarks might be nice to have, but, potential customers also need to determine the balance between power consumption and the efficiency of the system. For example, in the average office scenario, it might not be worth purchasing a noisy and power-hungry PC just because it ends up with a 2000 score in the SYSmark 2014 SE benchmarks. In order to provide a balanced perspective, SYSmark 2018 also allows vendors and decision makers to track the energy consumption during each workload. In the graphs below, we find the total energy consumed by the PC under test for a single iteration of each SYSmark 2018 workload. For reference, the calibration system consumes 5.36 Wh for productivity, 7.71 Wh for creativity, 5.61 Wh for responsiveness, and 18.68 Wh overall.

Despite being passively cooled, the ECS LIVA Z2 manages to put up a credible fight against the actively-cooled Arches Canyon NUC. Other than that, the scores are along expected lines, with the June Canyon NUC coming out on top (due to its use of the most powerful Gemini Lake SoC available).

UL Benchmarks - PCMark and 3DMark

This section deals with a selection of the UL Futuremark benchmarks - PCMark 10, PCMark 8, and 3DMark. While the first two evaluate the system as a whole, 3DMark focuses on the graphics capabilities.

PCMark 10

UL's PCMark 10 evaluates computing systems for various usage scenarios (generic / essential tasks such as web browsing and starting up applications, productivity tasks such as editing spreadsheets and documents, gaming, and digital content creation). We benchmarked select PCs with the PCMark 10 Extended profile and recorded the scores for various scenarios. These scores are heavily influenced by the CPU and GPU in the system, though the RAM and storage device also play a part. The power plan was set to Balanced for all the PCs while processing the PCMark 10 benchmark.

Other than the gaming scenario, the passively-cooled ECS LIVA Z2 turns out to be better than the Arches Canyon NUC. The June Canyon NUC comes out on by a huge margin.

PCMark 8

We continue to present PCMark 8 benchmark results (as those have more comparison points) while our PCMark 10 scores database for systems grows in size. 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. Despite its passive cooling, the ECS LIVA Z2 is a step ahead of the Arches Canyon NUC in all the workloads.

3DMark

UL's 3DMark comes with a diverse set of graphics workloads that target different Direct3D feature levels. Correspondingly, the rendering resolutions are also different. We use 3DMark 2.4.4264 to get an idea of the graphics capabilities of the system. In this section, we take a look at the performance of the Gemini Lake UCFF PCs across the different 3DMark workloads.

3DMark Ice Storm

This workload has three levels of varying complexity - the vanilla Ice Storm, Ice Storm Unlimited, and Ice Storm Extreme. It is a cross-platform benchmark (which means that the scores can be compared across different tablets and smartphones as well). All three use DirectX 11 (feature level 9) / OpenGL ES 2.0. While the Extreme renders at 1920 x 1080, the other two render at 1280 x 720. The graphs below present the various Ice Storm worloads' numbers for different systems that we have evaluated.

UL 3DMark - Ice Storm Workloads

3DMark Cloud Gate

The Cloud Gate workload is meant for notebooks and typical home PCs, and uses DirectX 11 (feature level 10) to render frames at 1280 x 720. The graph below presents the overall score for the workload across all the systems that are being compared.

3DMark Sky Diver

The Sky Diver workload is meant for gaming notebooks and mid-range PCs, and uses DirectX 11 (feature level 11) to render frames at 1920 x 1080. The graph below presents the overall score for the workload across all the systems that are being compared.

3DMark Fire Strike Extreme

The Fire Strike benchmark has three workloads. The base version is meant for high-performance gaming PCs. Similar to Sky Diver, it uses DirectX 11 (feature level 11) to render frames at 1920 x 1080. The Ultra version targets 4K gaming system, and renders at 3840 x 2160. However, we only deal with the Extreme version in our benchmarking - It renders at 2560 x 1440, and targets multi-GPU systems and overclocked PCs. The graph below presents the overall score for the Fire Strike Extreme benchmark across all the systems that are being compared.

3DMark Time Spy

The Time Spy workload has two levels with different complexities. Both use DirectX 12 (feature level 11). However, the plain version targets high-performance gaming PCs with a 2560 x 1440 render resolution, while the Extreme version renders at 3840 x 2160 resolution. The graphs below present both numbers for all the systems that are being compared in this review.

UL 3DMark - Time Spy Workloads

3DMark Night Raid

The Night Raid workload is a DirectX 12 benchmark test. It is less demanding than Time Spy, and is optimized for integrated graphics. The graph below presents the overall score in this workload for different system configurations.

In terms of GPU performance, being actively cooled helps the Arches Canyon NUC come in second place for almost all relevant cases

Miscellaneous Performance Metrics

This section looks at some of the other commonly used benchmarks representative of the performance of specific real-world applications.

3D Rendering - CINEBENCH R15

We use CINEBENCH R15 for 3D rendering evaluation. The program provides three benchmark modes - OpenGL, single threaded and multi-threaded. Evaluation of different PC configurations in all three modes provided us the following results.

The ECS LIVA Z2 suffers a bit in the OpenGL case, but, is neck and neck with the Arches Canyon NUC in the multi-threaded, and far ahead in the single-threaded case.

x265 Benchmark

Next up, we have some video encoding benchmarks using x265 v2.8. The appropriate encoder executable is chosen based on the supported CPU features. In the first case, we encode 600 1080p YUV 4:2:0 frames into a 1080p30 HEVC Main-profile compatible video stream at 1 Mbps and record the average number of frames encoded per second.

Our second test case is 1200 4K YUV 4:2:0 frames getting encoded into a 4Kp60 HEVC Main10-profile video stream at 35 Mbps. The encoding FPS is recorded. Unfortunately, both Gemini Lake systems were unable to complete this benchmark, exiting abruptly in the middle. The Arches Canyon encoded the 4K stream at 0.36 fps.

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 rates when utilizing all the available threads for the LZMA algorithm.

In addition to the CPU, the RAM also plays a role here. ECS's decision to equip only one of the two SODIMM slots with RAM could be the culprit here.

Cryptography Benchmarks

Cryptography has become an indispensable part of our interaction with computing systems. Almost all modern systems have some sort of hardware-acceleration for making cryptographic operations faster and more power efficient. In this sub-section, we look at two different real-world applications that may make use of this acceleration.

BitLocker is a Windows features that encrypts entire disk volumes. While drives that offer encryption capabilities are dealt with using that feature, most legacy systems and external drives have to use the host system implementation. Windows has no direct benchmark for BitLocker. However, we cooked up a BitLocker operation sequence to determine the adeptness of the system at handling BitLocker operations. We start off with a 2.5GB RAM drive in which a 2GB VHD (virtual hard disk) is created. This VHD is then mounted, and BitLocker is enabled on the volume. Once the BitLocker encryption process gets done, BitLocker is disabled. This triggers a decryption process. The times taken to complete the encryption and decryption are recorded. This process is repeated 25 times, and the average of the last 20 iterations is graphed below.

Since all the considered CPUs have AES-NI capabilities, the above benchmark is representative of the sustainable clock speed and also the RAM characteristics.

Creation of secure archives is best done through the use of AES-256 as the encryption method while password protecting ZIP files. We re-use the benchmark mode of 7-Zip to determine the AES256-CBC encryption and decryption rates using pure software as well as AES-NI. Note that the 7-Zip benchmark uses a 48KB buffer for this purpose.

This shows that the relative numbers are similar to the ones observed in other benchmarks.

Yet another cryptography application is secure network communication. OpenSSL can take advantage of the acceleration provided by the host system to make operations faster. It also has a benchmark mode that can use varying buffer sizes. We recorded the processing rate for a 8KB buffer using the hardware-accelerated AES256-CBC-HAC-SHA1 feature.

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). We present the results from our evaluation using the CPU mode only. The benchmark (v1.3) takes 84 photographs and does four stages of computation:

• Stage 1: Align Photographs (capable of OpenCL acceleration)
• 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, and others multithreaded, it is interesting to record the effects of CPU generations, speeds, number of cores, and DRAM parameters using this software.

Stage 3 is likely memory performance-limited, unlike the first two stages that are reflective of the single-threaded performance capabilities.

Dolphin Emulator

Wrapping up our application benchmark numbers is the new Dolphin Emulator (v5) benchmark mode results.

This is again a test of the CPU capabilities, and the ordering expected based on the previous results is seen here too.

Networking and Storage Performance

Networking and storage are two major aspects which influence our experience with any computing system. This section presents results from our evaluation of these aspects in the Intel NUC7PJYH. 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.

There is really not much to choose from between the three systems - our choice of storage drives (Crucial BX300) was the same for both the Gemini Lake units, and that is reflected in the above graphs.

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 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.

Despite all the WLAN modules being 1x1, the June Canyon NUC has much better numbers. In part, this is due to the availability of antenna diversity in the AC 9462 (which allows the better of two antennae to be utilized for communication).

Concluding Remarks

Earlier I briefly touched upon the HTPC capabilities of the two Gemini Lake PCs. Of the two, the ECS LIVA Z2 is particularly attractive due to its passively cooled nature. Gemini Lake has one of the most power-efficient and full-featured video decoders in the market - It is capable of even handling VP9 Profile 2 in hardware. The only disappointing aspect from a HTPC viewpoint is that HDR is not supported. Consumers looking for a full-featured HTPC setup have to look beyond Gemini Lake for their needs. That said, if HDR is not a concern (as is likely in most TVs installed in a secondary location), Gemini Lake in general and the ECS LIVA Z2 in particular are good choices to consider.

Moving on to the power consumption aspect, we recorded the maximum sustained at-wall numbers while running the AIDA64 System Stability Test.

These numbers show that Gemini Lake, despite running at the same clocks as Apollo Lake, has managed to provide better performance per watt. The raw numbers are approximately the same when two systems with similar cooling situations are compared.

The Intel June Canyon NUC7PJYH Build Components

Coming to the business end of the review, Gemini Lake has shown itself to be a solid successor to Apollo Lake. Doubling the internal cache has led to significant performance increase in many real-life workloads. This is particularly evident when comparing the performance numbers of the NUC7PJYH (June Canyon) and the older NUC6CAYH (Arches Canyon). In addition to the performance bump, we also have better Wi-Fi capabilities. The product is not perfect, though - the installation of a 2.5" drive is not as easy as it is in some of the other UCFF PCs such as the ECS LIVA Z2. With the rise in popularity of M.2 SSDs, Intel could as well replace the 2.5" drive caddy with a M.2 SATA SSD-only slot on the motherboard, and reduce the system volume. The default fan curves could always use some tweaking, but, that is more of a subjective issue. These are very minor issues.

Accessing the Drive Caddy and the SODIMM Slots in the ECS LIVA Z2

On the other hand, the move from the ECS LIVA Z to the Z2 is a mixed bag. While the all-important fanless nature is still present, the retaining of aspects such as the measly 32GB of eMMC (with Windows 10 Home pre-installed) do not make for a good user experience. 32GB of eMMC may suffice for certain single-purpose industrial application systems, but, as a general-purpose PC, consumers will need to install an additional 2.5" SATA drive (similar to what we had to do for processing this review). Some consumers may appreciate the replacement of the mini-DP port in the LIVA Z with a full-sized HDMI one in the LIVA Z2. However, the sacrifice of a second LAN port for this is not a particularly good trade-off. That said, the performance increase enabled by Gemini Lake still makes the ECS LIVA Z2 a good choice over the ECS LIVA Z for most applications.