ASRock’s AMD Ryzen NUC (4x4BOX-V1000M) Review

AMD Ryzen

As a long time Intel user I’m rather familiar with the Intel’s CPU hierarchy with the Celerons, Pentiums and Core CPUs. AMD has a rather large lineup of CPUs as well and Ryzen is just one of the products. And to top of it Ryzen is actually a marketing term that, in rough terms, corresponds to Intel’s Core CPUs. Coincidentally AMD has chosen to call their Ryzen CPU sub-families Ryzen 3, Ryzen 5 and Ryzen 7 in increasing order of performance. How thoughtful.

The Ryzen CPUs (and APUs if you want to get technical) are based on Zen microarchitecture. Zen was the first generation, Zen+ the second and Zen 2 is the latest microarchitecture. The Ryzen Embedded chips like the ones used in the ASRock 4X4BOX are also based on the first generation Zen microarchitecture and they were released in early 2018. Which leaves me wondering whether we could see a consumer-oriented AMD NUC-like PC with Zen 2 mobile CPU (such as Ryzen 5 3500U or Ryzen 7 3700U) anytime soon…

The Ryzen Embedded V1605B powered ASRock 4X4 BOX-V1000M costs $399 at NewEgg and the lower end 4X4 BOX-R1000V is about $100 cheaper.

Performance

Cinebench R15

The Cinebench R15 is quick benchmark utility based on Cinema 4 Suite. The multi-core test is done using all CPU cores and can benefit greatly of hyperthreading. It also has a simple OpenGL benchmark.

In the multi-core CPU benchmark the 4X4 BOX is trailing the i5 Bean Canyon NUC with a score of 671 and in the single-core benchmark the result of 141 is somewhat behind Bean Canyon as well.

In the OpenGL benchmark the 4X4 BOX trails previous NUCs by a significant margin and achieves a score of 53.69.

3DMark

In the 3DMark benchmark that tests gaming performance in various DirectX and CPU tests the tables are turned. The Ryzen V1605B powers the 4X4 BOX to a better result than any Bean Canyon NUC in both Sky Diver and Firestrike benchmarks. Only the Crimson Canyon NUC8i3CYSM with its Radeon RX 540 GPU manages to beat it in these tests. Then again in the Cloud Gate the 4X4 BOX beats the Crimson Canyon handily but the i7 Bean Canyon edges it.

Geekbench 3

Geekbench is another popular test for CPU performance. The ASRock V1000M is basically on par with the i5 Bean Canyon NUC here. Though even if the final scores are similar there are differencies in the individual tests. Study the detailed results and comparison at Geekbench site.

Geekbench 4 results were 3915 for single-core and 11603 for multi-core test.

Gaming

The Radeon GPU should suit casual gaming quite well. I tried my old and tested Dirt 3 again on this NUC as I do have material to compare this NUC to. In addition I ran the benchmark provided in the Shadow of the Tomb Raider which is a much more recent game.

Dirt 3

Dirt 3 is an older racing game but cranking up the detail level and the resolution still makes it possible to provide a proper challenge for the CPUs and GPUs found in NUC-sized PCs. It’s also nice to use a proper game for benchmarking instead of the synthetic benchmarks as that gives a true picture of the system performance in a certain application.

With my usual setup (1920×1080, high details, 60 Hz refresh rate, no vertical sync, 2xMSAA) Dirt 3 runs very well. The 4X4 BOX beats all Bean Canyon NUCs by a clear margin but trails the Crimson Canyon NUC again.

Shadow of the Tomb Raider

To get an idea how a more modern and demanding game would run on this mini PC I installed Shadow of the Tomb Raider trial from Steam. The trial version is convenient because it offers a benchmark that can be used to assess the performance of your PC with the game. Firing up the benchmark it became quickly apparent that there simply isn’t enough oomph to run this with the default settings (1920×1080, detail level high, TAA anti-aliasing). The average frame rate was 13. As this is the first system I test with Shadow of the Tomb Raider I don’t have anything to compare this to, but I don’t expect that any of the Bean Canyon NUCs would do better here.

In order to make the game playable I had to lower the resolution significantly (I tried 1024×600) and drop the detail level. At that point the game didn’t look that impressive anymore…

Power Consumption and Fan Noise

Time to hook up my trusty old power meter. I do use a consumer-grade simple EM240 power meter made by Brennenstuhl. There’s no guarantee that any of the readings are 100% accurate, but at least the figures should be comparable between units I’ve reviewed since the same device was used.

Task	Power
Stand-by	4.0 W
Idle, Windows 10 desktop	13.8 W
Watching a 4K video in Kodi, Windows 10	26.3 W
CPU-Z stress test	46.8 W
Shadow of the Tomb Raider gaming	49.5 W
Watching a 4K video, LibreELEC 9.2	29.8 W
Idle, LibreELEC 9.2 main menu	18.2 W

The power consumption figures are rather similar to the figures I measured for Bean Canyon i3 and i5 NUCs with the exception of idle and standby numbers that are significantly higher for the 4X4 BOX. Based on some research I’ve done the higher idle power consumption seems to be a feature of the first generation of Ryzen CPUs. The third generation Ryzen 3000 mobile CPUs apparently have improved significantly the idle power consumption even if still not matching Intel’s mobile CPUs.

What about the fan noise? Well, there’s quite some to be honest. Seeing the power consumption figures that match Bean Canyon NUC levels it’s not really a surprise considering that the fan is significantly smaller. I could compare the fan noise to a typical lower end non-metal laptop. My problem with the fan is that it kicks in very suddenly. While idling the fan is basically quiet and then suddenly it might boost up to full speed making significant noise. Of course the 4X4 BOX is an industrial model and intended for environments where things like fan noise might be less of a concern.

Keep reading

Move to the next page and read about HTPC capabilities, Linux support and Conclusions.