Verdict — Simple, But Maybe Too Simple
As a mini PC for businesses, the Cubi NUC 2MG is a plain and reliable option with no flashy aesthetics. Its integrated speaker, microphones, Co-Pilot+ support, and Power Link feature can come in handy in certain conditions or usage scenarios while the Lunar Lake-V CPU is generally cooler and more efficient than the current crop of AMD alternatives. We particularly appreciate the fast integrated MicroSD reader as many other mini PCs lack card readers.
Our main gripe about the Cubi 2MG is its relatively high price of $900 USD when faster AMD mini PCs like the Beelink SER9 Pro or GMK NucBox K12 are readily available for hundreds of dollars less. Such alternatives also come with useful features like removable RAM and support for more internal SSDs. MSI says it offers improved reliaibility, better customer support, and a longer three-year warranty as standard to justify the asking price, but performance-per-dollar is nonetheless relatively weak.
Fan behavior could also use some improvement. Unless if you're on Silent mode, the fan tends to pulse and be more noticeable even when running undemanding loads. The cycling GPU clock rates, while not an issue for most professional settings, is still odd to see when most other mini PCs don't exhibit the same problem.
Pros
Cons
The Cubi NUC 2MG succeeds last year's Cubi NUC 1M. It replaces the older Raptor Lake-U Core Ultra 7 150U CPU with newer Lunar Lake-V options while making other minor auxiliary changes. The Cubi series is optimized for businesses and retail environments as opposed to home offices or HTPC applications.
Our test unit is a midrange configuration with the Core Ultra 7 258V CPU, 32 GB of soldered RAM, and 1 TB SSD for approximately $900 USD. Lesser SKUs start with the Core Ultra 5 226V while the highest-end option ships with the Core Ultra 9 288V. A white version is available in some regions.
The model competes directly against the Asus NUC series which also targets environments where 24/7 operation and reliability take priority over raw performance.
More MSI reviews:
Specifications
Case
Chassis and dimensions remain mostly unchanged from the Cubi 1M. New to the Cubi 2MG, however, is the dedicated CoPilot button on the front of the unit. The problem with this feature is that the physical button can be difficult to access depending on how you position the system on your desk.
Power Link is supported for powering on the device via certain MSI monitors for convenience. The list of compatible monitors can be found here. Unfortunately, Power Link still requires an AC adapter for the mini PC and so it is not a true "one cable" solution.
Connectivity
Port options have changed slightly from the Cubi 1M. There is now just one HDMI port instead of two alongside fewer USB 3.0 ports.
Note that while where are two USB-C Thunderbolt 4 ports, only one of the ports supports Power Delivery-in of up to 100 W should the main AC adapter be unavailable.
SD Card Reader
| SD Card Reader | |
| average JPG Copy Test (av. of 3 runs) | |
| MSI Cubi NUC 1M (Angelbird AV Pro V60) | |
| MSI Cubi NUC AI+ 2MG (Angelbird AV Pro V60) | |
| maximum AS SSD Seq Read Test (1GB) | |
| MSI Cubi NUC AI+ 2MG (Angelbird AV Pro V60) | |
| MSI Cubi NUC 1M (Angelbird AV Pro V60) | |
Communication
| Networking | |
| iperf3 transmit AXE11000 | |
| MSI Cubi NUC AI+ 2MG | |
| GMK NucBox K12 | |
| Minisforum MS-A1 | |
| MSI Cubi NUC 1M | |
| Beelink SER9 Pro | |
| iperf3 receive AXE11000 | |
| MSI Cubi NUC AI+ 2MG | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC 1M | |
| iperf3 transmit AXE11000 6GHz | |
| Minisforum MS-A1 | |
| MSI Cubi NUC 1M | |
| MSI Cubi NUC AI+ 2MG | |
| GMK NucBox K12 | |
| iperf3 receive AXE11000 6GHz | |
| Minisforum MS-A1 | |
| MSI Cubi NUC AI+ 2MG | |
| MSI Cubi NUC 1M | |
| GMK NucBox K12 | |
Maintenance
The bottom panel comes off easily with just a Phillips screwdriver for quick servicing. Unfortunately, only one internal PCIe4 x4 M.2 2280 SSD is supported with no removable RAM. In contrast, the NucBox K12 mini PC supports up to three M.2 SSDs and two removable SODIMM RAM modules.
Sustainability
The chassis consists of 37.25 percent post-consumer recycled plastic while the inside packaging consists of FSC-certified cartons made from sustainably sourced wood pulp.
Accessories And Warranty
The retail box includes a removable VESA mounting bracket and a handy 19-inch corded power button for turning on the PC if it is situated in a hard-to-reach area.
A three-year limited manufacturer warranty comes standard compared to one one-year for most other mini PCs in the US.
Performance
Testing Conditions
We set MSI Center to Performance mode prior to running the benchmarks below.
Processor
The Core Ultra 7 258V is almost 50 percent faster than the older Core 7 150U in last year's Cubi NUC 1M when it comes to raw multi-thread performance. It also comes with an integrated NPU whereas the Nuc 1M did not for Co-Pilot+ support.
Upgrading to the Core Ultra 9 288V SKU is not expected to increase performance by any significant margin while downgrading to the Core Ultra 5 226V will only reduce performance by just a few percentage points.
Cinebench R15 Multi Loop
Cinebench R23: Multi Core | Single Core
Cinebench R20: CPU (Multi Core) | CPU (Single Core)
Cinebench R15: CPU Multi 64Bit | CPU Single 64Bit
Blender: v2.79 BMW27 CPU
7-Zip 18.03: 7z b 4 | 7z b 4 -mmt1
Geekbench 6.5: Multi-Core | Single-Core
Geekbench 5.5: Multi-Core | Single-Core
HWBOT x265 Benchmark v2.2: 4k Preset
LibreOffice : 20 Documents To PDF
R Benchmark 2.5: Overall mean
| CPU Performance rating | |
| Minisforum MS-A1 | |
| Asus NUC 14 Pro+ | |
| Beelink SER9 Pro | |
| Intel NUC 13 Pro Kit NUC13ANK -2! | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Cinebench R23 / Multi Core | |
| Asus NUC 14 Pro+ | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (7920 - 11097, n=23) | |
| MSI Cubi NUC 1M | |
| GMK NucBox M5 | |
| Ninkear M7, (16+512) | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Cinebench R23 / Single Core | |
| MSI Cubi NUC AI+ 2MG | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| MSI Cubi NUC 1M | |
| Average Intel Core Ultra 7 258V (1664 - 1957, n=23) | |
| Minisforum MS-A1 | |
| Asus NUC 14 Pro+ | |
| Beelink SER9 Pro | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Cinebench R20 / CPU (Multi Core) | |
| Asus NUC 14 Pro+ | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (3130 - 4314, n=23) | |
| Ninkear M7, (16+512) | |
| MSI Cubi NUC 1M | |
| GMK NucBox M5 | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Cinebench R20 / CPU (Single Core) | |
| MSI Cubi NUC AI+ 2MG | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| MSI Cubi NUC 1M | |
| Average Intel Core Ultra 7 258V (643 - 749, n=23) | |
| Minisforum MS-A1 | |
| Asus NUC 14 Pro+ | |
| Beelink SER9 Pro | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Cinebench R15 / CPU Multi 64Bit | |
| Asus NUC 14 Pro+ | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| Average Intel Core Ultra 7 258V (1280 - 1656, n=23) | |
| MSI Cubi NUC AI+ 2MG | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| MSI Cubi NUC 1M | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Cinebench R15 / CPU Single 64Bit | |
| MSI Cubi NUC AI+ 2MG | |
| Minisforum MS-A1 | |
| Average Intel Core Ultra 7 258V (258 - 298, n=23) | |
| Beelink SER9 Pro | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| Asus NUC 14 Pro+ | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Blender / v2.79 BMW27 CPU | |
| BMAX B4 Turbo | |
| Asus NUC 14 Essential | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| Average Intel Core Ultra 7 258V (315 - 452, n=23) | |
| MSI Cubi NUC AI+ 2MG | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| Beelink SER9 Pro | |
| Minisforum MS-A1 | |
| Asus NUC 14 Pro+ | |
| 7-Zip 18.03 / 7z b 4 | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| Asus NUC 14 Pro+ | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| Ninkear M7, (16+512) | |
| Average Intel Core Ultra 7 258V (28482 - 36724, n=23) | |
| MSI Cubi NUC AI+ 2MG | |
| GMK NucBox M5 | |
| MSI Cubi NUC 1M | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| 7-Zip 18.03 / 7z b 4 -mmt1 | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| MSI Cubi NUC 1M | |
| MSI Cubi NUC AI+ 2MG | |
| Asus NUC 14 Pro+ | |
| Average Intel Core Ultra 7 258V (5265 - 5869, n=23) | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Geekbench 6.5 / Multi-Core | |
| Asus NUC 14 Pro+ | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| Average Intel Core Ultra 7 258V (8964 - 11334, n=23) | |
| MSI Cubi NUC AI+ 2MG | |
| MSI Cubi NUC 1M | |
| GMK NucBox M5 | |
| Ninkear M7, (16+512) | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Geekbench 6.5 / Single-Core | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (2593 - 2795, n=23) | |
| Minisforum MS-A1 | |
| MSI Cubi NUC 1M | |
| Beelink SER9 Pro | |
| Asus NUC 14 Pro+ | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Geekbench 5.5 / Multi-Core | |
| Asus NUC 14 Pro+ | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| Average Intel Core Ultra 7 258V (7574 - 9918, n=23) | |
| MSI Cubi NUC 1M | |
| MSI Cubi NUC AI+ 2MG | |
| GMK NucBox M5 | |
| Ninkear M7, (16+512) | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Geekbench 5.5 / Single-Core | |
| MSI Cubi NUC AI+ 2MG | |
| Minisforum MS-A1 | |
| Average Intel Core Ultra 7 258V (1891 - 2046, n=23) | |
| MSI Cubi NUC 1M | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| Beelink SER9 Pro | |
| Asus NUC 14 Pro+ | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| HWBOT x265 Benchmark v2.2 / 4k Preset | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| Asus NUC 14 Pro+ | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (9.65 - 14.2, n=23) | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| LibreOffice / 20 Documents To PDF | |
| Asus NUC 14 Essential | |
| BMAX B4 Turbo | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| Asus NUC 14 Pro+ | |
| Beelink SER9 Pro | |
| Average Intel Core Ultra 7 258V (41.1 - 59.3, n=23) | |
| Minisforum MS-A1 | |
| MSI Cubi NUC AI+ 2MG | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| MSI Cubi NUC 1M | |
| R Benchmark 2.5 / Overall mean | |
| BMAX B4 Turbo | |
| Asus NUC 14 Essential | |
| Ninkear M7, (16+512) | |
| GMK NucBox M5 | |
| Beelink SER9 Pro | |
| Average Intel Core Ultra 7 258V (0.443 - 0.4998, n=23) | |
| MSI Cubi NUC AI+ 2MG | |
| Asus NUC 14 Pro+ | |
| Intel NUC 13 Pro Kit NUC13ANK | |
| MSI Cubi NUC 1M | |
| Minisforum MS-A1 | |
* ... smaller is better
AIDA64: FP32 Ray-Trace | FPU Julia | CPU SHA3 | CPU Queen | FPU SinJulia | FPU Mandel | CPU AES | CPU ZLib | FP64 Ray-Trace | CPU PhotoWorxx
| Performance rating | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V | |
| Ninkear M7, (16+512) | |
| MSI Cubi NUC 1M | |
| AIDA64 / FP32 Ray-Trace | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (8454 - 12781, n=23) | |
| Ninkear M7, (16+512) | |
| MSI Cubi NUC 1M | |
| AIDA64 / FPU Julia | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (40875 - 63401, n=23) | |
| Ninkear M7, (16+512) | |
| MSI Cubi NUC 1M | |
| AIDA64 / CPU SHA3 | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (1930 - 2817, n=23) | |
| Ninkear M7, (16+512) | |
| MSI Cubi NUC 1M | |
| AIDA64 / CPU Queen | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| Ninkear M7, (16+512) | |
| MSI Cubi NUC 1M | |
| Average Intel Core Ultra 7 258V (48816 - 49818, n=23) | |
| MSI Cubi NUC AI+ 2MG | |
| AIDA64 / FPU SinJulia | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| Ninkear M7, (16+512) | |
| MSI Cubi NUC AI+ 2MG | |
| MSI Cubi NUC 1M | |
| Average Intel Core Ultra 7 258V (3817 - 4440, n=23) | |
| AIDA64 / FPU Mandel | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (21670 - 33565, n=23) | |
| Ninkear M7, (16+512) | |
| MSI Cubi NUC 1M | |
| AIDA64 / CPU AES | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (30957 - 40753, n=23) | |
| Ninkear M7, (16+512) | |
| MSI Cubi NUC 1M | |
| AIDA64 / CPU ZLib | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (422 - 632, n=23) | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| AIDA64 / FP64 Ray-Trace | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (4258 - 6930, n=23) | |
| Ninkear M7, (16+512) | |
| MSI Cubi NUC 1M | |
| AIDA64 / CPU PhotoWorxx | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (44223 - 55458, n=23) | |
| Beelink SER9 Pro | |
| Minisforum MS-A1 | |
| GMK NucBox K12 | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
Stress Test
The CPU would stabilize at 3.2 GHz, 66 C, and about 38 W when running Prime95 stress. When running Cyberpunk 2077 to represent typical gaming stress, however, GPU clock rates would cycle between 800 MHz and 1950 MHz resulting in unstable frame times when gaming as shown by our GPU Performance section below even if the average frame rate is normal. This stuttering behavior would occur in both Balanced mode or Performance mode.
| Average CPU Clock (GHz) | GPU Clock (MHz) | Average CPU Temperature (°C) | |
| System Idle | -- | -- | 33 |
| Prime95 Stress | 3.2 | -- | 66 |
| Prime95 + FurMark Stress | 2.9 | 1200 | 71 |
| Cyberpunk 2077 Stress | 1.3 | 800 - 1950 | 63 |
System Performance
PCMark 10: Score | Essentials | Productivity | Digital Content Creation
CrossMark: Overall | Productivity | Creativity | Responsiveness
WebXPRT 3: Overall
WebXPRT 4: Overall
Mozilla Kraken 1.1: Total
| Performance rating | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V | |
| MSI Cubi NUC 1M | |
| MSI Cubi NUC AI+ 2MG | |
| Ninkear M7, (16+512) | |
| PCMark 10 / Score | |
| Minisforum MS-A1 | |
| GMK NucBox K12 | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V (6265 - 7314, n=22) | |
| Beelink SER9 Pro | |
| MSI Cubi NUC AI+ 2MG | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| PCMark 10 / Essentials | |
| Minisforum MS-A1 | |
| MSI Cubi NUC 1M | |
| GMK NucBox K12 | |
| Beelink SER9 Pro | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V (9295 - 11467, n=22) | |
| MSI Cubi NUC AI+ 2MG | |
| Ninkear M7, (16+512) | |
| PCMark 10 / Productivity | |
| Minisforum MS-A1 | |
| GMK NucBox K12 | |
| Beelink SER9 Pro | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V (7357 - 9619, n=22) | |
| MSI Cubi NUC AI+ 2MG | |
| Ninkear M7, (16+512) | |
| MSI Cubi NUC 1M | |
| PCMark 10 / Digital Content Creation | |
| Minisforum MS-A1 | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V (8914 - 10308, n=22) | |
| MSI Cubi NUC AI+ 2MG | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| CrossMark / Overall | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC 1M | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V (1551 - 1742, n=21) | |
| MSI Cubi NUC AI+ 2MG | |
| Ninkear M7, (16+512) | |
| CrossMark / Productivity | |
| Minisforum MS-A1 | |
| MSI Cubi NUC 1M | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V (1479 - 1684, n=21) | |
| MSI Cubi NUC AI+ 2MG | |
| Ninkear M7, (16+512) | |
| CrossMark / Creativity | |
| Minisforum MS-A1 | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V (1727 - 1977, n=21) | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| CrossMark / Responsiveness | |
| MSI Cubi NUC 1M | |
| Beelink SER9 Pro | |
| Minisforum MS-A1 | |
| GMK NucBox K12 | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V (1246 - 1411, n=21) | |
| MSI Cubi NUC AI+ 2MG | |
| Ninkear M7, (16+512) | |
| WebXPRT 3 / Overall | |
| Minisforum MS-A1 | |
| MSI Cubi NUC AI+ 2MG | |
| MSI Cubi NUC 1M | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V (219 - 317, n=21) | |
| Ninkear M7, (16+512) | |
| WebXPRT 4 / Overall | |
| MSI Cubi NUC AI+ 2MG | |
| Minisforum MS-A1 | |
| MSI Cubi NUC 1M | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V (199.5 - 291, n=21) | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| Ninkear M7, (16+512) | |
| Mozilla Kraken 1.1 / Total | |
| Ninkear M7, (16+512) | |
| GMK NucBox K12 | |
| Beelink SER9 Pro | |
| MSI Cubi NUC 1M | |
| Minisforum MS-A1 | |
| Average Intel Core Ultra 7 258V, Intel Arc Graphics 140V (468 - 559, n=23) | |
| MSI Cubi NUC AI+ 2MG | |
* ... smaller is better
| PCMark 10 Score | 6864 points | |
Help | ||
| AIDA64 / Memory Copy | |
| Average Intel Core Ultra 7 258V (104408 - 114336, n=23) | |
| MSI Cubi NUC AI+ 2MG | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| Minisforum MS-A1 | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| AIDA64 / Memory Read | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (82846 - 89417, n=23) | |
| MSI Cubi NUC 1M | |
| Minisforum MS-A1 | |
| GMK NucBox K12 | |
| Beelink SER9 Pro | |
| Ninkear M7, (16+512) | |
| AIDA64 / Memory Write | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Core Ultra 7 258V (108842 - 117989, n=23) | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| Minisforum MS-A1 | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| AIDA64 / Memory Latency | |
| Beelink SER9 Pro | |
| Average Intel Core Ultra 7 258V (95 - 101.1, n=23) | |
| MSI Cubi NUC AI+ 2MG | |
| Ninkear M7, (16+512) | |
| GMK NucBox K12 | |
| Minisforum MS-A1 | |
| MSI Cubi NUC 1M | |
* ... smaller is better
| DPC Latencies / LatencyMon - interrupt to process latency (max), Web, Youtube, Prime95 | |
| MSI Cubi NUC 1M | |
| Minisforum MS-A1 | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| Beelink SER9 Pro | |
* ... smaller is better
Storage Devices
SSD transfer rates with the primary 1 TB Phison are stable at almost 5900 MB/s even when under prolonged stress which is a slight improvement over the WD SSD in the original Cubi 1M.
| Drive Performance rating - Percent | |
| Minisforum MS-A1 | |
| MSI Cubi NUC 1M | |
| Beelink SER9 Pro | |
| MSI Cubi NUC AI+ 2MG | |
| GMK NucBox K12 | |
| Ninkear M7, (16+512) | |
* ... smaller is better
Disk Throttling: DiskSpd Read Loop, Queue Depth 8
GPU Performance
The Arc 140V in the Cubi 2MG performs below average by about 35 percent according to 3DMark benchmarks. The disappointing results are due to the cycling GPU clock rates and stuttering as mentioned in our Stress Test section above.
In GPU workloads where stable frame rates may not matter such as with renderers or oneAPI, the Cubi 2MG performs as expected relative to other PCs with the same CPU and GPU. The stuttering should therefore be a non-issue for most target professionals.
| 3DMark 11 Performance | 7781 points | |
| 3DMark Cloud Gate Standard Score | 31534 points | |
| 3DMark Fire Strike Score | 6005 points | |
| 3DMark Time Spy Score | 3053 points | |
Help | ||
* ... smaller is better
| Performance rating - Percent | |
| Beelink SER9 Pro | |
| Minisforum MS-A1 -1! | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| MSI Cubi NUC 1M -2! | |
| Ninkear M7, (16+512) -2! | |
| Cyberpunk 2077 2.2 Phantom Liberty - 1920x1080 Ultra Preset (FSR off) | |
| GMK NucBox K12 | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| MSI Cubi NUC AI+ 2MG | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| Baldur's Gate 3 - 1920x1080 Ultra Preset AA:T | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| Minisforum MS-A1 | |
| GTA V - 1920x1080 Highest Settings possible AA:4xMSAA + FX AF:16x | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC AI+ 2MG | |
| Minisforum MS-A1 | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| Final Fantasy XV Benchmark - 1920x1080 High Quality | |
| Beelink SER9 Pro | |
| MSI Cubi NUC AI+ 2MG | |
| Minisforum MS-A1 | |
| GMK NucBox K12 | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| Strange Brigade - 1920x1080 ultra AA:ultra AF:16 | |
| MSI Cubi NUC AI+ 2MG | |
| Beelink SER9 Pro | |
| GMK NucBox K12 | |
| MSI Cubi NUC 1M | |
| Dota 2 Reborn - 1920x1080 ultra (3/3) best looking | |
| Minisforum MS-A1 | |
| GMK NucBox K12 | |
| Beelink SER9 Pro | |
| MSI Cubi NUC AI+ 2MG | |
| MSI Cubi NUC 1M | |
| Ninkear M7, (16+512) | |
| X-Plane 11.11 - 1920x1080 high (fps_test=3) | |
| Minisforum MS-A1 | |
| GMK NucBox K12 | |
| Beelink SER9 Pro | |
| MSI Cubi NUC AI+ 2MG | |
| Ninkear M7, (16+512) | |
| F1 24 | |
| 1280x720 Ultra Low Preset AA:T | |
| MSI Cubi NUC AI+ 2MG | |
| Average Intel Arc Graphics 140V (n=1) | |
| 1920x1080 Low Preset AA:T AF:16x | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| Average Intel Arc Graphics 140V (54 - 99, n=24) | |
| MSI Cubi NUC AI+ 2MG | |
| Asus NUC 14 Pro+ | |
| Ninkear M7, (16+512) | |
| 1920x1080 Medium Preset AA:T AF:16x | |
| Minisforum MS-A1 | |
| Average Intel Arc Graphics 140V (49.2 - 88.2, n=23) | |
| Beelink SER9 Pro | |
| Asus NUC 14 Pro+ | |
| Ninkear M7, (16+512) | |
| 1920x1080 High Preset AA:T AF:16x | |
| Average Intel Arc Graphics 140V (34.4 - 66.4, n=25) | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
| MSI Cubi NUC AI+ 2MG | |
| Asus NUC 14 Pro+ | |
| Ninkear M7, (16+512) | |
| 1920x1080 Ultra High Preset AA:T AF:16x | |
| Average Intel Arc Graphics 140V (15.3 - 22, n=23) | |
| MSI Cubi NUC AI+ 2MG | |
| Minisforum MS-A1 | |
| Beelink SER9 Pro | |
Frame rates when idling on Cyberpunk 2077 at 1080p Ultra settings would cycle between 18 and 26 whereas repeating this test on the NucBox K12 would result in a stabler average frame rate of 23. This stuttering would occur when running other games in our test as well.
Cyberpunk 2077 ultra FPS Chart
| low | med. | high | ultra | |
|---|---|---|---|---|
| GTA V (2015) | 160.4 | 152.4 | 63.6 | 26.6 |
| Dota 2 Reborn (2015) | 98.1 | 80.4 | 59.9 | 65.8 |
| Final Fantasy XV Benchmark (2018) | 71.3 | 39.1 | 28.7 | |
| X-Plane 11.11 (2018) | 51.8 | 41.7 | 36 | |
| Strange Brigade (2018) | 240 | 71.8 | 59.7 | 57.3 |
| Baldur's Gate 3 (2023) | 36.1 | 29.5 | 29 | 26.7 |
| Cyberpunk 2077 2.2 Phantom Liberty (2023) | 33.9 | 26.6 | 23.2 | 19.4 |
| F1 24 (2024) | 72 | 53.8 | 16.6 |
Emissions
System Noise
Noise level
| Idle |
| 28 / 36.1 / 36.1 dB(A) |
| Load |
| 38.9 / 49.2 dB(A) |
![]() | ||
30 dB silent 40 dB(A) audible 50 dB(A) loud |
||
min: | ||
| MSI Cubi NUC AI+ 2MG Arc 140V, Core Ultra 7 258V | MSI Cubi NUC 1M Iris Xe G7 96EUs, Core 7 150U, WD PC SN560 SDDPNQE-1T00 | GMK NucBox K12 Radeon 780M, Ryzen 7 H 255, Crucial P3 Plus CT1000P3PSSD8 | Beelink SER9 Pro Radeon 780M, Ryzen 7 H 255, Crucial P3 Plus CT1000P3PSSD8 | Ninkear M7, (16+512) Vega 7, R5 7430U, Faspeed P8-512G-SE | Minisforum MS-A1 Radeon 780M, R7 8700G, Kingston OM8PGP41024Q-A0 | |
|---|---|---|---|---|---|---|
| Noise | 7% | -4% | 12% | 19% | -3% | |
| off / environment * (dB) | 24.8 | 23.5 5% | 23.5 5% | 23 7% | 24.8 -0% | 24.8 -0% |
| Idle Minimum * (dB) | 28 | 29.9 -7% | 35.4 -26% | 29.1 -4% | 25.3 10% | 33.9 -21% |
| Idle Average * (dB) | 36.1 | 29.9 17% | 35.4 2% | 30.1 17% | 25.7 29% | 34 6% |
| Idle Maximum * (dB) | 36.1 | 32 11% | 39 -8% | 31.7 12% | 25.7 29% | 34 6% |
| Load Average * (dB) | 38.9 | 33.5 14% | 42.4 -9% | 33.7 13% | 33.6 14% | 41.1 -6% |
| Cyberpunk 2077 ultra * (dB) | 37 | 39.1 -6% | 40.4 -9% | 33.8 9% | 28.7 22% | 42.2 -14% |
| Load Maximum * (dB) | 49.2 | 40.5 18% | 41.4 16% | 35.2 28% | 34.3 30% | 46.2 6% |
| Witcher 3 ultra * (dB) | 28.8 |
* ... smaller is better
Temperature
Speakers
MSI Cubi NUC AI+ 2MG audio analysis
(-) | not very loud speakers (70.5 dB)
Bass 100 - 315 Hz
(±) | reduced bass - on average 10.4% lower than median
(-) | bass is not linear (16% delta to prev. frequency)
Mids 400 - 2000 Hz
(±) | higher mids - on average 6.9% higher than median
(-) | mids are not linear (19.7% delta to prev. frequency)
Highs 2 - 16 kHz
(+) | balanced highs - only 4.3% away from median
(±) | linearity of highs is average (12.2% delta to prev. frequency)
Overall 100 - 16.000 Hz
(±) | linearity of overall sound is average (29.5% difference to median)
Compared to same class
» 70% of all tested devices in this class were better, 13% similar, 17% worse
» The best had a delta of 13%, average was 25%, worst was 64%
Compared to all devices tested
» 87% of all tested devices were better, 3% similar, 10% worse
» The best had a delta of 4%, average was 24%, worst was 134%
Apple MacBook Pro 16 2021 M1 Pro audio analysis
(+) | speakers can play relatively loud (84.7 dB)
Bass 100 - 315 Hz
(+) | good bass - only 3.8% away from median
(+) | bass is linear (5.2% delta to prev. frequency)
Mids 400 - 2000 Hz
(+) | balanced mids - only 1.3% away from median
(+) | mids are linear (2.1% delta to prev. frequency)
Highs 2 - 16 kHz
(+) | balanced highs - only 1.9% away from median
(+) | highs are linear (2.7% delta to prev. frequency)
Overall 100 - 16.000 Hz
(+) | overall sound is linear (4.6% difference to median)
Compared to same class
» 0% of all tested devices in this class were better, 0% similar, 100% worse
» The best had a delta of 5%, average was 17%, worst was 45%
Compared to all devices tested
» 0% of all tested devices were better, 0% similar, 100% worse
» The best had a delta of 4%, average was 24%, worst was 134%
Energy Management
Power Consumption
Overall consumption is noticeably lower than an AMD H series mini PC like the GMK NucBox K12 even though our integrated Arc 140V is roughly comparable to the competing integrated Radeon 780M when gaming. Performance-per-watt generally favors the MSI Cubi albeit at the cost of raw processor performance when compared to many AMD Zen 4 or Zen 5 alternatives.
Consumption when running Cyberpunk 2077 would cycle between 28 W and 49 W for seemingly no reason which aligns with our unstable frame rates and GPU clock rates mentioned above.
| Off / Standby | |
| Idle | |
| Load |
|
Key:
min: | |
| MSI Cubi NUC AI+ 2MG Core Ultra 7 258V, Arc 140V, , , x, | MSI Cubi NUC 1M Core 7 150U, Iris Xe G7 96EUs, WD PC SN560 SDDPNQE-1T00, , x, | GMK NucBox K12 Ryzen 7 H 255, Radeon 780M, Crucial P3 Plus CT1000P3PSSD8, , x, | Beelink SER9 Pro Ryzen 7 H 255, Radeon 780M, Crucial P3 Plus CT1000P3PSSD8, , x, | Ninkear M7, (16+512) R5 7430U, Vega 7, Faspeed P8-512G-SE, , x, | Minisforum MS-A1 R7 8700G, Radeon 780M, Kingston OM8PGP41024Q-A0, , x, | |
|---|---|---|---|---|---|---|
| Power Consumption | -24% | -99% | -56% | -47% | -121% | |
| Idle Minimum * (Watt) | 2.9 | 6.4 -121% | 9.9 -241% | 7 -141% | 9.3 -221% | 8.8 -203% |
| Idle Average * (Watt) | 5.7 | 8.3 -46% | 10.1 -77% | 7.1 -25% | 10.7 -88% | 12.6 -121% |
| Idle Maximum * (Watt) | 10.2 | 10.4 -2% | 13.1 -28% | 7.5 26% | 12.6 -24% | 14 -37% |
| Load Average * (Watt) | 38.7 | 38.5 1% | 69.8 -80% | 62.3 -61% | 37.7 3% | 93.2 -141% |
| Cyberpunk 2077 ultra external monitor * (Watt) | 41.1 | 36.1 12% | 86.3 -110% | 80.1 -95% | 32.3 21% | 102.1 -148% |
| Load Maximum * (Watt) | 60 | 52.5 12% | 92.8 -55% | 82.9 -38% | 43.4 28% | 105.4 -76% |
* ... smaller is better
Power Consumption Cyberpunk / Stress Test
Power Consumption external Monitor
Notebookcheck Rating
Basic and reliable mini PC for business, but it's arguably too basic considering the high asking price.
MSI Cubi NUC AI+ 2MG
- 09/29/2025 v8
Allen Ngo
Potential Competitors in Comparison
Image | Model / Review | Price | Weight | Height | Display |
|---|---|---|---|---|---|
| MSI Cubi NUC AI+ 2MG Intel Core Ultra 7 258V ⎘ Intel Arc Graphics 140V ⎘ 32 GB Memory | Amazon: $899.00 List Price: 900 USD | 669 g | 50.1 mm | x | |
| MSI Cubi NUC 1M Intel Core 7 150U ⎘ Intel Iris Xe Graphics G7 96EUs ⎘ 16 GB Memory, 1024 GB SSD | Amazon: List Price: 950 USD | 660 g | 50.1 mm | x | |
| GMK NucBox K12 AMD Ryzen 7 H 255 ⎘ AMD Radeon 780M ⎘ 32 GB Memory, 1024 GB SSD | Amazon: $549.98 | 938 g | 78 mm | x | |
| Beelink SER9 Pro AMD Ryzen 7 H 255 ⎘ AMD Radeon 780M ⎘ 32 GB Memory, 1024 GB SSD | Amazon: $649.00 List Price: 520 USD | 812 g | 44.7 mm | x | |
| Ninkear M7, (16+512) AMD Ryzen 5 7430U ⎘ AMD Radeon RX Vega 7 ⎘ 16 GB Memory, 512 GB SSD | Amazon: $299.99 List Price: 399€ | 400 g | 36 mm | x | |
| Minisforum MS-A1 AMD Ryzen 7 8700G ⎘ AMD Radeon 780M ⎘ 32 GB Memory, 1024 GB SSD | Amazon: 1. $1,499.90 MINISFORUM MS-A1 Mini Workst... 2. $689.00 MINISFORUM Mini Workstation ... 3. $635.00 MINISFORUM Mini PC MS-01-S13... List Price: 829 EUR | 1.7 kg | 48 mm | x |
Transparency
The selection of devices to be reviewed is made by our editorial team. The test sample was provided to the author as a loan by the manufacturer or retailer for the purpose of this review. The lender had no influence on this review, nor did the manufacturer receive a copy of this review before publication. There was no obligation to publish this review. As an independent media company, Notebookcheck is not subjected to the authority of manufacturers, retailers or publishers.
This is how Notebookcheck is testing
Every year, Notebookcheck independently reviews hundreds of laptops and smartphones using standardized procedures to ensure that all results are comparable. We have continuously developed our test methods for around 20 years and set industry standards in the process. In our test labs, high-quality measuring equipment is utilized by experienced technicians and editors. These tests involve a multi-stage validation process. Our complex rating system is based on hundreds of well-founded measurements and benchmarks, which maintains objectivity. Further information on our test methods can be found here.



























































