
BOSGAME M2 迷你电脑评测:修正了原版 M1 的一大缺陷
同样的外观,全新的处理器。
M2 解决了我们在 M1 上报道过的音频问题,同时将基本处理器选项从 Ryzen 7 提升到了 Ryzen 9。 不过,我们认为制造商本应再接再厉,更好地区分这两款机型。Allen Ngo, 👁 Allen Ngo (translated by DeepL / Ninh Duy) Published 🇺🇸 🇫🇷 ...
结论 - 与 M1 性能基本相同
Bosgame M2 解决了原版 M1 中出现的音频延迟问题。 原版 M1 的音频延迟问题。因此,较新的型号可能会吸引重视扬声器的用户,因为这种功能在大多数迷你 PC 上并不常见。除此之外,尽管处理器 "升级 "了,但性能基本相同。
对于未来可能的改版,我们希望看到该系统不再使用专有的交流适配器,而是改用通用的 USB-C Power Delivery。同样令人遗憾的是,M2 继续使用 AMD Zen 4,而 Zen 5 已经集成了人工智能加速功能。从这个意义上说,M2 感觉更像是 M1 的一个独立 SKU,而不是真正的继任者。或许 "M1 Pro "或 "M1 Ryzen 9 Edition "才是更合适的名称。
Pros
Cons
价格和供应情况
亚马逊现已开始销售Bosgame M2,售价 600 美元,配备 32 GB 内存和 1 TB SSD。用户也可以直接从制造商处购买https://www.bosgamepc.com/products/bosgame-mini-pc-m2-amd-ryzen-97940hs价格为 700 美元。
可能的替代品比较
Image | Model / Review | Price | Weight | Height | Display |
---|---|---|---|---|---|
BOSGAME M2 AMD Ryzen 9 7940HS ⎘ AMD Radeon 780M ⎘ 32 GB Memory | Amazon: $699.00 List Price: 600 USD | 819 g | 51 mm | x | |
BOSGAME M1 AMD Ryzen 7 7840HS ⎘ AMD Radeon 780M ⎘ 32 GB Memory | Amazon: 1. $619.00 BOSGAME Mini Gaming PC M1, M... 2. $499.99 BOSGAME Gaming Mini PC Ryzen... 3. $59.99 PowerHOOD AC Power Adapter C... List Price: 690 USD | 766 g | 51 mm | x | |
GMK NucBox M7 Pro AMD Ryzen 9 PRO 6950H ⎘ AMD Radeon 680M ⎘ 32 GB Memory | Amazon: List Price: 440 USD | 621 g | 58.4 mm | x | |
Beelink SER9 AMD Ryzen AI 9 HX 370 ⎘ AMD Radeon 890M ⎘ 32 GB Memory, 1024 GB SSD | Amazon: $999.00 List Price: 1000 USD | 814 g | 44.7 mm | x | |
Asus NUC 14 Pro Intel Core Ultra 5 125H ⎘ Intel Arc 7-Core iGPU ⎘ , 2048 GB SSD | Amazon: 1. $627.99 ASUS NUC 14 Pro Tall Barebon... 2. $1,000.00 ASUS NUC 14 Pro+ Full System... 3. $909.00 ASUS NUC 14 Pro AI Mini PC, ... List Price: 500 EUR | 525 g | 37 mm | x | |
Intel NUC 13 Pro Desk Edition Kit Intel Core i7-1360P ⎘ Intel Iris Xe Graphics G7 96EUs ⎘ 32 GB Memory, 512 GB SSD | Amazon: 1. $582.00 Intel NUC 13 Pro Kit NUC13AN... 2. $666.99 Intel NUC 13 Pro Kit NUC13AN... 3. $706.00 Intel NUC 13 Pro Kit NUC13AN... List Price: 800€ | 558 g | 37 mm | x |
Bosgame M2 继承了 Bosgame M1,将 Ryzen 7 7840HS 升级到了 Ryzen 9 7940HS。除此之外,它从上到下基本上都是同一款迷你 PC。
我们的测试机配备了上述 Ryzen 9 CPU、32 GB 内存和 1 TB NVMe SSD,售价约为 600 美元。在撰写本文时,其他 SKU 尚未广泛上市。
Bosgame M2 的替代产品包括其他高性能迷你 PC,如 Beelink SER9, 华硕 NUC 14 Pro或 GMK NucBox M7 Pro.
更多 Bosgame 评论:
» Notebookcheck多媒体笔记本电脑Top 10排名
» Notebookcheck游戏笔记本电脑Top 10排名
» Notebookcheck低价办公/商务笔记本电脑Top 10排名
» Notebookcheck高端办公/商务笔记本电脑Top 10排名
» Notebookcheck工作站笔记本电脑Top 10排名
» Notebookcheck亚笔记本电脑Top 10排名
» Notebookcheck超级本产品Top 10排名
» Notebookcheck变形本产品Top 10排名
» Notebookcheck平板电脑Top 10排名
» Notebookcheck智能手机Top 10排名
» Notebookcheck评测过最出色的笔记本电脑屏幕
» Notebookcheck售价500欧元以下笔记本电脑Top 10排名
» Notebookcheck售价300欧元以下笔记本电脑Top 10排名
规格
案例
金属机身与去年的 M1 保持一致。用户必须在机身底部寻找 M2 标签,以便从视觉上区分该机型与前代 M1。
连接性
端口选项没有改变,包括使用专有的交流电桶适配器端口,而不是完全依赖通用的 USB-C Power Delivery。
交流
英特尔 AX210 标配支持 Wi-Fi 6E 和蓝牙 5.3。BE200 最好能提供 Wi-Fi 7 连接,以便更好地将 M2 与原来的 M1 型号区分开来。
Networking | |
iperf3 transmit AXE11000 | |
BOSGAME M1 | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
GMK NucBox M7 Pro | |
BOSGAME M2 | |
Beelink SER9 | |
iperf3 receive AXE11000 | |
BOSGAME M1 | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
BOSGAME M2 | |
Beelink SER9 | |
iperf3 transmit AXE11000 6GHz | |
Intel NUC 13 Pro Desk Edition Kit | |
Asus NUC 14 Pro | |
BOSGAME M1 | |
BOSGAME M2 | |
iperf3 receive AXE11000 6GHz | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
BOSGAME M1 | |
BOSGAME M2 |
可持续性
制造商没有提到在生产过程中采取了任何可持续发展措施。包装本身主要是硬纸板和一些塑料包装。
配件和保修
零售包装盒包括一个交流适配器、可拆卸的 VESA 安装支架、HDMI 线缆和用户手册。
如果在美国购买,制造商提供一年有限保修。
性能
测试条件
在运行以下基准测试之前,我们将 Windows 设置为性能模式。没有预装特殊的 "控制中心 "应用程序。
处理器
......(英文 Ryzen 9 7940HS 性能几乎与 Ryzen 7 7840HS.它们的纸面规格仅相差几百 MHz,因此对于熟悉 Ryzen 系列的发烧友来说,结果应该不足为奇。
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 6.4: 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 | |
Beelink SER9 | |
Beelink SER8 | |
Average AMD Ryzen 9 7940HS | |
BOSGAME M2 | |
BOSGAME M1 | |
Intel NUC 13 Pro Desk Edition Kit | |
Geekom XT13 Pro | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Beelink SER6 Pro -2! | |
GMK NucBox M6 | |
MSI Cubi NUC 1M | |
Beelink SER5 |
Cinebench R23 / Multi Core | |
Beelink SER9 | |
Average AMD Ryzen 9 7940HS (15556 - 18131, n=11) | |
Beelink SER8 | |
BOSGAME M2 | |
BOSGAME M1 | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
Geekom XT13 Pro | |
Beelink SER6 Pro | |
GMK NucBox M6 | |
Beelink SER5 | |
MSI Cubi NUC 1M |
Cinebench R23 / Single Core | |
Beelink SER9 | |
Intel NUC 13 Pro Desk Edition Kit | |
MSI Cubi NUC 1M | |
Average AMD Ryzen 9 7940HS (1733 - 1832, n=11) | |
Beelink SER8 | |
BOSGAME M2 | |
Asus NUC 14 Pro | |
Geekom XT13 Pro | |
BOSGAME M1 | |
GMK NucBox M7 Pro | |
Beelink SER6 Pro | |
GMK NucBox M6 | |
Beelink SER5 |
Cinebench R20 / CPU (Multi Core) | |
Beelink SER9 | |
Average AMD Ryzen 9 7940HS (6277 - 7111, n=11) | |
Beelink SER8 | |
BOSGAME M1 | |
BOSGAME M2 | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro | |
Geekom XT13 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
Beelink SER6 Pro | |
GMK NucBox M6 | |
Beelink SER5 | |
MSI Cubi NUC 1M |
Cinebench R20 / CPU (Single Core) | |
Beelink SER9 | |
Intel NUC 13 Pro Desk Edition Kit | |
MSI Cubi NUC 1M | |
Average AMD Ryzen 9 7940HS (679 - 714, n=11) | |
Beelink SER8 | |
BOSGAME M1 | |
BOSGAME M2 | |
Geekom XT13 Pro | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Beelink SER6 Pro | |
GMK NucBox M6 | |
Beelink SER5 |
Cinebench R15 / CPU Multi 64Bit | |
Beelink SER9 | |
Average AMD Ryzen 9 7940HS (2470 - 2848, n=11) | |
Beelink SER8 | |
BOSGAME M1 | |
BOSGAME M2 | |
Geekom XT13 Pro | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
Beelink SER6 Pro | |
GMK NucBox M6 | |
Beelink SER5 | |
MSI Cubi NUC 1M |
Cinebench R15 / CPU Single 64Bit | |
Beelink SER9 | |
Beelink SER8 | |
Average AMD Ryzen 9 7940HS (258 - 288, n=11) | |
BOSGAME M1 | |
Intel NUC 13 Pro Desk Edition Kit | |
GMK NucBox M7 Pro | |
MSI Cubi NUC 1M | |
BOSGAME M2 | |
Asus NUC 14 Pro | |
Geekom XT13 Pro | |
GMK NucBox M6 | |
Beelink SER6 Pro | |
Beelink SER5 |
Blender / v2.79 BMW27 CPU | |
MSI Cubi NUC 1M | |
Beelink SER5 | |
GMK NucBox M6 | |
Beelink SER6 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
Geekom XT13 Pro | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
BOSGAME M2 | |
BOSGAME M1 | |
Average AMD Ryzen 9 7940HS (182 - 202, n=10) | |
Beelink SER8 | |
Beelink SER9 |
7-Zip 18.03 / 7z b 4 | |
Beelink SER9 | |
Beelink SER8 | |
Average AMD Ryzen 9 7940HS (66408 - 73434, n=11) | |
BOSGAME M1 | |
BOSGAME M2 | |
GMK NucBox M7 Pro | |
Geekom XT13 Pro | |
Beelink SER6 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
GMK NucBox M6 | |
Beelink SER5 | |
MSI Cubi NUC 1M | |
Asus NUC 14 Pro |
7-Zip 18.03 / 7z b 4 -mmt1 | |
Beelink SER8 | |
Average AMD Ryzen 9 7940HS (6310 - 6776, n=11) | |
BOSGAME M1 | |
BOSGAME M2 | |
Beelink SER9 | |
GMK NucBox M7 Pro | |
MSI Cubi NUC 1M | |
Intel NUC 13 Pro Desk Edition Kit | |
Geekom XT13 Pro | |
Beelink SER6 Pro | |
GMK NucBox M6 | |
Asus NUC 14 Pro | |
Beelink SER5 |
Geekbench 6.4 / Multi-Core | |
Beelink SER9 | |
Beelink SER8 | |
BOSGAME M2 | |
BOSGAME M1 | |
Average AMD Ryzen 9 7940HS (9899 - 13470, n=9) | |
Asus NUC 14 Pro | |
Geekom XT13 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
GMK NucBox M7 Pro | |
GMK NucBox M6 | |
MSI Cubi NUC 1M | |
Beelink SER5 |
Geekbench 6.4 / Single-Core | |
Beelink SER9 | |
BOSGAME M2 | |
Beelink SER8 | |
MSI Cubi NUC 1M | |
BOSGAME M1 | |
Average AMD Ryzen 9 7940HS (2426 - 2746, n=9) | |
Intel NUC 13 Pro Desk Edition Kit | |
Geekom XT13 Pro | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro | |
GMK NucBox M6 | |
Beelink SER5 |
Geekbench 5.5 / Multi-Core | |
Beelink SER9 | |
BOSGAME M2 | |
Beelink SER8 | |
BOSGAME M1 | |
Average AMD Ryzen 9 7940HS (11207 - 12500, n=12) | |
Geekom XT13 Pro | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
GMK NucBox M7 Pro | |
Beelink SER6 Pro | |
GMK NucBox M6 | |
MSI Cubi NUC 1M | |
Beelink SER5 |
Geekbench 5.5 / Single-Core | |
Beelink SER8 | |
BOSGAME M2 | |
Average AMD Ryzen 9 7940HS (1823 - 2032, n=12) | |
Beelink SER9 | |
BOSGAME M1 | |
MSI Cubi NUC 1M | |
Intel NUC 13 Pro Desk Edition Kit | |
Geekom XT13 Pro | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro | |
Beelink SER6 Pro | |
GMK NucBox M6 | |
Beelink SER5 |
HWBOT x265 Benchmark v2.2 / 4k Preset | |
Beelink SER9 | |
BOSGAME M2 | |
Beelink SER8 | |
Average AMD Ryzen 9 7940HS (20.1 - 22.6, n=10) | |
BOSGAME M1 | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Beelink SER6 Pro | |
Geekom XT13 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
GMK NucBox M6 | |
Beelink SER5 | |
MSI Cubi NUC 1M |
LibreOffice / 20 Documents To PDF | |
Beelink SER5 | |
Asus NUC 14 Pro | |
GMK NucBox M6 | |
Beelink SER9 | |
Beelink SER6 Pro | |
GMK NucBox M7 Pro | |
BOSGAME M1 | |
Intel NUC 13 Pro Desk Edition Kit | |
Geekom XT13 Pro | |
Beelink SER8 | |
Average AMD Ryzen 9 7940HS (36.9 - 54.8, n=11) | |
BOSGAME M2 | |
MSI Cubi NUC 1M |
R Benchmark 2.5 / Overall mean | |
Beelink SER5 | |
GMK NucBox M6 | |
Beelink SER6 Pro | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro | |
Geekom XT13 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
MSI Cubi NUC 1M | |
BOSGAME M1 | |
Beelink SER8 | |
Average AMD Ryzen 9 7940HS (0.4159 - 0.4467, n=11) | |
BOSGAME M2 | |
Beelink SER9 |
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 6.4: 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
* ... 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 | |
Beelink SER9 | |
BOSGAME M2 | |
BOSGAME M1 | |
Average AMD Ryzen 9 7940HS | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / FP32 Ray-Trace | |
Beelink SER9 | |
Average AMD Ryzen 9 7940HS (25226 - 29109, n=9) | |
BOSGAME M2 | |
BOSGAME M1 | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / FPU Julia | |
Beelink SER9 | |
Average AMD Ryzen 9 7940HS (106512 - 123214, n=9) | |
BOSGAME M2 | |
BOSGAME M1 | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / CPU SHA3 | |
Beelink SER9 | |
BOSGAME M2 | |
Average AMD Ryzen 9 7940HS (4566 - 5119, n=9) | |
BOSGAME M1 | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / CPU Queen | |
Average AMD Ryzen 9 7940HS (118174 - 124783, n=9) | |
BOSGAME M2 | |
BOSGAME M1 | |
Beelink SER9 | |
GMK NucBox M7 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
Asus NUC 14 Pro |
AIDA64 / FPU SinJulia | |
Beelink SER9 | |
BOSGAME M2 | |
Average AMD Ryzen 9 7940HS (14663 - 15551, n=9) | |
BOSGAME M1 | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / FPU Mandel | |
Beelink SER9 | |
Average AMD Ryzen 9 7940HS (56839 - 66238, n=9) | |
BOSGAME M2 | |
BOSGAME M1 | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / CPU AES | |
BOSGAME M2 | |
BOSGAME M1 | |
Average AMD Ryzen 9 7940HS (86309 - 171350, n=9) | |
GMK NucBox M7 Pro | |
Beelink SER9 | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / CPU ZLib | |
Beelink SER9 | |
Average AMD Ryzen 9 7940HS (866 - 1152, n=9) | |
BOSGAME M2 | |
BOSGAME M1 | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / FP64 Ray-Trace | |
Beelink SER9 | |
Average AMD Ryzen 9 7940HS (13175 - 15675, n=9) | |
BOSGAME M2 | |
BOSGAME M1 | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / CPU PhotoWorxx | |
Beelink SER9 | |
Asus NUC 14 Pro | |
BOSGAME M2 | |
BOSGAME M1 | |
Average AMD Ryzen 9 7940HS (29922 - 41567, n=9) | |
GMK NucBox M7 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
CrossMark: Overall | Productivity | Creativity | Responsiveness
WebXPRT 3: Overall
WebXPRT 4: Overall
Mozilla Kraken 1.1: Total
PCMark 10 / Score | |
Beelink SER9 | |
BOSGAME M2 | |
Average AMD Ryzen 9 7940HS, AMD Radeon 780M (7048 - 7520, n=4) | |
BOSGAME M1 | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
PCMark 10 / Essentials | |
BOSGAME M1 | |
BOSGAME M2 | |
Intel NUC 13 Pro Desk Edition Kit | |
Average AMD Ryzen 9 7940HS, AMD Radeon 780M (10448 - 11114, n=4) | |
Beelink SER9 | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro |
PCMark 10 / Productivity | |
Average AMD Ryzen 9 7940HS, AMD Radeon 780M (9919 - 10717, n=4) | |
BOSGAME M2 | |
Beelink SER9 | |
BOSGAME M1 | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
PCMark 10 / Digital Content Creation | |
Beelink SER9 | |
BOSGAME M1 | |
BOSGAME M2 | |
Average AMD Ryzen 9 7940HS, AMD Radeon 780M (9119 - 9929, n=4) | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
CrossMark / Overall | |
BOSGAME M1 | |
Average AMD Ryzen 9 7940HS, AMD Radeon 780M (1726 - 1861, n=4) | |
BOSGAME M2 | |
Beelink SER9 | |
Intel NUC 13 Pro Desk Edition Kit | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro |
CrossMark / Productivity | |
BOSGAME M1 | |
Average AMD Ryzen 9 7940HS, AMD Radeon 780M (1672 - 1781, n=4) | |
BOSGAME M2 | |
Beelink SER9 | |
Intel NUC 13 Pro Desk Edition Kit | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro |
CrossMark / Creativity | |
BOSGAME M1 | |
Average AMD Ryzen 9 7940HS, AMD Radeon 780M (1914 - 2072, n=4) | |
BOSGAME M2 | |
Beelink SER9 | |
Intel NUC 13 Pro Desk Edition Kit | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro |
CrossMark / Responsiveness | |
BOSGAME M1 | |
BOSGAME M2 | |
Beelink SER9 | |
Average AMD Ryzen 9 7940HS, AMD Radeon 780M (1382 - 1562, n=4) | |
Intel NUC 13 Pro Desk Edition Kit | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro |
WebXPRT 3 / Overall | |
BOSGAME M2 | |
Average AMD Ryzen 9 7940HS, AMD Radeon 780M (302 - 342, n=5) | |
BOSGAME M1 | |
GMK NucBox M7 Pro | |
Intel NUC 13 Pro Desk Edition Kit | |
Beelink SER9 | |
Asus NUC 14 Pro |
WebXPRT 4 / Overall | |
BOSGAME M2 | |
Intel NUC 13 Pro Desk Edition Kit | |
Average AMD Ryzen 9 7940HS, AMD Radeon 780M (234 - 274, n=4) | |
BOSGAME M1 | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Beelink SER9 |
Mozilla Kraken 1.1 / Total | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
BOSGAME M1 | |
BOSGAME M2 | |
Average AMD Ryzen 9 7940HS, AMD Radeon 780M (487 - 514, n=5) | |
Intel NUC 13 Pro Desk Edition Kit | |
Beelink SER9 |
* ... smaller is better
PCMark 10 Score | 7377 points | |
Help |
AIDA64 / Memory Copy | |
Beelink SER9 | |
BOSGAME M2 | |
BOSGAME M1 | |
Asus NUC 14 Pro | |
Average AMD Ryzen 9 7940HS (59712 - 78739, n=9) | |
GMK NucBox M7 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / Memory Read | |
Beelink SER9 | |
Asus NUC 14 Pro | |
BOSGAME M1 | |
BOSGAME M2 | |
Average AMD Ryzen 9 7940HS (50399 - 61836, n=9) | |
GMK NucBox M7 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / Memory Write | |
Beelink SER9 | |
BOSGAME M2 | |
BOSGAME M1 | |
Average AMD Ryzen 9 7940HS (72541 - 91567, n=9) | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro | |
Intel NUC 13 Pro Desk Edition Kit |
AIDA64 / Memory Latency | |
Asus NUC 14 Pro | |
Beelink SER9 | |
BOSGAME M2 | |
Intel NUC 13 Pro Desk Edition Kit | |
Average AMD Ryzen 9 7940HS (87.6 - 111.1, n=9) | |
BOSGAME M1 | |
GMK NucBox M7 Pro |
* ... smaller is better
DPC 延迟
DPC Latencies / LatencyMon - interrupt to process latency (max), Web, Youtube, Prime95 | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
BOSGAME M1 | |
Intel NUC 13 Pro Desk Edition Kit | |
BOSGAME M2 |
* ... smaller is better
存储设备
我们的测试设备配备了相同的 金士顿 OM8SEP41024N-A0NVMe 固态硬盘。最高传输速率约为 4000 MB/s,没有明显的性能节流问题。
Drive Performance Rating - Percent | |
Intel NUC 13 Pro Desk Edition Kit | |
Asus NUC 14 Pro | |
BOSGAME M2 | |
Beelink SER9 | |
BOSGAME M1 | |
GMK NucBox M7 Pro |
* ... smaller is better
Disk Throttling: DiskSpd Read Loop, Queue Depth 8
图形处理器性能
Ryzen 7 7840HS 和 Ryzen 9 7940HS 都集成了相同的 Radeon 780M图形处理器,两者在图形性能上没有重大差异。
3DMark Performance Rating - Percent | |
Beelink SER9 | |
BOSGAME M1 | |
BOSGAME M2 | |
Beelink SER8 | |
Asus NUC 14 Pro | |
Average AMD Radeon 780M | |
GMK NucBox M7 Pro -1! | |
Beelink SER6 Pro | |
Geekom XT13 Pro -1! | |
Intel NUC 13 Pro Desk Edition Kit -1! | |
GMK NucBox M6 -1! | |
MSI Cubi NUC 1M -1! | |
Beelink SER5 -1! |
3DMark 11 Performance | 14017 points | |
3DMark Cloud Gate Standard Score | 38111 points | |
3DMark Fire Strike Score | 8145 points | |
3DMark Time Spy Score | 3541 points | |
3DMark Steel Nomad Score | 511 points | |
Help |
* ... smaller is better
Cyberpunk 2077 2.2 Phantom Liberty | |
1280x720 Low Preset (FSR off) | |
Beelink SER9 | |
BOSGAME M1 | |
BOSGAME M2 | |
Beelink SER8 | |
Average AMD Radeon 780M (53.6 - 80.6, n=23) | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Geekom XT13 Pro | |
MSI Cubi NUC 1M | |
1920x1080 Low Preset (FSR off) | |
Beelink SER9 | |
BOSGAME M1 | |
BOSGAME M2 | |
Beelink SER8 | |
Average AMD Radeon 780M (29.3 - 52.1, n=44) | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
GMK NucBox M6 | |
Geekom XT13 Pro | |
MSI Cubi NUC 1M | |
1920x1080 Medium Preset (FSR off) | |
Beelink SER9 | |
BOSGAME M1 | |
BOSGAME M2 | |
Beelink SER8 | |
Average AMD Radeon 780M (16.3 - 37.5, n=40) | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
GMK NucBox M6 | |
Geekom XT13 Pro | |
MSI Cubi NUC 1M | |
1920x1080 High Preset (FSR off) | |
BOSGAME M1 | |
BOSGAME M2 | |
Beelink SER9 | |
Beelink SER8 | |
Average AMD Radeon 780M (13.8 - 29.6, n=41) | |
GMK NucBox M7 Pro | |
Asus NUC 14 Pro | |
Geekom XT13 Pro | |
GMK NucBox M6 | |
MSI Cubi NUC 1M | |
1920x1080 Ultra Preset (FSR off) | |
BOSGAME M1 | |
BOSGAME M2 | |
Beelink SER9 | |
Beelink SER8 | |
Average AMD Radeon 780M (11.5 - 26.3, n=36) | |
Asus NUC 14 Pro | |
GMK NucBox M7 Pro | |
Geekom XT13 Pro | |
MSI Cubi NUC 1M |
Cyberpunk 2077 ultra FPS Chart
low | med. | high | ultra | |
---|---|---|---|---|
GTA V (2015) | 166.6 | 155.8 | 69.4 | 28.5 |
Dota 2 Reborn (2015) | 155.8 | 122.7 | 98.4 | 88.1 |
Final Fantasy XV Benchmark (2018) | 83.8 | 41.5 | 30.1 | |
X-Plane 11.11 (2018) | 66.1 | 49.1 | 41.4 | |
Far Cry 5 (2018) | 88 | 53 | 49 | 46 |
Strange Brigade (2018) | 175.1 | 70.4 | 59.7 | 51.2 |
Tiny Tina's Wonderlands (2022) | 76.9 | 55.9 | 37.8 | 30.3 |
F1 22 (2022) | 107.3 | 93.3 | 69.1 | 19.3 |
Baldur's Gate 3 (2023) | 44.1 | 36.5 | 30.6 | 29.6 |
Cyberpunk 2077 2.2 Phantom Liberty (2023) | 46.3 | 36.3 | 29.2 | 25.8 |
排放量
系统噪音
根据我们的测量,风扇噪音和特性与 M1 基本相同。游戏时,系统噪音略高于 40 dB(A),比价格较高的 Beelink SER9 但也比性能较弱的 华硕 NUC 14 Pro.
Noise Level
Idle |
| 31.5 / 32.3 / 34.9 dB(A) |
Load |
| 39 / 42.5 dB(A) |
![]() | ||
30 dB silent 40 dB(A) audible 50 dB(A) loud |
||
min: ![]() ![]() ![]() |
BOSGAME M2 Radeon 780M, R9 7940HS | BOSGAME M1 Radeon 780M, R7 7840HS | GMK NucBox M7 Pro Radeon 680M, R9 PRO 6950H | Beelink SER9 Radeon 890M, Ryzen AI 9 HX 370, Crucial P3 Plus CT1000P3PSSD8 | Asus NUC 14 Pro Arc 7-Core, Ultra 5 125H, Kingston KC3000 2048G SKC3000D2048G | Intel NUC 13 Pro Desk Edition Kit Iris Xe G7 96EUs, i7-1360P, Samsung PM9A1 MZVL2512HCJQ | |
---|---|---|---|---|---|---|
Noise | 1% | -3% | 15% | 3% | 4% | |
off / environment * | 23 | 23.2 -1% | 23 -0% | 23.2 -1% | 24 -4% | 25.2 -10% |
Idle Minimum * | 31.5 | 32 -2% | 35.9 -14% | 26.6 16% | 26 17% | 28.1 11% |
Idle Average * | 32.3 | 32 1% | 36.7 -14% | 26.8 17% | 26.2 19% | 28.2 13% |
Idle Maximum * | 34.9 | 32 8% | 36.7 -5% | 26.8 23% | 26.2 25% | 28.3 19% |
Load Average * | 39 | 38.7 1% | 38.4 2% | 33.4 14% | 42.4 -9% | 42 -8% |
Cyberpunk 2077 ultra * | 40.3 | 38.7 4% | 34.6 14% | 47.7 -18% | ||
Load Maximum * | 42.5 | 42 1% | 39.8 6% | 33.8 20% | 47.8 -12% | 42.1 1% |
Witcher 3 ultra * | 41.7 |
* ... smaller is better
压力测试
CPU 平均时钟 (GHz) | GPU 时钟 (MHz) | CPU 平均温度 (°C) | |
系统闲置 | -- | -- | 32 |
Prime95 压力 | 4.0 | 800 | 87 |
Prime95 + FurMark Stress | 4.2 | 2078 | 74 |
赛博朋克 2077》压力 | 1.2 | 2800 | 66 |
发言人
BOSGAME M2 audio analysis
(±) | speaker loudness is average but good (81.2 dB)
Bass 100 - 315 Hz
(±) | reduced bass - on average 13.6% lower than median
(+) | bass is linear (5.9% delta to prev. frequency)
Mids 400 - 2000 Hz
(±) | higher mids - on average 5.2% higher than median
(+) | mids are linear (3.7% delta to prev. frequency)
Highs 2 - 16 kHz
(+) | balanced highs - only 2.8% away from median
(+) | highs are linear (3.3% delta to prev. frequency)
Overall 100 - 16.000 Hz
(±) | linearity of overall sound is average (16.3% difference to median)
Compared to same class
» 11% of all tested devices in this class were better, 16% similar, 74% worse
» The best had a delta of 13%, average was 25%, worst was 64%
Compared to all devices tested
» 24% of all tested devices were better, 5% similar, 71% 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 18%, 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%
能源管理
耗电量
Off / Standby | ![]() ![]() |
Idle | ![]() ![]() ![]() |
Load |
![]() ![]() |
![]() ![]() ![]() ![]() ![]() ![]() ![]() | |
Key:
min: ![]() ![]() ![]() |
BOSGAME M2 R9 7940HS, Radeon 780M, , , x, | BOSGAME M1 R7 7840HS, Radeon 780M, , , x, | GMK NucBox M7 Pro R9 PRO 6950H, Radeon 680M, , , x, | Beelink SER9 Ryzen AI 9 HX 370, Radeon 890M, Crucial P3 Plus CT1000P3PSSD8, , x, | Asus NUC 14 Pro Ultra 5 125H, Arc 7-Core, Kingston KC3000 2048G SKC3000D2048G, , x, | Intel NUC 13 Pro Desk Edition Kit i7-1360P, Iris Xe G7 96EUs, Samsung PM9A1 MZVL2512HCJQ, , x, | |
---|---|---|---|---|---|---|
Power Consumption | -4% | -58% | 14% | -2% | -10% | |
Idle Minimum * | 6.4 | 7.3 -14% | 14.6 -128% | 5.7 11% | 5.1 20% | 7 -9% |
Idle Average * | 7.3 | 7.3 -0% | 15 -105% | 5.9 19% | 6.4 12% | 9.2 -26% |
Idle Maximum * | 7.5 | 8.4 -12% | 16 -113% | 6 20% | 14.7 -96% | 11.8 -57% |
Load Average * | 75 | 72.7 3% | 80.4 -7% | 68.7 8% | 55.5 26% | 45 40% |
Cyberpunk 2077 ultra external monitor * | 83 | 86.2 -4% | 78.2 6% | 59.1 29% | ||
Load Maximum * | 99.9 | 99.1 1% | 89.5 10% | 82 18% | 103.3 -3% | 97.2 3% |
Witcher 3 ultra * | 80.5 |
* ... smaller is better
Power Consumption Cyberpunk / Stress Test
Power Consumption external Monitor
笔记本检查评级
BOSGAME M2
- 01/27/2025 v8
Allen Ngo
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. We never accept compensation or payment in return for our reviews. 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.