Verdict - AMD version is the worse choice overall
MSI's Claw is one of the few gaming handhelds with an Intel processor, offering a really good, albeit expensive, overall package thanks to efficient Lunar Lake processors. However, we don't really understand why an AMD-powered Claw model is now being launched on the market. The AMD version offers no real added value, especially since the new case is less ergonomic, and the reduction in the size of the shoulder buttons and triggers does not really make sense.
In general, we are surprised that MSI uses a different case. The development costs would have been better spent on software optimization, as there is a need to catch up, especially with the energy profiles. In manual mode, it is now possible to freely adjust the CPU parameters, but the previous sub-profiles with sensible TDP gradations have simply been removed. If you don't want to experiment too much yourself, you'll have to make do with the AI Engine mode.
In terms of price, the two variants are on the same level at 999 euros, but the AMD model only has 24 GB RAM instead of 32 GB for the Intel variant. This also makes the Claw A8 100 euros more expensive than the current Asus ROG Xbox Ally X with the same hardware (but smaller 7-inch display), which offers advantages above all in terms of comfort and energy profiles. For those who prefer the larger 8-inch screen, we recommend the Intel version of the Claw.
Pros
Cons
Price and availability
The MSI Claw A8 is available from online retailers such as Notebooksbilliger for an RRP of 999 euros (approximately US$1173).
Table of Contents
MSI offers the Claw 8 AI+, one of the few Intel gaming handhelds on the market, and is now expanding its portfolio with an AMD model. The Claw A8 is based on the AMD Ryzen Z2 Extreme, which includes the Radeon 890M, and even uses a new case. In addition to the familiar 8-inch IPS display, 24 GB of RAM, and a 1 TB SSD are also used. The RRP is still €999 and is therefore identical to the Intel model, but also €100 more expensive than the Asus ROG Xbox Ally X with identical features (but a smaller 7-inch display).
Specifications
Case & equipment
MSI has decided to equip the AMD Claw model with a completely new case. This is much more angular and is available in two colors, white and neon green, with a dark grey back. Visually, the white model does not look as high-quality as the silver version of the Intel model (although a white version is now available as well). The case is made of plastic, and the workmanship and stability are also good, but the ergonomics with the more angular handles are slightly worse; the Claw A8 is simply not quite as comfortable as the Intel version during longer sessions.
However, there are no differences between the two versions in terms of features: In addition to the two Thunderbolt 4 ports, there is a fast microSD card reader and a fast Wi-Fi 7 module. One advantage of the AMD version, however, is its internal layout: the cooling has been arranged differently, leaving space for a large M.2 2280 SSD.
| SD Card Reader | |
| average JPG Copy Test (av. of 3 runs) | |
| MSI Claw 8 AI+ (Angelbird AV Pro V60) | |
| MSI Claw A8 BZ2EM (Angelbird AV Pro V60) | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| Asus ROG Xbox Ally X (Angelbird AV Pro V60) | |
| Average of class Handheld / UMPC (20.9 - 201, n=13, last 2 years) | |
| GPD Pocket 4 (AV Pro SD microSD 128 GB V60) | |
| maximum AS SSD Seq Read Test (1GB) | |
| Asus ROG Xbox Ally X (Angelbird AV Pro V60) | |
| MSI Claw 8 AI+ (Angelbird AV Pro V60) | |
| MSI Claw A8 BZ2EM (Angelbird AV Pro V60) | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (74.7 - 261, n=12, last 2 years) | |
| Lenovo Legion Go | |
| GPD Pocket 4 (AV Pro SD microSD 128 GB V60) | |
| Networking | |
| MSI Claw A8 BZ2EM | |
| iperf3 transmit AXE11000 | |
| iperf3 receive AXE11000 | |
| Asus ROG Xbox Ally X | |
| iperf3 transmit AXE11000 | |
| iperf3 receive AXE11000 | |
| GPD Pocket 4 | |
| iperf3 transmit AXE11000 | |
| iperf3 receive AXE11000 | |
| iperf3 transmit AXE11000 6GHz | |
| iperf3 receive AXE11000 6GHz | |
| Zotac ZGC-G1A1W-01 | |
| iperf3 transmit AXE11000 | |
| iperf3 receive AXE11000 | |
| iperf3 transmit AXE11000 6GHz | |
| iperf3 receive AXE11000 6GHz | |
| MSI Claw 8 AI+ | |
| iperf3 transmit AXE11000 | |
| iperf3 receive AXE11000 | |
| Lenovo Legion Go | |
| iperf3 transmit AXE11000 | |
| iperf3 receive AXE11000 | |
| iperf3 transmit AXE11000 6GHz | |
| iperf3 receive AXE11000 6GHz | |
Sustainability
MSI does not provide any information on sustainability, the use of recycled materials, or environmental certifications. The packaging is mainly made of cardboard and is recyclable.
Software
MSI equips the Claw A8 with the well-known M-Center software. In addition, you can now launch directly into Xbox full-screen mode, which we already reported on in our review of the Asus ROG Xbox Ally X. In addition, the most important functions are now integrated into the Xbox Game Bar and can be accessed at any time with a single tap.
Input devices - Hall effect triggers & sticks
As with the Claw 8 AI+, the AMD model also uses very precise Hall-effect analog sticks and triggers. However, we are not entirely satisfied with the ergonomics here, as both the shoulder buttons and the triggers are narrower than on the Intel version. Of course, it depends a little on the size of your hands, but for us, it wasn't as comfortable to use, and the finger position wasn't natural. The shoulder buttons are very "clicky," quite loud, and precise. The option buttons next to the screen, on the other hand, feel rather cheap and are also very loud.
Display - IPS with 120 Hz & VRR
The 8-inch IPS screen (1,920 x 1,200 pixels) with 120 Hz and VRR does not differ from the Intel model. The overall picture quality is still very good, but the relatively slow response times remain a weak point. PWM flickering is not present.
Our analysis with the professional CalMAN software shows that the factory calibration again exhibits a blue cast, along with a color temperature that is too cool. This benefits the maximum brightness, which averages ~530 cd/m². With our own calibration (the profile is available for free download in the box below), we can significantly improve the display, but the brightness is about 50 cd/m² lower.
| |||||||||||||||||||||||||
Brightness Distribution: 90 %
Center on Battery: 540 cd/m²
Contrast: 1588:1 (Black: 0.34 cd/m²)
ΔE ColorChecker Calman: 2.9 | ∀{0.5-29.43 Ø4.78}
calibrated: 1.3
ΔE Greyscale Calman: 4.7 | ∀{0.09-98 Ø5}
73% AdobeRGB 1998 (Argyll 3D)
99.8% sRGB (Argyll 3D)
71.1% Display P3 (Argyll 3D)
Gamma: 2.18
CCT: 7507 K
| MSI Claw A8 BZ2EM PN8007QB1-2, IPS, 1920x1200, 8", 120 Hz | Asus ROG Xbox Ally X TL070FVXS01_0, IPS, 1920x1080, 7", 120 Hz | GPD Pocket 4 YHB03P24, IPS, 2560x1600, 8.8", 144 Hz | Zotac ZGC-G1A1W-01 ZDZ0501, OLED, 1920x1080, 7", 120 Hz | MSI Claw 8 AI+ PN8007QB1-2, IPS, 1920x1200, 8", 120 Hz | Lenovo Legion Go Legion Go Display, IPS, 2560x1600, 8.8", 144 Hz | |
|---|---|---|---|---|---|---|
| Display | -7% | 18% | 26% | -0% | 18% | |
| Display P3 Coverage (%) | 71.1 | 66 -7% | 92.7 30% | 99.43 40% | 71.1 0% | 92.1 30% |
| sRGB Coverage (%) | 99.8 | 93.1 -7% | 99.6 0% | 99.93 0% | 98.5 -1% | 99.56 0% |
| AdobeRGB 1998 Coverage (%) | 73 | 68.3 -6% | 90 23% | 99.89 37% | 72.9 0% | 90.09 23% |
| Response Times | 49% | -16% | 62% | 1% | 8% | |
| Response Time Grey 50% / Grey 80% * (ms) | 28.1 ? | 11.4 ? 59% | 27.9 ? 1% | 1.08 ? 96% | 23.7 ? 16% | 20.8 ? 26% |
| Response Time Black / White * (ms) | 13 ? | 8 ? 38% | 17.2 ? -32% | 9.3 ? 28% | 14.9 ? -15% | 14.4 ? -11% |
| PWM Frequency (Hz) | 120 ? | 700 | ||||
| PWM Amplitude * (%) | 10 | |||||
| Screen | -4% | -18% | 45% | -7% | -38% | |
| Brightness middle (cd/m²) | 540 | 490 -9% | 511.2 -5% | 833 54% | 524 -3% | 411 -24% |
| Brightness (cd/m²) | 530 | 459 -13% | 524 -1% | 830 57% | 514 -3% | 413 -22% |
| Brightness Distribution (%) | 90 | 86 -4% | 84 -7% | 90 0% | 93 3% | 80 -11% |
| Black Level * (cd/m²) | 0.34 | 0.33 3% | 0.4 -18% | 0.064 81% | 0.33 3% | 0.42 -24% |
| Contrast (:1) | 1588 | 1485 -6% | 1278 -20% | 13016 720% | 1588 0% | 979 -38% |
| Colorchecker dE 2000 * | 2.9 | 2.8 3% | 4.54 -57% | 6.7 -131% | 3.5 -21% | 5.94 -105% |
| Colorchecker dE 2000 max. * | 6.1 | 6.5 -7% | 8.96 -47% | 12.04 -97% | 7.3 -20% | 10.56 -73% |
| Colorchecker dE 2000 calibrated * | 1.3 | 1.5 -15% | 0.95 27% | 2.95 -127% | 1.3 -0% | 2.95 -127% |
| Greyscale dE 2000 * | 4.7 | 4.2 11% | 6.4 -36% | 11.99 -155% | 5.9 -26% | 0.92 80% |
| Gamma | 2.18 101% | 2.22 99% | 2.26 97% | 2.213 99% | 2.22 99% | 1.93 114% |
| CCT | 7507 87% | 7381 88% | 8125 80% | 8033 81% | 7852 83% | 8064 81% |
| Total Average (Program / Settings) | 13% /
3% | -5% /
-10% | 44% /
43% | -2% /
-5% | -4% /
-20% |
* ... smaller is better
Display Response Times
| ↔ Response Time Black to White | ||
|---|---|---|
| 13 ms ... rise ↗ and fall ↘ combined | ↗ 5.1 ms rise | |
| ↘ 7.9 ms fall | ||
| The screen shows good response rates in our tests, but may be too slow for competitive gamers. In comparison, all tested devices range from 0.1 (minimum) to 240 (maximum) ms. » 32 % of all devices are better. This means that the measured response time is better than the average of all tested devices (20.2 ms). | ||
| ↔ Response Time 50% Grey to 80% Grey | ||
| 28.1 ms ... rise ↗ and fall ↘ combined | ↗ 13.3 ms rise | |
| ↘ 14.8 ms fall | ||
| The screen shows relatively slow response rates in our tests and may be too slow for gamers. In comparison, all tested devices range from 0.165 (minimum) to 636 (maximum) ms. » 37 % of all devices are better. This means that the measured response time is better than the average of all tested devices (31.6 ms). | ||
Screen Flickering / PWM (Pulse-Width Modulation)
| Screen flickering / PWM not detected | |||
In comparison: 53 % of all tested devices do not use PWM to dim the display. If PWM was detected, an average of 8111 (minimum: 5 - maximum: 343500) Hz was measured. | |||
Performance - AMD handheld with 24 GB RAM
MSI equips the AMD version of the Claw with 24 GB of RAM, while the Intel model has 32 GB for the same price.
Test conditions
MSI still offers the three energy modes: Runtime, AI Engine, and Manual. While the Intel model still had three predefined sub-profiles for manual mode (8, 17 or 30 watts), these are no longer available for the AMD model. Instead, the three parameters SPL, SPPT & FPPT can be controlled individually. This is also good in principle, but we don't understand why the previous sub-profiles were removed, as there is now no quick-change option. We therefore carried out the benchmarks and measurements in AI mode, which is activated by default. However, MSI should address this.
Processor
The AMD Ryzen Z2 Extreme offers a total of 8 cores (3x Zen 5 & 5x Zen 5c), and in AI mode, the processor can briefly consume up to 50 watts, but then quickly levels out at just 20 watts. The pure CPU performance is therefore somewhat weaker than that of the competitors. To be honest, this is not so important when gaming, and if you need more CPU power, you can switch to manual mode. The performance remains constant in battery mode.
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 | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Lenovo Legion Go -2! | |
| Average AMD Ryzen Z2 Extreme | |
| Zotac ZGC-G1A1W-01 | |
| MSI Claw 8 AI+ | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC | |
| Cinebench R23 / Multi Core | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (12151 - 14586, n=2) | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC (3984 - 16841, n=17, last 2 years) | |
| MSI Claw 8 AI+ | |
| Cinebench R23 / Single Core | |
| GPD Pocket 4 | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme (1954 - 1970, n=2) | |
| Asus ROG Xbox Ally X | |
| MSI Claw 8 AI+ | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (1071 - 2006, n=16, last 2 years) | |
| Cinebench R20 / CPU (Multi Core) | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (4726 - 5825, n=2) | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC (1521 - 6577, n=15, last 2 years) | |
| MSI Claw 8 AI+ | |
| Cinebench R20 / CPU (Single Core) | |
| GPD Pocket 4 | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme (764 - 772, n=2) | |
| Asus ROG Xbox Ally X | |
| MSI Claw 8 AI+ | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (357 - 787, n=15, last 2 years) | |
| Cinebench R15 / CPU Multi 64Bit | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (2070 - 2261, n=2) | |
| MSI Claw A8 BZ2EM | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| Average of class Handheld / UMPC (689 - 2533, n=15, last 2 years) | |
| MSI Claw 8 AI+ | |
| Cinebench R15 / CPU Single 64Bit | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| Asus ROG Xbox Ally X | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (155.1 - 304, n=14, last 2 years) | |
| Average AMD Ryzen Z2 Extreme (190.7 - 275, n=2) | |
| MSI Claw A8 BZ2EM | |
| Blender / v2.79 BMW27 CPU | |
| MSI Claw 8 AI+ | |
| Average of class Handheld / UMPC (189 - 588, n=14, last 2 years) | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme (234 - 296, n=2) | |
| Zotac ZGC-G1A1W-01 | |
| Asus ROG Xbox Ally X | |
| GPD Pocket 4 | |
| 7-Zip 18.03 / 7z b 4 | |
| GPD Pocket 4 | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (45321 - 52461, n=2) | |
| Average of class Handheld / UMPC (19684 - 66618, n=14, last 2 years) | |
| MSI Claw A8 BZ2EM | |
| MSI Claw 8 AI+ | |
| 7-Zip 18.03 / 7z b 4 -mmt1 | |
| GPD Pocket 4 | |
| Zotac ZGC-G1A1W-01 | |
| MSI Claw 8 AI+ | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (5434 - 5590, n=2) | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC (3653 - 6143, n=15, last 2 years) | |
| Geekbench 6.5 / Multi-Core | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (10979 - 12734, n=3) | |
| Lenovo Legion Go | |
| Lenovo Legion Go | |
| MSI Claw A8 BZ2EM | |
| MSI Claw 8 AI+ | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (3950 - 13426, n=16, last 2 years) | |
| Geekbench 6.5 / Single-Core | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| MSI Claw 8 AI+ | |
| Lenovo Legion Go | |
| Lenovo Legion Go | |
| Average AMD Ryzen Z2 Extreme (1888 - 2790, n=3) | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (1333 - 2904, n=16, last 2 years) | |
| MSI Claw A8 BZ2EM | |
| Geekbench 5.5 / Multi-Core | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (10464 - 11222, n=2) | |
| MSI Claw A8 BZ2EM | |
| Lenovo Legion Go | |
| Lenovo Legion Go | |
| MSI Claw 8 AI+ | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (3932 - 11975, n=16, last 2 years) | |
| Geekbench 5.5 / Single-Core | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| MSI Claw 8 AI+ | |
| Lenovo Legion Go | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Average AMD Ryzen Z2 Extreme (1323 - 2017, n=2) | |
| Average of class Handheld / UMPC (875 - 2134, n=16, last 2 years) | |
| MSI Claw A8 BZ2EM | |
| HWBOT x265 Benchmark v2.2 / 4k Preset | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Lenovo Legion Go | |
| Average AMD Ryzen Z2 Extreme (14.7 - 17.9, n=2) | |
| Zotac ZGC-G1A1W-01 | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC (6.21 - 20.8, n=14, last 2 years) | |
| MSI Claw 8 AI+ | |
| LibreOffice / 20 Documents To PDF | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (57.4 - 68.7, n=2) | |
| Average of class Handheld / UMPC (41.1 - 92.8, n=14, last 2 years) | |
| MSI Claw A8 BZ2EM | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| R Benchmark 2.5 / Overall mean | |
| Average of class Handheld / UMPC (0.4167 - 0.756, n=14, last 2 years) | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| Asus ROG Xbox Ally X | |
| MSI Claw 8 AI+ | |
| Average AMD Ryzen Z2 Extreme (0.4395 - 0.452, n=2) | |
| MSI Claw A8 BZ2EM | |
| GPD Pocket 4 | |
* ... 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 | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Zotac ZGC-G1A1W-01 | |
| Average AMD Ryzen Z2 Extreme | |
| Lenovo Legion Go | |
| Average of class Handheld / UMPC | |
| MSI Claw A8 BZ2EM -1! | |
| MSI Claw 8 AI+ | |
| AIDA64 / FP32 Ray-Trace | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Average AMD Ryzen Z2 Extreme (17272 - 22536, n=2) | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC (6544 - 26450, n=14, last 2 years) | |
| MSI Claw 8 AI+ | |
| AIDA64 / FPU Julia | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Average AMD Ryzen Z2 Extreme (64269 - 88336, n=2) | |
| Average of class Handheld / UMPC (36464 - 103637, n=14, last 2 years) | |
| MSI Claw A8 BZ2EM | |
| MSI Claw 8 AI+ | |
| AIDA64 / CPU SHA3 | |
| GPD Pocket 4 | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (2946 - 3783, n=2) | |
| Average of class Handheld / UMPC (1157 - 4484, n=14, last 2 years) | |
| MSI Claw A8 BZ2EM | |
| MSI Claw 8 AI+ | |
| AIDA64 / CPU Queen | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| GPD Pocket 4 | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme (76357 - 77437, n=2) | |
| Asus ROG Xbox Ally X | |
| Average of class Handheld / UMPC (41744 - 107867, n=14, last 2 years) | |
| MSI Claw 8 AI+ | |
| AIDA64 / FPU SinJulia | |
| GPD Pocket 4 | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (11040 - 12594, n=2) | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC (4425 - 16980, n=14, last 2 years) | |
| MSI Claw 8 AI+ | |
| AIDA64 / FPU Mandel | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Average AMD Ryzen Z2 Extreme (33241 - 46842, n=2) | |
| Average of class Handheld / UMPC (19153 - 55457, n=14, last 2 years) | |
| MSI Claw A8 BZ2EM | |
| MSI Claw 8 AI+ | |
| AIDA64 / CPU AES | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (n=1) | |
| Average of class Handheld / UMPC (18552 - 149047, n=13, last 2 years) | |
| Zotac ZGC-G1A1W-01 | |
| GPD Pocket 4 | |
| Lenovo Legion Go | |
| MSI Claw 8 AI+ | |
| AIDA64 / CPU ZLib | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (721 - 990, n=2) | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (356 - 1112, n=14, last 2 years) | |
| MSI Claw A8 BZ2EM | |
| Lenovo Legion Go | |
| MSI Claw 8 AI+ | |
| AIDA64 / FP64 Ray-Trace | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Average AMD Ryzen Z2 Extreme (9159 - 12084, n=2) | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC (3585 - 14181, n=14, last 2 years) | |
| MSI Claw 8 AI+ | |
| AIDA64 / CPU PhotoWorxx | |
| MSI Claw 8 AI+ | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (52205 - 53205, n=2) | |
| MSI Claw A8 BZ2EM | |
| GPD Pocket 4 | |
| Average of class Handheld / UMPC (19472 - 54726, n=14, last 2 years) | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
System performance
There are no problems with the general system performance, and the performance in the synthetic benchmarks is also very good overall.
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 | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| Zotac ZGC-G1A1W-01 | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M | |
| Asus ROG Xbox Ally X | |
| Average of class Handheld / UMPC | |
| Lenovo Legion Go -5! | |
| PCMark 10 / Score | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M (7068 - 8082, n=2) | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| Asus ROG Xbox Ally X | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (3592 - 8082, n=14, last 2 years) | |
| Lenovo Legion Go | |
| PCMark 10 / Essentials | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M (10311 - 10390, n=2) | |
| Asus ROG Xbox Ally X | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| Average of class Handheld / UMPC (7542 - 11262, n=14, last 2 years) | |
| PCMark 10 / Productivity | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M (9567 - 14802, n=2) | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| MSI Claw 8 AI+ | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| Average of class Handheld / UMPC (5015 - 14802, n=14, last 2 years) | |
| PCMark 10 / Digital Content Creation | |
| MSI Claw 8 AI+ | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M (9317 - 9715, n=2) | |
| MSI Claw A8 BZ2EM | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (3326 - 10026, n=14, last 2 years) | |
| Lenovo Legion Go | |
| CrossMark / Overall | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| Zotac ZGC-G1A1W-01 | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M (1532 - 1596, n=2) | |
| Asus ROG Xbox Ally X | |
| Average of class Handheld / UMPC (840 - 1847, n=14, last 2 years) | |
| CrossMark / Productivity | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| Zotac ZGC-G1A1W-01 | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M (1453 - 1517, n=2) | |
| Asus ROG Xbox Ally X | |
| Average of class Handheld / UMPC (874 - 1719, n=14, last 2 years) | |
| CrossMark / Creativity | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| MSI Claw A8 BZ2EM | |
| Zotac ZGC-G1A1W-01 | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M (1712 - 1772, n=2) | |
| Asus ROG Xbox Ally X | |
| Average of class Handheld / UMPC (808 - 2077, n=14, last 2 years) | |
| CrossMark / Responsiveness | |
| GPD Pocket 4 | |
| Zotac ZGC-G1A1W-01 | |
| MSI Claw 8 AI+ | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M (1291 - 1359, n=2) | |
| Asus ROG Xbox Ally X | |
| Average of class Handheld / UMPC (759 - 1610, n=14, last 2 years) | |
| WebXPRT 3 / Overall | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| Asus ROG Xbox Ally X | |
| Average of class Handheld / UMPC (170.3 - 325, n=15, last 2 years) | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M (205 - 288, n=2) | |
| MSI Claw A8 BZ2EM | |
| WebXPRT 4 / Overall | |
| MSI Claw 8 AI+ | |
| GPD Pocket 4 | |
| Zotac ZGC-G1A1W-01 | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M (229 - 231, n=2) | |
| Asus ROG Xbox Ally X | |
| Average of class Handheld / UMPC (107 - 291, n=14, last 2 years) | |
| Mozilla Kraken 1.1 / Total | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme, AMD Radeon 890M (517 - 756, n=2) | |
| Average of class Handheld / UMPC (428 - 931, n=14, last 2 years) | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Asus ROG Xbox Ally X | |
| MSI Claw 8 AI+ | |
| GPD Pocket 4 | |
* ... smaller is better
| PCMark 10 Score | 8082 points | |
Help | ||
| AIDA64 / Memory Copy | |
| MSI Claw 8 AI+ | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme (92176 - 92683, n=2) | |
| Asus ROG Xbox Ally X | |
| GPD Pocket 4 | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (32617 - 107929, n=15, last 2 years) | |
| AIDA64 / Memory Read | |
| MSI Claw A8 BZ2EM | |
| Average AMD Ryzen Z2 Extreme (100379 - 101256, n=2) | |
| Asus ROG Xbox Ally X | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| Average of class Handheld / UMPC (25380 - 101256, n=15, last 2 years) | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| AIDA64 / Memory Write | |
| MSI Claw 8 AI+ | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (104064 - 104172, n=2) | |
| MSI Claw A8 BZ2EM | |
| GPD Pocket 4 | |
| Average of class Handheld / UMPC (25595 - 117883, n=15, last 2 years) | |
| AIDA64 / Memory Latency | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (99.8 - 147.6, n=15, last 2 years) | |
| Lenovo Legion Go | |
| Asus ROG Xbox Ally X | |
| Average AMD Ryzen Z2 Extreme (106.4 - 114, n=2) | |
| MSI Claw A8 BZ2EM | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
* ... smaller is better
DPC latencies
| DPC Latencies / LatencyMon - interrupt to process latency (max), Web, Youtube, Prime95 | |
| MSI Claw 8 AI+ | |
| Zotac ZGC-G1A1W-01 | |
| GPD Pocket 4 | |
| MSI Claw A8 BZ2EM | |
| Asus ROG Xbox Ally X | |
* ... smaller is better
Storage
The AMD model now offers space for a larger M.2 2280 SSD, which is particularly practical if you want to install a larger model. The built-in Micron 2500 SSD (PCIe 4.0) with 1 TB (887 GB free) delivers very high overall performance and remains completely stable even under prolonged load.
* ... smaller is better
Continuous read performance: DiskSpd Read Loop, Queue Depth 8
GPU performance
The Radeon 890M offers very good performance overall, although you can also tease out a bit more in manual mode. In gaming, the two variants of the MSI Claw are on a par, and it doesn't really make a difference whether you opt for the AMD or Intel variant. The gaming performance also remains very stable overall.
| 3DMark Performance rating - Percent | |
| Asus ROG Xbox Ally X | |
| Average of class Handheld / UMPC | |
| MSI Claw 8 AI+ | |
| Average AMD Radeon 890M | |
| MSI Claw A8 BZ2EM | |
| GPD Pocket 4 | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| 3DMark 11 Performance | 13403 points | |
| 3DMark Fire Strike Score | 8497 points | |
| 3DMark Time Spy Score | 3640 points | |
| 3DMark Steel Nomad Score | 597 points | |
| 3DMark Steel Nomad Light Score | 3228 points | |
Help | ||
| Blender / v3.3 Classroom CPU | |
| MSI Claw 8 AI+ | |
| Average of class Handheld / UMPC (290 - 1026, n=14, last 2 years) | |
| MSI Claw A8 BZ2EM | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| Asus ROG Xbox Ally X | |
| GPD Pocket 4 | |
| Average AMD Radeon 890M (214 - 475, n=20) | |
* ... smaller is better
| Performance rating - Percent | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| MSI Claw 8 AI+ -1! | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go -2! | |
| Cyberpunk 2077 - 1920x1080 Ultra Preset (FSR off) | |
| MSI Claw 8 AI+ | |
| Asus ROG Xbox Ally X | |
| GPD Pocket 4 | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC (11.5 - 91.4, n=21, last 2 years) | |
| Zotac ZGC-G1A1W-01 | |
| Baldur's Gate 3 - 1920x1080 Ultra Preset AA:T | |
| Asus ROG Xbox Ally X | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC (13.5 - 118.8, n=19, last 2 years) | |
| Zotac ZGC-G1A1W-01 | |
| GTA V - 1920x1080 Highest AA:4xMSAA + FX AF:16x | |
| GPD Pocket 4 | |
| MSI Claw A8 BZ2EM | |
| Asus ROG Xbox Ally X | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (14 - 32.8, n=15, last 2 years) | |
| Lenovo Legion Go | |
| Final Fantasy XV Benchmark - 1920x1080 High Quality | |
| Asus ROG Xbox Ally X | |
| MSI Claw 8 AI+ | |
| GPD Pocket 4 | |
| MSI Claw A8 BZ2EM | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| Average of class Handheld / UMPC (17.9 - 34.8, n=21, last 2 years) | |
| Strange Brigade - 1920x1080 ultra AA:ultra AF:16 | |
| Asus ROG Xbox Ally X | |
| MSI Claw 8 AI+ | |
| GPD Pocket 4 | |
| MSI Claw A8 BZ2EM | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (30 - 58.8, n=22, last 2 years) | |
| Lenovo Legion Go | |
| Dota 2 Reborn - 1920x1080 ultra (3/3) best looking | |
| GPD Pocket 4 | |
| MSI Claw 8 AI+ | |
| Lenovo Legion Go | |
| Asus ROG Xbox Ally X | |
| MSI Claw A8 BZ2EM | |
| Zotac ZGC-G1A1W-01 | |
| Average of class Handheld / UMPC (30.8 - 85, n=20, last 2 years) | |
| X-Plane 11.11 - 1920x1080 high (fps_test=3) | |
| GPD Pocket 4 | |
| Asus ROG Xbox Ally X | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC (22.7 - 50.5, n=20, last 2 years) | |
| Lenovo Legion Go | |
| Zotac ZGC-G1A1W-01 | |
| MSI Claw 8 AI+ | |
Cyberpunk 2077 ultra FPS chart
| low | med. | high | ultra | |
|---|---|---|---|---|
| GTA V (2015) | 150.2 | 132.9 | 73.9 | 32.7 |
| Dota 2 Reborn (2015) | 129.9 | 108.5 | 87.2 | 77.7 |
| Final Fantasy XV Benchmark (2018) | 83.3 | 44.2 | 32 | |
| X-Plane 11.11 (2018) | 57.1 | 47.8 | 38.2 | |
| Strange Brigade (2018) | 176.7 | 78.5 | 65.2 | 55.4 |
| Baldur's Gate 3 (2023) | 44.9 | 35.6 | 30.2 | 28.7 |
| Cyberpunk 2077 (2023) | 44.9 | 35.8 | 28.9 | 25 |
| F1 24 (2024) | 85.1 | 76 | 57.1 | 14.9 |
Emissions & Energy
Noise emissions
Compared to the Intel version, we could no longer detect fan pulsation under load. However, there is still an audible noise of around 29 dB(A) at idle. When gaming, we measured between 36-39 dB(A), depending on the title, which is not yet disturbing. There were no other electronic noises.
Noise level
| Idle |
| 24 / 24 / 29 dB(A) |
| Load |
| 39.1 / 39.1 dB(A) |
![]() | ||
30 dB silent 40 dB(A) audible 50 dB(A) loud |
||
min: | ||
| MSI Claw A8 BZ2EM Ryzen Z2 Extreme, Radeon 890M | Asus ROG Xbox Ally X Ryzen Z2 Extreme, Radeon 890M | GPD Pocket 4 Ryzen AI 9 HX 370, Radeon 890M | Zotac ZGC-G1A1W-01 R7 8840U, Radeon 780M | MSI Claw 8 AI+ Core Ultra 7 258V, Arc 140V | Lenovo Legion Go Z1 Extreme, Radeon 780M | |
|---|---|---|---|---|---|---|
| Noise | -6% | -10% | -8% | -3% | 6% | |
| off / environment * (dB) | 24 | 24.2 -1% | 22.8 5% | 25.6 -7% | 24.7 -3% | 23 4% |
| Idle Minimum * (dB) | 24 | 24.2 -1% | 25.8 -8% | 25.6 -7% | 24.7 -3% | 23 4% |
| Idle Average * (dB) | 24 | 24.2 -1% | 25.8 -8% | 28.4 -18% | 24.7 -3% | 23 4% |
| Idle Maximum * (dB) | 29 | 29 -0% | 31 -7% | 33.4 -15% | 28.9 -0% | 25 14% |
| Load Average * (dB) | 39.1 | 41.4 -6% | 47 -20% | 37.3 5% | 37 5% | |
| Cyberpunk 2077 ultra * (dB) | 36.6 | 43.4 -19% | 43.3 -18% | 37.6 -3% | 41.4 -13% | |
| Load Maximum * (dB) | 39.1 | 43.4 -11% | 45.1 -15% | 38.9 1% | 41.4 -6% | 37 5% |
| Witcher 3 ultra * (dB) | 46 |
* ... smaller is better
Temperature
(+) The maximum temperature on the upper side is 39.2 °C / 103 F, compared to the average of 37.8 °C / 100 F, ranging from 26.6 to 55 °C for the class Handheld / UMPC.
(+) The bottom heats up to a maximum of 37 °C / 99 F, compared to the average of 37.2 °C / 99 F
(+) In idle usage, the average temperature for the upper side is 29.4 °C / 85 F, compared to the device average of 32.3 °C / 90 F.
(±) 3: The average temperature for the upper side is 37.7 °C / 100 F, compared to the average of 32.3 °C / 90 F for the class Handheld / UMPC.
(±) The palmrests and touchpad can get very hot to the touch with a maximum of 39.5 °C / 103.1 F.
(-) The average temperature of the palmrest area of similar devices was 31.7 °C / 89.1 F (-7.8 °C / -14 F).
| MSI Claw A8 BZ2EM Ryzen Z2 Extreme, Radeon 890M | Asus ROG Xbox Ally X Ryzen Z2 Extreme, Radeon 890M | GPD Pocket 4 Ryzen AI 9 HX 370, Radeon 890M | Zotac ZGC-G1A1W-01 R7 8840U, Radeon 780M | MSI Claw 8 AI+ Core Ultra 7 258V, Arc 140V | Lenovo Legion Go Z1 Extreme, Radeon 780M | |
|---|---|---|---|---|---|---|
| Heat | -1% | 12% | -17% | 11% | ||
| Maximum Upper Side * (°C) | 39.2 | 40.8 -4% | 33.8 14% | 50 -28% | 36.6 7% | |
| Maximum Bottom * (°C) | 37 | 40.8 -10% | 35.2 5% | 50 -35% | 34.8 6% | |
| Idle Upper Side * (°C) | 30 | 29.9 -0% | 26.6 11% | 33 -10% | 27.5 8% | |
| Idle Bottom * (°C) | 34.9 | 30.6 12% | 28.6 18% | 33 5% | 27.2 22% |
* ... smaller is better
Speaker
MSI Claw A8 BZ2EM audio analysis
(+) | speakers can play relatively loud (83.6 dB)
Bass 100 - 315 Hz
(±) | reduced bass - on average 12% lower than median
(±) | linearity of bass is average (7.9% delta to prev. frequency)
Mids 400 - 2000 Hz
(+) | balanced mids - only 4.5% away from median
(+) | mids are linear (3.7% delta to prev. frequency)
Highs 2 - 16 kHz
(+) | balanced highs - only 3.1% away from median
(+) | highs are linear (2.9% delta to prev. frequency)
Overall 100 - 16.000 Hz
(+) | overall sound is linear (11.9% difference to median)
Compared to same class
» 0% of all tested devices in this class were better, 9% similar, 91% worse
» The best had a delta of 12%, average was 17%, worst was 23%
Compared to all devices tested
» 8% of all tested devices were better, 2% similar, 89% worse
» The best had a delta of 4%, average was 24%, worst was 134%
Asus ROG Xbox Ally X audio analysis
(+) | speakers can play relatively loud (86.6 dB)
Bass 100 - 315 Hz
(±) | reduced bass - on average 14.8% lower than median
(±) | linearity of bass is average (7.7% delta to prev. frequency)
Mids 400 - 2000 Hz
(±) | higher mids - on average 5.4% higher than median
(+) | mids are linear (3.5% delta to prev. frequency)
Highs 2 - 16 kHz
(+) | balanced highs - only 4% away from median
(+) | highs are linear (4.4% delta to prev. frequency)
Overall 100 - 16.000 Hz
(+) | overall sound is linear (12.8% difference to median)
Compared to same class
» 9% of all tested devices in this class were better, 14% similar, 77% worse
» The best had a delta of 12%, average was 17%, worst was 23%
Compared to all devices tested
» 11% of all tested devices were better, 3% similar, 86% worse
» The best had a delta of 4%, average was 24%, worst was 134%
MSI Claw 8 AI+ audio analysis
(+) | speakers can play relatively loud (85 dB)
Bass 100 - 315 Hz
(±) | reduced bass - on average 14.5% lower than median
(±) | linearity of bass is average (8.8% delta to prev. frequency)
Mids 400 - 2000 Hz
(+) | balanced mids - only 2.1% away from median
(+) | mids are linear (2.1% delta to prev. frequency)
Highs 2 - 16 kHz
(+) | balanced highs - only 2.4% away from median
(+) | highs are linear (4.2% delta to prev. frequency)
Overall 100 - 16.000 Hz
(+) | overall sound is linear (11.9% difference to median)
Compared to same class
» 0% of all tested devices in this class were better, 9% similar, 91% worse
» The best had a delta of 12%, average was 17%, worst was 23%
Compared to all devices tested
» 8% of all tested devices were better, 2% similar, 89% worse
» The best had a delta of 4%, average was 24%, worst was 134%
Power consumption
In idle mode, the AMD version requires slightly more power than the Intel model, so Lunar Lake has an advantage here. However, the AMD version is somewhat more economical in AI mode when gaming. We measure a maximum of 64 watts, which quickly levels off at around 49 watts, so the 65-watt power supply is sufficiently dimensioned overall. However, the standby consumption of 2.4 watts is clearly too high.
| Off / Standby | |
| Idle | |
| Load |
|
Key:
min: | |
| MSI Claw A8 BZ2EM Ryzen Z2 Extreme, Radeon 890M, Micron 2500 1TB MTFDKBA1T0QGN, IPS, 1920x1200, 8" | Asus ROG Xbox Ally X Ryzen Z2 Extreme, Radeon 890M, Samsung MZVMA1T0HCLD, IPS, 1920x1080, 7" | GPD Pocket 4 Ryzen AI 9 HX 370, Radeon 890M, , IPS, 2560x1600, 8.8" | Zotac ZGC-G1A1W-01 R7 8840U, Radeon 780M, Phison PS5021-E21 512GB, OLED, 1920x1080, 7" | MSI Claw 8 AI+ Core Ultra 7 258V, Arc 140V, Kioxia BG6 KBG60ZNS1T02, IPS, 1920x1200, 8" | Lenovo Legion Go Z1 Extreme, Radeon 780M, WD PC SN740 512GB SDDPMQD-512G-1101, IPS, 2560x1600, 8.8" | Average AMD Radeon 890M | Average of class Handheld / UMPC | |
|---|---|---|---|---|---|---|---|---|
| Power Consumption | -6% | -13% | -29% | 7% | -47% | -16% | -1% | |
| Idle Minimum * (Watt) | 5.3 | 3.7 30% | 6 -13% | 7.3 -38% | 3.2 40% | 11.5 -117% | 5.78 ? -9% | 6.01 ? -13% |
| Idle Average * (Watt) | 8.7 | 6.1 30% | 8.8 -1% | 8.3 5% | 6.3 28% | 12 -38% | 8 ? 8% | 8.26 ? 5% |
| Idle Maximum * (Watt) | 9.1 | 6.2 32% | 9.2 -1% | 12.1 -33% | 6.6 27% | 17.4 -91% | 9.55 ? -5% | 10.3 ? -13% |
| Load Average * (Watt) | 42.2 | 56.1 -33% | 53.1 -26% | 40.6 4% | 45 -7% | 64.5 ? -53% | 41.8 ? 1% | |
| Cyberpunk 2077 ultra * (Watt) | 35.5 | 54.5 -54% | 50.1 -41% | 51.2 -44% | 47 -32% | |||
| Cyberpunk 2077 ultra external monitor * (Watt) | 37.6 | 55 -46% | 45.7 -22% | 50.2 -34% | 46.7 -24% | |||
| Load Maximum * (Watt) | 64 | 66.4 -4% | 56.4 12% | 59.3 7% | 54 16% | 78.6 ? -23% | 54.7 ? 15% | |
| Witcher 3 ultra * (Watt) | 42.6 |
* ... smaller is better
Power consumption Cyberpunk / Stress test
Power consumption with external monitor
Battery life
The battery capacity remains unchanged at 80 Wh, and the runtimes are also very good. Depending on the TDP setting, you can expect between 1.5 - 3 hours when gaming, and significantly longer runtimes are possible, especially for emulated retro titles. In the WLAN test at 150 cd/m², we measured around 18.5 hours and 12 hours at full brightness. These are good values, but the Intel variant was even better.
| Battery Runtime | |
| WiFi Websurfing | |
| MSI Claw 8 AI+ | |
| MSI Claw A8 BZ2EM | |
| Asus ROG Xbox Ally X | |
| Average of class Handheld / UMPC (5.43 - 23.6, n=15, last 2 years) | |
| GPD Pocket 4 | |
| Zotac ZGC-G1A1W-01 | |
| Lenovo Legion Go | |
| WiFi Websurfing max. Brightness | |
| MSI Claw 8 AI+ | |
| Asus ROG Xbox Ally X | |
| MSI Claw A8 BZ2EM | |
| Average of class Handheld / UMPC (6.63 - 16.6, n=6, last 2 years) | |
| 1920x1080 Cyberpunk 2077 ultra 150cd | |
| MSI Claw A8 BZ2EM | |
| Asus ROG Xbox Ally X | |
| Average of class Handheld / UMPC (0.85 - 2.78, n=11, last 2 years) | |
| MSI Claw 8 AI+ | |
| Zotac ZGC-G1A1W-01 | |
Notebookcheck overall rating
MSI Claw A8 BZ2EM
- 12/13/2025 v8
Andreas Osthoff
Possible alternatives in comparison
Image | Model / Review | Price | Weight | Height | Display |
|---|---|---|---|---|---|
| MSI Claw A8 BZ2EM AMD Ryzen Z2 Extreme ⎘ AMD Radeon 890M ⎘ 24 GB Memory, 1024 GB SSD | Amazon: 1. $19.99 Hzycwgone Case for MSI Claw ... 2. $19.99 CDCHGAM Protective Case for ... 3. $18.99 JYGamhanbil Protective Case ... List Price: 999€ | 765 g | 24 mm | 8.00" 1920x1200 283 PPI IPS | |
| Asus ROG Xbox Ally X AMD Ryzen Z2 Extreme ⎘ AMD Radeon 890M ⎘ 24 GB Memory, 1024 GB SSD | Amazon: 1. $33.24 JSAUX Docking Station for RO... 2. $9.99 amFilm 3 Pack Screen Protect... 3. $6.99 TQLGY 2 Pack Tempered Glass ... List Price: 899€ | 718 g | 50.9 mm | 7.00" 1920x1080 315 PPI IPS | |
| GPD Pocket 4 AMD Ryzen AI 9 HX 370 ⎘ AMD Radeon 890M ⎘ 32 GB Memory | Amazon: List Price: 1400 USD | 777 g | 22.2 mm | 8.80" 2560x1600 343 PPI IPS | |
| Zotac ZGC-G1A1W-01 AMD Ryzen 7 8840U ⎘ AMD Radeon 780M ⎘ 16 GB Memory, 512 GB SSD | Amazon: 1. $12.98 Puccy 3 Pack Screen Protecto... 2. $25.99 PowerHOOD 65W USB-C AC/DC Ad... 3. $14.98 Puccy 2 Pack Anti Blue Light... | 692 g | 35 mm | 7.00" 1920x1080 315 PPI OLED | |
| MSI Claw 8 AI+ Intel Core Ultra 7 258V ⎘ Intel Arc Graphics 140V ⎘ 32 GB Memory, 512 GB SSD | Amazon: 1. $9.99 TQLGY 2 Pack Tempered Glass ... 2. $17.99 CDCHGAM Protector Cover Case... 3. $18.99 Spigen Tempered Glass Screen... List Price: 999€ | 793 g | 30 mm | 8.00" 1920x1200 283 PPI IPS | |
| Lenovo Legion Go AMD Ryzen Z1 Extreme ⎘ AMD Radeon 780M ⎘ 16 GB Memory, 512 GB SSD | Amazon: List Price: 800€ | 845 g | 41 mm | 8.80" 2560x1600 343 PPI IPS |
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.

















































