Community member Robman86 compared the similarly equipped Razer Blade 18 and Asus ROG Strix Scar 18 models in a review of the GeForce RTX 4090 laptop GPU. The main focus of his analysis was on the CPU, because the or its TDP quickly makes the difference in the notebook.
Besides the GPU also counts the TDP of the CPU
Like the MSI GT77HX Titan 13V tested by the editors, the Razer Blade 18 and Asus ROG Strix Scar 18 rely on a GeForce RTX 4090 laptop GPU, which can draw on the maximum power budget of 150 watts plus up to 25 watts of Dynamic Boost 2.0. In games, however, all three are by no means equally fast if the CPU makes the difference – and that makes them faster in the notebook than in the desktop PC, as Robman86 shows.
< p class="p text-width">The most interesting thing for me was how much power consumption the (strong) CPUs can/may dissipate during dual load in a tolerable (4X dbA) fan profile (“Balanced” or similar). And how much in the maximum loud fan profile (because some/many use headsets, this is probably used more often than you think). Because a GPU limit can be ironed out relatively quickly (e.g. reduction of details or FSR/DLSS level), but a CPU limit is really bad, the reduction of details is usually of little use, and a CPU limit has it really very bad frame times.
Robman86
On paper high-end CPUs
The Blade 18 has an Intel Core i9-13950HX and 32 GB DDR5-5600, the Scar 18 has an Intel Core i9-13980HX with a slower but more generous 64 GB DDR5-4800, while the editing pattern of the MSI notebook (different than the model in stores) also uses a Core i9-13950HX, but only operates it with DDR5-4000 (64 GB) in accordance with the specification. On paper, all three notebooks are very close in terms of CPUs, and the Blade 18 has clear advantages over the MSI GT77HX Titan when it comes to memory – the Scar is right in the middle.
For comparison, Robman86 used his desktop PC with Intel Core i9-13900K and RTX 4080 and the older Alienware Area 51m R2 gaming notebook with desktop CPU (Intel Core i9-10900K) and GeForce RTX 2080 Mobile. At first glance, this seems like a mistake considering the very different graphics cards (although the 4090 laptop GPU and 4080 desktop are based on the same chip, the TDP and memory are not comparable), but Robman86 ended up addressing the CPU impact in Play. To isolate him, he chose a Far Cry 6 savegame that focuses extremely on the CPU – the game already does it.
The Razer Blade 18 was first run in the “Balanced” profile (not the louder gaming profile), the Asus ROG Strix Scar 18 in the “Performance” profile (not the louder “Turbo”). The results after a “warm up” period were as follows:
- Alienware Area 51m R2 (Intel Core i9-10900K; 125 W): 92 FPS
- Razer Blade 18 (Intel Core i9-13950HX; 45 W): 103 FPS
- Asus Scar G18 (Intel Core i9-13980HX; 55 W): 118 FPS
- Mini-ITX desktop (Intel Core i9-13900K; 105 W): 161 FPS
The difference to the desktop processor, which on paper clocks hardly any higher, is striking, but not surprising in view of the power loss: Robman86 was able to observe after a short time that the CPU power limit “PL1” to 45 W on the Razer and 55 W in the Asus notebook and thus the frame rate decreases, while the desktop CPU consumes approx. 105 watts in the test sequence without throttling – and the three generations old Comet Lake CPU in the Alienware notebook is at an impressive 125 watts not that far behind.
What particularly annoys Robman86: Although the CPUs in the new notebooks are clearly throttled, the volume is high – the CPUs are not yet throttled due to the temperature, the CPU temperature was high but not yet critically high. The situation is different with the quieter model from Alienware, in which the CPU was allowed to run up to the temperature limit.
Robman86 then chose the loudest profile of the notebooks (Razer: Gaming, Asus: Turbo), which enabled both notebooks to operate their CPUs at over 100 W for a short time and later to supply them with 65 watts permanently.
- Alienware Area 51m R2 (Intel Core i9-10900K ; 125 W): 92 FPS
- Razer Blade 18 (Intel Core i9-13950HX; 65 W): 110 FPS
- Asus Scar G18 (Intel Core i9-13980HX; 65 W) : 118 fps
- Mini-ITX desktop (Intel Core i9-13900K; 105 W): 161 FPS
Not quite, because the frames per second increased only in the case of the Razer Blade 18 by seven FPS, the Asus Scar G18 still reaches 118 FPS. The background noise increased significantly.
What brought more was to reduce the power consumption of the GPU. At less than 70 watts (set in Afterburner), the CPUs in the notebooks from Asus and Razer were also allowed to draw 90 to 100 watts permanently, but then the GPU became a bottleneck in this scene. And at over 80 watts, the CPUs were immediately capped again at 65 watts.
Gaming notebooks are not gaming PCs
< p class="p text-width">Robman86 has impressively and vividly documented the topic of “power sharing” in gaming notebooks, which was also repeatedly addressed by the editors. The CPU and GPU share a chassis with a cooling system. In order not to overload this, OEMs keep the combined power consumption of the CPU and GPU under control, the GPU usually has priority and the CPU lags behind. However, this lowers the CPU limit, in significantly more games than on desktop PCs the processor is suddenly the clock generator, not the GPU – simply because it clocks much lower.
The problem is exacerbated by the fact that many OEMs are now preventing the CPU from running within the temperature limit. As a result, a gaming notebook with a slim chassis that costs several thousand euros and whose GPU and CPU are hardly inferior to a desktop PC on paper scratches the 60 FPS limit in certain scenes.
This is nothing for Robman86 and neither are alternative gaming notebooks with a larger chassis and even louder cooler. In the end, every customer has to know that for himself. However, knowing the disadvantages of using powerful CPUs and GPUs in a compact chassis can only help buyers with this decision – and Robman86 wrote down exactly that knowledge.
More impressions on Blade 16 and Scar 18
In addition to this detour on the performance of gaming notebooks and their design-related limitations compared to gaming PCs, Robman86 has also summarized many other impressions of the two notebooks from Razer and Asus in his post.
Give me the hints!
In the category “From the community” the editors keep adding interesting reader articles from the ComputerBase forum the center.
The following reader articles from the ComputerBase forum have recently secured a place on the start page:
- Zombie Boards and Co: Masterchief79's graphics card modding and repair blog
- AMD Radeon RX 6900 XT optimized for efficiency
- GPU coolant: WLP, cheese , potatoes and ketchup in comparison
- In search of the strongest AGP platform
- The 3dfx T-Buffer
- Steam Deck Field reports
- Do-it-yourself Airbus A320 cockpit
- Dual Xeon CPU versus Ryzen 1700X
CB radio podcast episode #8: If the graphics memory should simply not grow