Nvidia has added a Pascal architecture at the lower end of their 1000 series through the GT-1030 entry graphics card, also known as MX150. This promises a significant performance gain compared to the previous generation. We compare the MX150 with the Maxwell-based 940MX of the previous generation to determine the extent to which entry-level gamers benefit from the new chip.
Introduction
Last year the GPU sector saw great progress with the new Pascal architecture. Because the most important part of designing the notebooks is portability, problems such as temperature development level prevented the use of more powerful GPUs. However, modern notebook GPUs deliver a good performance and can almost keep up with their desktop counterparts. In order to meet different target groups, there are notebooks with integrated graphics card such as Intel HD graphics and those with independent GPUs: AMD or Nvidia. It’s easy to decide between notebooks with integrated and independent graphics cards; Integrated graphic processors are suitable only for office use and very light gaming, but a dedicated video card can be used for workstation or gaming purpose. In this article will be compared two of the most popular standalone GPUs for ultrabooks to medium notebooks – the Nvidia GeForce MX150 and its predecessor, the Nvidia GeForce 940MX GDDR5. Both are similar in terms of price. But do really MX150 deliver so much more power than its predecessor, that an update could be justified?
Nvidia GeForce MX150
Nvidia’s Pascal architecture is successful. This is especially true for the top-end GPUs starting with the GeForce GTX 1050. It is likely that notebook GPUs achieve a performance comparable to their more expensive desktop counterparts, and Nvidia has decided to completely omit the “mobile” name. Notebooks with GTX 1050 and higher are more expensive and are aimed at creators of content and games. The MX150 (mobile GT 1030) strives to make the benefits of Pascal available to the masses who need ample power for 4K video playback and playing the latest AAA tracks with low settings. It claims to be an alternative to the Iris-Pro / Iris-Plus graphics chips integrated into the latest Sky Lake / Kaby Lake offerings and can potentially enable ultrabooks as well as budget laptops to the mentioned tasks.
The MX150 is based on the 16-nm GP108 chip and has a maximum of 2GB GDDR5-VRAM, a core clock of 1.468 MHz (boost clock: 1.532 MHz) and supports the APIs DirectX 12, Vulkan and OpenGL 4.5. The TDP is fixed at 25 watts, but OEMs can adjust the clock rates for special TDP or performance targets. The GP108 has 16 raster operation pipelines (ROPs) and the same number of 384 CUDA cores as the Maxwell-based 940MX (8 ROPs). Combined with the advantages of the Pascal architecture, there are significant improvements compared to the Maxwell predecessor GM108, at least on paper.
Nvidia GeForce 940MX with GDDR5
The GeForce 940MX is based on the older 28-nm Maxwell architecture. It is an updated GeForce 940M core with faster GDDR5-VRAM and higher clock rates. This GPU is perfect for light to medium graphics loads and can deal with games of the previous year with low settings pretty well. Content creators and developers can use it for a light workload.
There are variants of the 940MX-s, which depend largely on the OEM. The 940MX with GM107 chip has 512 CUDA cores, but lower clock rates (795 – 861 MHz) while the GM108 uses 384 CUDA cores at elevated clock rates (1,122 – 1,242 MHz). The 940MX supports vendor-specific Nvidia technologies such as GPU Boost 2.0, Optimus, PhysX and GameWorks. It can be equipped with up to 4 GB of VRAM and its TDP moves between 15 and 30 watts, depending on how the OEM adapts the clock rates. The GPU supports DirectX 12 up to FeatureLevel 11_0 and OpenGL 4.5 APIs.
The table below compares the specifications of the two GPUs:
Comparison of the Specifications between GeForce MX150 and GeForce 940MX
| GeForce MX150 | GeForce 940MX | |
|---|---|---|
| Codename | N17P-G1 | N16S-GTR-B/S |
| Architecture | Pascal 14 nm | Maxwell 28 nm |
| CUDA-Cores | 384 | 384 |
| Raster Operation Pipelines (ROPs) | 16 | 8 |
| Clock rate | 1.468 MHz (Base) 1.532 MHz (Boost) | 1.122 MHz (Base) 1.242 MHz (Boost) |
| Memory Bus Bandwidth | 64-bit | 64-bit |
| Memory-Type | GDDR5 | GDDR5/DDR3 |
| Max VRAM | 2.048 MB | 4.096 MB |
| API-Support | DirectX 12 (FL 12_1), OpenGL 4.5, Vulkan | DirectX 12 (FL 11_0), OpenGL 4.5 |
| VR-Ready | No | No |
| G-Sync-Support | Yes | Yes |
| TDP | ~ 25W | 15-30W |
The above comparison of the specifications shows that the MX150 brings some technological advantages of the Pascal architecture, such as more up-to-date API support and a lower 14 nm manufacturing process. This leads to lower / comparable TDPs with the same number of cores. The newer architecture uses higher clock rates and more ROPs. Compared to the previous generation, it should, therefore, deliver significantly higher performance and better benchmark results.
Benchmarks
To get a more realistic look at the performance, different well-known games were performed on both systems. Both, Rise of the Tomb Raider and The Witcher 3 are very GPU intensive and give a sense of what gamers can expect. The MX150 creates at least 20 FPS in both Tomb Raider and The Witcher 3 with low and medium settings. At high settings, the FPS of both cards is not high enough to play. Similar to the synthetic benchmarks, the performance gains of the MX150 are about 50 percent higher.
| Rise of the Tomb Raider | |
| 1920×1080 High Preset AA:FX AF:4x | |
| MX150 | |
| 940MX | |
| 1366×768 Medium Preset AF:2x | |
| MX150 | |
| 940MX | |
| 1024×768 Lowest Preset | |
| MX150 | |
| 940MX | |
| The Witcher 3 | |
| 1920×1080 High Graphics & Postprocessing (Nvidia HairWorks Off) | |
| MX150 | |
| 940MX | |
| 1366×768 Medium Graphics & Postprocessing | |
| MX150 | |
| 940MX | |
| 1024×768 Low Graphics & Postprocessing | |
| MX150 | |
| 940MX | |
According to some synthetic and gaming benchmarks, one thing is clear: the MX150 is clearly an improvement over its Maxwell predecessor and could soon be the GPU of choice for entry-level gaming notebooks. Benchmarks reflect Nvidia’s performance growth (about 33%). Although the percentage increase is clear, the actual noticeable difference is quite moderate. Considering that the MX150 has essentially the same number of CUDA cores and a similar VRAM bandwidth, it should not be surprising. What is gratifying is that the MX150 actually delivers a higher performance – despite similar TDP values like the 940MX. This clearly shows that the Pascal architecture is very efficient even with less powerful GPUs. Since OEMs can adjust the core clocks, it is expected that more energy-efficient notebooks will be equipped with MX150. The observable performance increase can be attributed to a higher number of ROPs, higher core clocks, and Pascal architecture advantages. For more information on how the two cards perform in other gaming benchmarks, please refer to our special pages on the MX150 and 940MX.
Looking at the gaming performance, the MX150 also seems to be well suited for professional applications such as video and 3D rendering. Your DirectCompute and OpenCL performance are equal to the Quadro M1000M. The Quadro has more CUDA cores and a higher memory bandwidth, but the MX150 benefits from higher core clocks and supports innovative APIs like DirectX 12. Therefore, the MX150 in combination with a suitable CPU could be attractive for prosumer with a tight budget.
Not to be ignored is that Nvidia offers these GPUs as alternatives to integrated GPU solutions that are found in most ultrabooks and entry-level laptops. Compared to such iGPU solutions, the advantages of the MX150 are quickly apparent.
Although the new generation delivers higher power, there is a downfall. Unlike the CPUs used for this comparison, ultrabooks and budget gaming laptops are now powered by U-Series CPUs with 15 watts of TDP. Thus, they are optimized to maximize battery life and minimize heat generation. OEMs can also reduce the clock rates of the MX150 to keep the temperatures low. Therefore, the actual performance of the GPU may be variable.
Is it worth upgrading to MX150?
Is it necessary to replace your existing 940MX laptop with one with MX150? Not really. Although the MX150 brings a fairly obvious performance gain and an improved architecture, this may not have an effect in everyday use. Anyone looking for a new Ultrabook or entry-level gaming notebook can definitely consider one with MX150. Although the MX150 is not specifically made for Workstation users, it is nice to know that performance increase in this area as well. We generally recommend that buyers avoid the older Maxwell GPUs as best they can. Maxwell GPUs have clearly exceeded their zenith and the newer Pascal chips offer better performance and – more importantly – higher energy efficiency. But what if they get an excellent offer for a 940MX laptop? Well, if it is irresistible and you are sure that the notebook (in the meantime) will meet its performance requirements, it could be a purchase value. The 940MX is by no means “out of date”, and Nvidia will still offer driver updates for some time. Keep in mind, however, that the MX150 is better in every respect.
GPUs like the MX150 offer more than any iGPUs, and it is gratifying that Nvidia is also extending its latest architecture improvements to its entry-level chips. The MX150 is in every respect better than its Maxwell counterpart 940MX. Although the MX150 fades compared to its more powerful relatives such as the GTX 1050 and better, it provides enough power for most common tasks, allowing some gaming to be done.