Ten months after the first RDNA 3-powered laptop GPU, AMD is finally bringing its pinnacle Navi 31 processor to mobile configurations. The new Radeon RX 7900M is advertised to be 7% faster on average than Nvidia's RTX 4080 Laptop, so high-end gaming fans on the go may be all over Team Red. [Specs-wise, it's a bit of a slower Radeon RX 7900 GRE; like all models in the Radeon RX 7900 series that use the RDNA 3-based Navi 31, this new GPU is a medium-sized chunk of silicon with all the graphics circuitry (called the GCD), L3 cache and memory interface (MCD), surrounded by a number of smaller chiplets.

The only other thing they have in common is that the MCD is almost identical. In fact, the only difference is the number of active compute units (CUs) and the clock speeds used; in the case of the RX 7900M, there are 72 CUs and the game clock is 1,825 MHz. For reference, the RX 7900 GRE has 80 CUs and a game clock of 1,880 MHz.

The RX 7900M has a boost clock of "up to 2,090 MHz" because that is the value set by the notebook vendor. The maximum clock speed is limited by the power the GPU can consume, and in the case of this product, the specifications list a total graphics power of "up to 180W". Therefore, if the laptop's upper limit is lower than this, the boost clock will be lower than the AMD value.

Another point in common with GRE cards is the number of MCDs. This, along with the total width of the memory bus and the amount of L3 Infinity Cache, determines how much VRAM the GPU can access; in the case of the RX 7900M, it is 16GB of GDDR6, all running at 16Gbps (576GB/sec bandwidth), with a 76MB cache.

For me, the best thing about all the RX 7900 series GPUs is that they all have the same number of shader engines (SEs). This is like an apartment with all the necessary features, which for a graphics chip means one that manages threads, rasterizes triangles, crunches shaders, and performs the rendering process.

Since the RX 7900 has six SEs, every RX 7900 also has 192 ROPs (render operation units), but the RX 7900M has the fewest number of shaders, so the GPU is rarely ROP-limited in games This is usually not a major issue. Normally this is not a major issue, but it is a significant increase in ROP count compared to Nvidia's RTX 4080 Laptop, which has 80 ROPs.

Having more ROPs is good for high-resolution games, especially at high frame rates.

By raw metrics, the RX 7900M is definitely better than the RTX 4080 Laptop, but it has to be; Nvidia uses the AD104 for this model, which is a smaller GPU than the Navi 31 even in this truncated form. It also has a lower peak power limit of 150W, but that doesn't matter much if the RX 7900M laptop costs as much as the RTX 4080 laptop.

The first laptop with an RX 7900M was the Alienware m18. It is only available in certain configurations using an AMD Ryzen 9 7945HX, 32GB of DDR5-5200, and 1TB of storage, priced at $2800, $50 less than a configuration using a Ryzen 9 7845HX and RTX 4080 Laptop GPU.

If you don't care about using ray tracing in your games, the RX 7900M configuration is superior. However, if you want the best performance with the latest rendering techniques, you should opt for the RTX 4080.

Personally, I would still choose AMD's new GPU, as I think the RX 7900M is a better match for the 18" 1920 x 1200, 480 Hz screen used in the Alienware m18 laptop, and for pure gaming the RX 7900M looks very good. Since $2,800 is a lot of money for a laptop, though, it's best to wait for reviews and benchmarks before committing to a purchase.

I am happy that we can finally get a Navi 31 GPU in a laptop, but 10 months is too long to bring this to market. Nvidia has been launching its entire mobile RTX 40 series since the beginning of this year, and finding a laptop with a decent AMD GPU is hard enough as it is. Many vendors are producing Nvidia-powered gaming laptops, and discounts are frequent.

If Dell does not join many others in adopting the RX 7900M, AMD's excellent graphics processors are unlikely to see the light of day.