RTX 50 Mobile: Big Memory Upgrades for Laptops

Recent reports indicate that Nvidia may introduce the GeForce RTX 50 series as early as the fourth quarter of this year. Unlike the previous GeForce RTX 30/40 series, initially only the flagship GeForce RTX 5090 will be offered, with other models postponed until next year. Similar to data center products, the GeForce cards based on the Blackwell architecture will be manufactured using TSMC’s custom 4NP process, an improvement over the existing 4N process.

According to Moore’s Law is Dead, Nvidia plans to extend the GeForce RTX 50 series to mobile platforms as well. This iteration sees Nvidia upgrading the memory configuration of its mid-to-high-end mobile graphics cards, enhancing both the bandwidth and capacity.

Diverging from the strategy used with the RTX 4090 (AD103) and RTX 4080 (AD104), which employed different chips, both the mobile versions of the RTX 5080 and RTX 5090 will utilize the GB203 chip, paired with GDDR7 memory, each featuring a 256-bit memory bus and 16GB of memory. This represents a significant improvement over the RTX 4080 mobile card’s 192-bit and 12GB configuration.

The upgrade in memory configuration is not limited to the higher-tier RTX 5080; it also includes the RTX 5070 (or RTX 5070 Ti), which will see an increase to a 192-bit memory bus and 12GB of memory. Compared to the current RTX 4070’s 128-bit and 8GB, this is a notable improvement. However, for the mainstream RTX 5060 mobile card, Nvidia will maintain a 128-bit and 8GB setup, which might lead to performance bottlenecks in certain games.

Beyond performance, pricing remains a concern for gamers, especially since the GeForce RTX 40 series mobile graphics cards were significantly more expensive than their predecessors, and there is a clear hope that Nvidia will not continue to raise prices.