Not enough memory: Why the Nvidia RTX GPU is the main weakness of the Lenovo ThinkPad P1 G9

When looking at the laptop market nowadays, dedicated graphics processing units (GPUs) have become a bit of rarity. This is easy to prove: The German price-comparison site Geizhals.de lists a total of 6718 laptops, but only 1739 of these have a dedicated GPU. In general, laptops with dedicated GPUs deliver a higher GPU performance, so in general, having one should be an advantage, even though a dedicated GPU instead of an integrated GPU typically also means a higher power consumption.

There is, however, a catch: A dedicated GPU relies on its dedicated video memory (VRAM) instead of the system RAM. This used to be an advantage, but with many newer architectures, like Apple's M series or Nvidia's own RTX Spark, the system memory is unified, which means their integrated GPU can access far more memory than a typical dGPU.

Indeed, many dGPUs are still sadly limited when it comes to the amount of VRAM. And this hits one of our recently reviewed laptops especially hard: The Lenovo ThinkPad P1 Gen 9. This is a premium 16-inch workstation laptop for professional use, such as CAD rendering. It fastest dGPU is the Nvidia RTX Pro 2000 Blackwell - a laptop dGPU that is limited to 8 GB of VRAM.

Only up to the Nvidia RTX 2000 Pro - a too limited dGPU

For a laptop that costs thousands of dollars this is a big problem - 8 GB VRAM simply do not cut it for demanding applications, like running local AI. Older models of the P1, like the ThinkPad P1 Gen 6, had dGPUs with 16 GB of VRAM, as the P1 used to offer up to the RTX 5000 series of GPUs. The P1 Gen 9 is not even available with the RTX Pro 3000 Blackwell, which would at least include up to 12 GB of VRAM.

As it is, the dedicated dGPU of the ThinkPad P1 Gen 9 is an advantage for certain applications, but just too weak for others, which one would normally expect an expensive workstation like the P1 Gen 9 would be able to handle.