NVIDIA and AMD are releasing their next-gen GPUs by the end of this year. With months remaining on the GPUs’ launch, both the companies don’t operate with real SKUs name just yet. Instead, they use PCI ID, which may be exclusive to certain SKUs and hence help in tracking down upcoming GPUs.
The GPU IDs for the NVIDIA GeForce RTX 40 series lineup, which will be based on the Ada Lovelace GPU architecture, were published over at LaptopVideo2Go and discovered by @Harukaze5719. The PCI IDs are now added to the public PCI ID list.
Although not all of the GPUs mentioned are to be launched in the market, some device IDs are set for preliminary engineering boards that never go out to retail.
ID 2684 supposedly belongs to NVIDIA GeForce RTX 4090. Based on AD102, RTX 4090 is the flagship of the 40-series GPUs as of now, featuring 16384 CUDA Cores and 24GB of GDDR6X memory. The GPU will come packed with an insane 96 MB of L2 cache and a total of 384 ROPs. The card allegedly features a 60% faster base clock than RTX 3090.
In addition to the desktop RTX 4090, the rest of the list contains RTX 40 mobile SKUs indicated by the GN part in their names. The various SKUs will be based on AD103, AD104, AD106, and AD107 chips.
Another leaker kopite7kimi added that the GN21-X11 series SKUs will feature the AD103 GPU, which will be followed by the GN21-X9 with AD104, GN21-X6 with AD106, and GN21-X4/X2 with the AD107 GPUs.There is also mention of the newly released GTX 1630 in the list.
The RTX 4080 will feature the AD103 GPU and an estimated 10240 CUDA cores. It will feature 16 GB of GDDR6X memory on a 256-bit memory bus with a speed of 21Gbps. The RTX 4080 has a TDP of approximately 420W.
Similarly, the RTX 4070 will be paired with the AD104 and will feature 7168 CUDA cores and a 160-bit memory bus interface with 10GB of GDDR6 memory support clocking at 18 Gbps. Its expected TDP is 300W.
NVIDIA GeForce RTX 40 series is rumored to be launching from September onwards. However, we will have to wait until Q4’22 or early Q1’23 for the mobile variants.
NVIDIA Corporation is an American multinational technology company incorporated in Delaware, based in Santa Clara, California. They design graphics processing units (GPUs) for the gaming and professional markets, as well as system on chip units (SoCs) for the mobile computing and automotive market.
Best known for the “GeForce” lines of GPUs, they are a direct competitor to AMD’s “Radeon” series. NVIDIA has also expanded its offerings with its handheld game consoles Shield Portable, Shield Tablet, and Shield Android TV and its cloud gaming service GeForce Now.