RTX 5000 workstation GPU reported to have 15360 cores and 32GB memory

News about Nvidia’s RTX 4000 workstation GPUs comes at a sweet time. As for the RTX 6000 series based on the “Ada Lovelace” architecture, the professional GPU bears 18176 cores.

The newcomer in the lineup, the 4000, has a third of the cores which translates to a smaller physical size without the requirement for auxiliary power, featuring an ultra-efficient AD104 GPU requiring only 70W and no power connectors, it comes atop an SFF design.

The latest news about Nvidia comes from a leak by reputed Nvidia hardware leakers @kopite7kimi. They state that it’s probable that RTX 5000 comes with 15360 FP32, boasting 32GB GDDR6 memory.

This is quite a lot, given that most consumer GPUs go only up to 24GB, which is true with the RTX 40-series GPUs. However, the 5000 series probably won’t see the latest GDDRX6 technology.

Combining this information with the fact that the AD103 offers a maximum of 10240 cores, it can be safely concluded that the 5000 series also use the AD102.

It was also stated that to match the 32GB config, Nvidia would have to cut down the memory bus to 256-bit. The memory bandwidth is expected to be mathematically around 640GBps.

It is also expected that Nvidia will optimize the card in terms of TGP, which is expected to be anywhere in the region of 200-250W. The pricing also depends on the market but is expected to be $3000-$4500.

There’s a pretty large gap in terms of specs between the 4000 series and the 6000 series. This gap will probably be filled by the upcoming RTX 5000 and 4500 series.

It is also to be noted that despite the mammoth core count of the 6000 series GPU for workstations, it must be kept in mind that the GPU houses 144SMs. This means that there is a possibility for a GPU with as many as 18432 cores.

This remains a rumor, with no new information officially being released. However, given that it uses the GPU architecture that has been in the market for a while now, i.e., the AD102, the release could be imminent.

About Nvidia

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.