NVIDIA’s next-gen GeForce RTX 40 ‘Ada Lovelace’ GPUs are said to demand about a kilowatt of power, so get your PC ready, especially your power supplies. According to various sources, NVIDIA’s next-generation GeForce RTX 40 GPUs will be incredibly fast and power-hungry based on the Ada Lovelace graphics architecture. According to the newest speculations, the flagship GPU, the AD102, could consume over 800 Watts of electricity, according to Kopite7kimi and Greymon55.
According to the most recent information, GeForce RTX 40 ‘Ada Lovelace’ GPUs could be ready as early as September 2022, according to Greymon55. However, the most intriguing aspect of the report is the supposed power figures, not the launch. According to both leakers, the flagship AD102 GPU will have various SKUs for the RTX 4080, RTX 4080 Ti, and RTX 4090 desktop graphics cards. It appears that these GPUs will have multiple power targets, with the entry-level GPU hitting 450 Watts of peak consumption, followed by the Ti model at roughly 600 Watts, and the flagship RTX 4090 at around 850 Watts.
Although both leakers claim that these aren’t final specifications and that the power figures may change in retail variations, there’s reason to assume that they could be correct. NVIDIA is already funding development on the upcoming PCIe Gen 5 connector, allowing up to 600W power input per connector.
The GeForce RTX 3090 Ti is one example, with a TGP of 450W and the distinction of being the first desktop graphics card to use such a connection interface. The next-generation cards are expected to use the same PCIe standard, although the top type may get two Gen 5 connectors to help with the 800W power need.
Several PSU manufacturers have already begun shipping their brand new Gen 5 power supplies, including the necessary connectors to support the next-generation GPUs. Still, they only have one primary Gen 5 connector, which means that if NVIDIA uses a second 16-pin port, users will need to use a 2x 8-pin to 1x 16-pin adapter, which will be included with most of these PSUs.
Of course, these are all rumours for the time being, but based on their past stories and leaks, both leakers have a high level of reliability, so this could wind up being true. Given that NVIDIA’s Ada Lovelace GeForce RTX 40 series portfolio aims for a massive 2x performance boost to fight against AMD’s rDNA three products, the green team may be going all out in terms of power and pricing, in addition to performance.
Previously rumoured specifications revealed a significant upgrade to the basic specifications. According to Kopite’s preliminary specs (which may change), the NVIDIA AD102 “ADA GPU” appears to contain 18432 CUDA Cores. This is nearly double the number of cores in Ampere, which was already a significant improvement over Turing. A clock speed of 2.2 GHz would result in 81 TFLOPs of computational performance (FP32). This is more than twice as fast as the current RTX 3090, with 36 TFLOPs of FP32 processing power.
Also Read:
Prices of Crypto Mining GPUs of both Nvidia and AMD plummeted as the Crypto wave cools down
