NVIDIA’s 40 Series GPUs have finally been released after months of anticipation and debate among its add-in board partners. GeForce RTX 4090 and 4080 were announced by the firm today during the GTC 2022 keynote. The two GPUs, according to NVIDIA, greatly improve ray tracing performance by fully using the new “Ada Lovelace” architecture. Together with Taiwan Semiconductor Manufacturing Business (TSMC), the company created a brand-new “4N” production method that, according to NVIDIA, is up to twice as energy-efficient as the 8nm method it utilised for its 30 Series cards.
Ada Lovelace now supports third-generation RT Cores from NVIDIA, as well as Shader Execution Reordering, DLSS 3.0, and other cutting-edge rendering technologies. As a result, ray tracing speed has been greatly enhanced. The ray tracing performance in some games should be two to three times higher than it was with NVIDIA’s Ampere GPUs, the company claimed. The business showed off Cyberpunk 2077 operating at a nearly constant 100 frames per second with all of the game’s ray tracing capabilities turned up to the highest setting. The new design, according to NVIDIA, has resulted in up to a two-fold increase in rasterization performance.
On October 12th, the GeForce RTX 4090, which costs $1,599., will be on sale, marking the launch of the first of NVIDIA’s new Ada Lovelace GPUs. NVIDIA asserts that the 3090 Ti is two to four times quicker than its newest flagship, the GTX 1080 Ti while using the same amount of power. The new flagship has 24GB of GDDR6X memory. Fortunately, it starts off at $100 more expensive than its predecessor. 16,384 CUDA Cores with a base clock speed of 2.23GHz were able to fit into the RTX 4090 by NVIDIA.
NVIDIA is also releasing two RTX 4080 models in addition to the 4090. 12GB of GDDR6X memory is available on the entry-level model, which costs $899, while 16GB of memory costs $1,199. November will see the arrival of both variants. NVIDIA will, however, exclusively market the more costly model’s Founders Edition variant. You’ll need to contact the company’s partners to get the 12GB version, which may make it challenging to locate units with a starting price of $899.
The 16GB 3080 has 9,728 Cuda Cores and a base clock speed of 2.21GHz, with a top boost speed of 2.51GHz. A relatively modest 7,680 CUDA Cores are included in the 12GB variant, but the base and boost speeds are 100MHz quicker. Fortunately, you probably won’t need to replace your power supply if you intend to upgrade from a 3070 or 3080 because NVIDIA suggests a 700-watt PSU for the 12GB edition and a 750-watt PSU for its more potent sister. In the event that you choose to purchase a new power supply, you should hold off until additional ATX 3.0 PSUs are released later this year. The reason for this is that, at the very least, the Founders Edition devices will support the new PCIe Gen-5 16-pin connection standard. In spite of this, NVIDIA will also supply an 8-pin adaptor for those who don’t wish to rewire their machine.
The 40 Series GPUs from the NVIDIA release coincide with a trying period for the business. The demand from gamers and bitcoin miners made it difficult for much of the market to get the most recent GeForce GPUs at MSRP. All of that has altered recently as a result of the recent crypto market crisis and Ethereum’s much-awaited conversion to proof-of-stake coin minting. Due to the aforementioned circumstances, the secondhand market became swamped with 30 Series GPUs, making it extremely difficult for the company’s AIB partners and merchants to sell new video cards at MSRP.