The last time NVIDIA released the GeForce RTX 30 Series was back in September 2020. After waiting for more than two years, the green GPU company is finally releasing the next generation of its GeForce RTX GPUs. But prepare your wallet and bank account, because it is more expensive than ever.
NVIDIA CEO Jensen Huang finally formally introduced RTX 4090, RTX 4080 16GB and RTX 4080 12GB models at NVIDIA GTC Keynote, after much anticipation from gamers and tech-enthusiasts.
The GPUs were fabricated based on NVIDIA ‘Ada Lovelace’ architecture. This architecture offers greater raster, raytracing, and AI-accelerated computation than Ampere and is built on a custom TSMC 4N process.
Third-generation RT cores and fourth-generation Tensor Cores are introduced under the new architecture. Additionally, DLSS 3 upscaling technology is supported by the latest GPUs. Dual AV1 encoders on NVIDIA RTX 40 GPUs provide up to 8K 60 FPS recording for game streaming.
VideoCardz reported that the RTX 4090 has a AD102-300 GPU with 16384 CUDA cores and is capable of boosting the clock speed up to 2,520MHz. This translates to a single-precision performance of 82.6 TFLOPS, or 2.3 times greater than the RTX 3090. The flagship Ada-based GPU will come with 24GB of GDDR6X memory, with a peak bandwidth of 1 TB/s. This new card is more power hungry than the previous generation, with a default TDP of 450W, it will demand at least 100W more power than the 3090 Ti.
Meanwhile, the RTX4080 are basically two different GPUs but for some unknown reasons, NVIDIA decided to give them the same name.
The GeForce RTX 4080 (16GB GDDR6X) has an AD103-300 GPU, boasting 9,728 CUDA cores, 780 Tensor-TFLOPs, 113 RT-TFLOPs, and 49 Shader-TFLOPs of power, making it twice as fast as the GeForce 3080 Ti, using nearly 10% less power. This GPU has a TDP of 320W.
RTX 4080 (12GB GDDR6X) is supposed to be RTX 4070, because they have completely different specs from the 16GB version. This version has an AD104-400 GPU, 7,680 CUDA cores, 639 Tensor-TFLOPs, 92 RT-TFLOPs, and 40 Shader-TFLOPs. This GPU has a TDP of 285W.
NVIDIA said the RTX 4090 offers up to 4x more performance when compared to RTX 3090 Ti and the RTX 4080 is up to 2x faster than RTX 3080 Ti and more powerful than RTX 3090 Ti.
Brands from around the world such as ASUS, Colorful, Gainward, Galaxy, GIGABYTE, INNO3D, MSI, Palit, PNY, and ZOTAC will be releasing their own versions of RTX 40 Series GPUs. NVIDIA will also be releasing limited Founders Edition RTX 4090 and RTX 4080 (16GB) cards.
Pricing and Availability
According to NVIDIA’s website, the price of RTX 4090 will start from RM8,400 and available on 12 October 2022. While the RTX 4080 (16GB) price will start from RM6,300 and RM4,730 for the 12GB version. These two GPUs will be coming in November 2022.