Nvidia has officially made public its RTX 50 series GPUs on the market today. After months of rumors and leaks the next-generation RTX Blackwell GPUs are now officially available and there are four coming soon.
Nvidia President Jensen Huang revealed the RTX 50-series GPUs at a CES keynote on Thursday evening. Huang announced a $1,999 RTX 5090 an $999 RTX 5080 and a $749 RTX 5070 Ti and an affordable price of $549 for the RTX 5070. New Nvidia RTX 5090 as well as RTX 5080 will launch on January 30th. an RTX 5070 Ti and RTX 5070 are expected to be available in February.
Its RTX 50-series GPUs feature a brand new layout for the Founders Edition, with just two flow-through fans that double as well and the vapor chamber is 3D as well as GDDR7 memory. Each of the RTX 50-series GPUs comes with PCIe Gen 5 and comes with DisplayPort 2.1b connectors that can drive displays with up to 8K resolution and at 165Hz.
Surprisingly, this RTX 5090 Founders Edition will come with a two-slot GPU that fits into smaller form factor computers an enormous change from the dimensions that was the RTX 4090. The RTX 5090 features 32GB of GDDR7, a capacity of a memory speed that is 1,792GB/sec as well as a staggering 21760 CUDA cores.
It's a GPU that Nvidia promises to be two times more efficient than the RTX 4090 thanks to DLSS 4 and the Blackwell technology. However, it comes with a price in terms of power consumption, since Nvidia states that the RTX5090 will come with a total graphic capacity of 575 Watts as well as an optimum PSU requirement for 1000 watts. This is 125 watts more power than what you get from the RTX 4090, however, I'm hoping that it will be a lot less. RTX 5090 is much more energy efficient, meaning you'll never use 575 watts.
Nvidia showed Cyberpunk 2077 running on an RTX 5090 equipped with the DLSS 4 setting at 238fps, compared to 106fps with an RTX 4090 running the DLSS 3.5. Both GPUs run the game with Ray Tracing enabled.
RTX 5080 is intended to be twice as fast as the RTX 4080. It will come with 16GB of GDDR7 memory with a 960GB/sec memory bandwidth of 10752 CUDA cores. The RTX 5080 is expected to have a capacity of 360 Watts. Nvidia recommends an 850-watt power source. Nvidia promises significant performance improvements with its RTX 5080 over the prior RTX 4080 model due to these specifications.
Nvidia is also launching the new RTX-5070 Ti as well as the RTX 5070. The RTX 5070T comes with 16GB of GDDR7 memory and the memory bandwidth is 896GB/s as well as the capacity of 8,960 CUDA cores. The RTX 5070 comes with 12GB of GDDR7 memory, a memory bandwidth of 672 GB/sec as well as 6144 CUDA cores. The RTX 5070T will have a capacity of 300 Watts, and require a 750-watt Power Supply and the RTX 5070 comes with an overall graphics capacity of 250 Watts. It only requires a 650-watt power supply.
Nvidia claims that it will announce that the RTX 5070 Ti is expected to be two times quicker than RTX 4070 Ti, as well as the RTX 5070, is expected to be twice faster than that of the RTX 4070. Huang also claimed at CES it will be RTX 5070, which would offer " RTX 4090 performance at $549," however this is likely due to DLSS 4 enhancements, not just rasterization speed.
Nvidia will also be bringing its RTX 50-series of laptops, featuring its RTX 5090 notebook GPU launching featuring 24GB of GDDR7 memory. Its RTX 5080 GPU for laptops comes having 16GB of GDDR7 RAM. Then there's the RTX 5070 Ti comes with 12GB GDDR7 RAM as well and the RTX 5070 will come with just 8GB of memory in GDDR7. RTX 50 series laptops will be available from March 1st by various PC manufacturers.
Huang showed off the Nvidia RTX Blackwell GPUs with a live rendering demonstration at the start of Nvidia's CES opening keynote today. The demonstration featured new RTX Neural Materials, RTX Neural Faces, text that can be animated, and DLSS 4. "The new generation of DLSS can generate beyond frames, it can predict the future," Huang claims. Huang. "We used GeForce to enable AI, and now AI is revolutionizing GeForce."
The new RTX Neural shaders by Nvidia can reduce the texture in games, and RTX Neural Faces aim to enhance the quality of faces with the help of Artificial Intelligence (AI) generative. The next version of DLSS features Multi Frame Generation, which creates three additional frames in a traditional frame and allows for the addition of frame rates by as much as 8x on top of traditional rendering as per Nvidia.
DLSS 4 includes the ability to use transformers in real time to enhance image quality, decrease ghosting, and increase resolution when moving. It is also able to reduce ghosting and increase detail in motion. DLSS 4 upgrade will even run on older RTX GPUs since features are upgraded to new models of the transformer AI.
The announcement of Nvidia's RTX 50-series launch announcement came nearly two years following its announcement that the RTX 4090 as well as RTX 4080 were introduced and built upon the Nvidia Ada Lovelace architecture. Nvidia's RTX 40 series of GPUs focuses on improving ray tracer performance using Deep Learning Super Sampling (DLSS) version 3 along with the RTX 4090 has delivered amazing performance improvements over the prior RTX 3090 graphics card.

.png)
.png)