NVIDIA‘s latest mid-range GPU, the GeForce RTX 4070, is out. Reviews went live recently, and it got a somewhat lukewarm reception from the review community. Unfortunately for us, our 4070 card didn’t make it in time, and it’s still in transit. Anyway, some reviewers praise its performance and its efficiency, but some were not impressed with its pricing or the lackluster performance improvement from the previous generation. Speaking of its performance, the RTX 4070 performs on par with the RTX 3080 on average while consuming only around 180+W on average. The RTX 4070 is an ideal GPU for 1080p and 1440p gaming, especially since recent and latest AAA-titled games have become graphics-intensive. Check out more details below.
NVIDIA GeForce RTX 4070 Performance – TL;DR
Since our GeForce RTX 4070 is still in transit, our review may not go live until next week. So, I decided to curate the information from the reviewer’s guide and the reviews that went live recently.
In 1080p resolution, the RTX 4070 is about 3% faster than the RTX 3080 10GB; about 16% faster than the RTX 3070 Ti; and about 66% faster than the much older RTX 2070 Super. If we compare it to its direct predecessor, the RTX 3070, the RTX 4070 is about 22% faster on average.
By the way, the RTX 4070 and the RTX 3070 have the same amount of CUDA cores, which is 5888. But the RTX 4070 has a faster GDDR6X memory and a larger 12GB capacity. Not to mention, it has a boost clock speed of 2.4-2.5GHz. Despite having the same number of CUDA cores, the RTX 4070 is based on the newer and better Ada Lovelace architecture, 4th gen Tensor Cores, and 3rd gen RT Cores.
Moving to 1440p resolution, the RTX 4070 and RTX 3080 10GB performs identically. Sometimes, the RTX 3080 beats the RTX 4070 by a couple of percents. Meanwhile, the RTX 4070 is around 18% faster than the RTX 3070 Ti.
But how does it compare with the RTX 4070 Ti and RTX 4080? Well, the RTX 4070 Ti is about 20% faster than the RTX 4070 on average, depending on the scenario. And compared to the RTX 4080, the RTX 4080 is almost 50% faster than the RTX 4070 on average. But the RTX 4080 is double the price of the RTX 4070.
That’s about it for now. Hopefully, our review will go live very soon. Meanwhile, below are NVIDIA’s official news release and the specs of the RTX 40 series GPUs. Stay tuned.
Find out how NVIDIA’s GeForce RTX 4070 performs in our review here.
Pricing and Availability
You can also check out some of the best RTX 4070 graphics cards here.
NVIDIA GeForce RTX 4070 Brings Power of Ada Lovelace Architecture and DLSS 3 to Millions More Gamers and Creators, Starting at US$599
NVIDIA today announced the GeForce RTX™ 4070 GPU, delivering all the advancements of the NVIDIA® Ada Lovelace architecture — including DLSS 3 neural rendering, real-time ray-tracing technologies, and the ability to run most modern games at over 100 frames per second at 1440p resolution — starting at US$599.
Today’s PC gamers increasingly want super-premium visual experiences, with 83% of GeForce RTX 40 Series desktop gamers enabling ray tracing and 79% turning on DLSS, a revolutionary breakthrough in AI-powered graphics that massively boosts performance, now in its third generation.
Matt Wuebbling, vice president of global GeForce marketing at NVIDIA:
“Making sure games are ‘RTX ON’ with both ray tracing and DLSS has become a top priority for the majority of PC gamers,”. “With prices starting at US$599, the RTX 4070 is an amazing upgrade for gamers running on previous-generation GPUs.”
Play at 100+ FPS at 1440p with DLSS 3
The RTX 4070 allows the latest games such as A Plague Tale: Requiem, Dying Light 2, Microsoft Flight Simulator, Warhammer 40,000: Darktide and other demanding titles to run at over 100 frames per second, thanks to DLSS 3, as well as popular games such as Battlefield 2042, Call of Duty: Modern Warfare II and Marvel’s Guardians of the Galaxy that support DLSS 2.
With DLSS 3, AI-powered frame generation creates new, high-quality frames for smoother gameplay, rendering seven out of every eight pixels in a scene. Exceptional responsiveness is maintained through NVIDIA Reflex, which reduces input lag between devices and monitors.
Ray-tracing performance has also significantly improved, thanks to advancements like Shader Execution Reordering (SER), cutting-edge Opacity Micromap, and Displaced Micro-Mesh Engines. These innovations enable even the most demanding games to simultaneously implement multiple ray-tracing effects or even full ray tracing, also known as path tracing, for unparalleled realism and immersion.
This potent combination of third-generation ray tracing and DLSS has resulted in a 16x leap in ray-tracing operations per pixel over the past five years. Adoption rates have soared, with over 400 RTX games and applications available, with the adoption of DLSS 3 happening 7x faster than its predecessor.
Class-Leading Performance and Features
Compared to the RTX 2070 SUPER, the GeForce RTX 4070 is on average 2.6x faster with DLSS 3, and on average 1.4x faster than the GeForce RTX 3080 with DLSS 3.
DLSS 3 provides Ada Lovelace GPUs with a tremendous performance boost, but the GeForce RTX 4070 also excels in traditional games that do not include more advanced features such as ray tracing and DLSS.
In these rasterized games, the GeForce RTX 4070 is on par with the GeForce RTX 3080 while running at nearly half the power — and offering an additional 2GB of memory. Additionally, the RTX 4070’s memory subsystem has been enhanced with 36MB of L2 cache and 12GB of ultra-high-speed GDDR6X memory.
Perfect for Content Creators
The GeForce RTX 4070 GPU comes backed by the NVIDIA Studio platform, which enables content creators to work faster and smarter with RTX acceleration and AI tools. Serving creators from livestreamers to video editors, the platform supercharges 100 creative apps, provides lasting stability with NVIDIA Studio Drivers and includes a powerful suite of Studio apps, such as NVIDIA Omniverse™, Canvas and Broadcast.
3D modelers rendering 4K scenes with the AI-powered DLSS 3 in NVIDIA Omniverse — a hub to interconnect existing 3D workflows, replacing linear pipelines with live-sync creation for real-time collaboration — can expect 2.8x faster performance than with last generation’s GeForce RTX 3070 Ti.
Broadcasters deploying the eighth-generation NVIDIA video encoder NVENC, with support for AV1, will enjoy 40% better efficiency. Livestreams will appear as if bitrate was increased by 40% — a big boost in image quality for popular broadcast apps like OBS Studio. AV1 is already supported in Discord with a full release coming to YouTube via OBS Studio soon.
Video editors will see 20% faster video export speeds, thanks to the eighth-generation NVIDIA encoder on the GeForce RTX 4070, along with AV1 incorporated into top video editing apps.
And all creators can benefit from the new fourth-generation Tensor Cores for AI tools, providing a significant performance increase from the last generation.
NVIDIA GeForce RTX 40 Series GPU Specifications
|Graphics Card||RTX 4070||RTX 4070 Ti||RTX 4080||RTX 4090|
|Fabrication Process||TSMC 4N NVIDIA Custom Process||TSMC 4N NVIDIA Custom Process||TSMC 4N NVIDIA Custom Process||TSMC 4 nm NVIDIA Custom Process|
|Transistor Count||35.8 billion||35.8 billion||45.9 billion||76.3 billion|
|Architecture||Ada Lovelace||Ada Lovelace||Ada Lovelace||Ada Lovelace|
|Graphics Processing Clusters||4||5||7||11|
|Texture Processing Clusters||23||30||38||64|
|Tensor Cores||184 (4th Gen)||240 (4th Gen)||304 (4th Gen)||512 (4th Gen)|
|RT Cores||46 (3rd Gen)||60 (3rd Gen)||76 (3rd Gen)||128 (3rd Gen)|
|Boost Clock||2475 MHz||2610 MHz||2505 MHz||2520 MHz|
|Memory Clock||10500 MHz||10500 MHz||11200 MHz||10501 MHz|
|Memory Data Rate||21 Gbps||21 Gbps||22.4 Gbps||21 Gbps|
|L2 Cache Size||36864 KB||49152 KB||65536 KB||73728 KB|
|Total Video Memory||12 GB GDDR6X||12 GB GDDR6X||16 GB GDDR6X||24 GB GDDR6X|
|Total Memory Bandwidth||504 GB/s||504 GB/s||716.8 GB/s||1008 GB/s|
|Texture Rate||455.4 Gigatexels/sec||626 Gigatexels/sec||761.5 Gigatexels/sec||1290.2 GigaTexels/sec|
|Connectors||3 x DisplayPort 1 x HDMI||3 x DisplayPort 1 x HDMI||3 x DisplayPort 1 x HDMI||3 x DisplayPort 1 x HDMI|
|Minimum Power Supply||650 Watts||700 Watts||750 Watts||850 Watts|
|Total Graphics Power (TGP)||200 Watts||285 Watts||320 Watts||450 Watts|
|PCI Express Interface||Gen 4||Gen 4||Gen 4||Gen 4|
|Release Date||April 2023||January 2023||November 2022||October 2022|