NVIDIA GeForce RTX 4070 Graphics Card Specs, Performance, Price & Availability – Everything We Know So Far
NVIDIA GeForce RTX 4070 will be the next-generation high-end gaming graphics card, offering the latest graphics architecture based on Ada Lovelace GPUs. The graphics card will be replacing the RTX 3070, a very popular gaming graphics card in the $500-$600 US segment.
NVIDIA GeForce RTX 4070 Graphics Card - Purely Designed For Enthusiast Gamers
While there's no denying the enthusiasm around the higher-end GeForce RTX 4090 & GeForce RTX 4080 series graphics cards that offer the best of the best gaming performance, the RTX 4070 series graphics cards will be designed around the $500 US segment which is a high-end price range that still offers lots of performance at hand. It's simple, the RTX 4090 series will be aimed at users who want the best of the best without worrying about the amount of money they are spending while the RTX 4080 series is aimed at users who want the best gaming performance at the best possible price. The RTX 4070 will be the sweet spot for high-end gaming, offering a buttery smooth 2K game experience.
The previous GeForce RTX 3070 was touted to offer a huge improvement over the RTX 2070 and was said to offer performance faster than the RTX 2080 Ti but ended up mostly on par with the Turing flagship with only the RTX 3070 Ti exceeding the performance of the previous Turing GPU flagship. It looks like the RTX 4070 will be placed in a similar position where it might offer graphics performance on par or close to the RTX 3080 Ti but a 'Ti' variant going further ahead in graphics performance.
We should expect similar things with the next-generation gaming solution too but an important factor to consider is that GPUs are becoming more power-hungry and more pricey. It is a trend that might continue into the future as we get better products but in return, there's always a cost to pay for end consumers. So starting with what we know so far, first we should take a look at the brand new Ada Lovelace or AD10* class GPUs that will be powering the next-gen GeForce RTX 40 series cards.
NVIDIA's AD104 'Ada Lovelace' GPU - The Next-Gen Powerhouse
Starting with the GPU configuration, the NVIDIA GeForce RTX 4070 series graphics cards are said to utilize the AD104 GPU core. The GPU is said to measure around 300mm2 and will utilize the TSMC 4N process node which is an optimized version of TSMC's 5nm (N5) node designed for the green team.
The NVIDIA Ada Lovelace AD104 GPU is expected to feature up to 5 GPC (Graphics Processing Clusters). This is the one less GPC than the GA104 GPU. Each GPU will consist of 6 TPCs and 2 SMs which is the same configuration as the existing chip. Each SM (Streaming Multiprocessor) will house four sub-cores which is also the same as the GA102 GPU. What's changed is the FP32 & the INT32 core configuration. Each sub-core will include 128 FP32 units but combined FP32+INT32 units will go up to 192. This is because the FP32 units don't share the same sub-core as the IN32 units. The 128 FP32 cores are separate from the 64 INT32 cores.
So in total, each sub-core will consist of 32 FP32 plus 16 INT32 units for a total of 48 units. Each SM will have a total of 128 FP32 units plus 64 INT32 units for a total of 192 units. And since there are a total of 60 SM units (12 per GPC), we are looking at 7,680 FP32 Units and 3,840 INT32 units for a total of 11,520 cores. Each SM will also include two Wrap Schedules (32 thread/CLK) for 64 wraps per SM. This is a 50% increase on the cores (FP32+INT32) and a 33% increase in Wraps/Threads vs the GA102 GPU.
NVIDIA AD103 'Ada Lovelace' Gaming GPU 'SM' Block Diagram (Image Credits: Kopite7kimi):
Moving over to the cache, this is another segment where NVIDIA has given a big boost over the existing Ampere GPUs. The Ada Lovelace GPUs will pack 192 KB of L1 cache per SM, an increase of 50% over Ampere. That's a total of 2 MB of L1 cache on the top AD104 GPU. The L2 cache will be increased to 48 MB as mentioned in the leaks. This is a 12x increase over the Ampere GA104 GPU that hosts just 4 MB of L2 cache. The cache will be shared across the GPU.
Finally, we have the ROPs which are also increased to 32 per GPC, an increase of 2x over Ampere. You are looking at up to 160 ROPs versus just 96 on the GA104 Ampere GPU. There are also going to be the latest 4th Generation Tensor and 3rd Generation RT (Raytracing) cores infused on the Ada Lovelace GPUs which will help boost DLSS & Raytracing performance to the next level. Overall, the Ada Lovelace AD103 GPU will offer:
- 5 GPCs vs 6 GPCs on GA104
- +25% Cores vs GA104 GPU
- 50% More L1 Cache (Versus Ampere GA104)
- Twice More L2 Cache (Versus Ampere GA104)
- +66% ROPs (Versus Ampere GA104)
- 4th Gen Tensor & 3rd Gen RT Cores
NVIDIA AD104 'Ada Lovelace' Gaming GPU Block Diagram Mock-Up (Image Credits: SemiAnalysis):
Do note that clock speeds, which are said to be between the 2-3 GHz range, aren't taken into the equation so they will also play a major role in improving the per-core performance versus Ampere.
NVIDIA GeForce RTX 4070 Graphics Cards Specifications
As we saw with the GA104 GPU, NVIDIA can have various configurations of the AD104 GPU for its GeForce RTX 4070 series lineup. We realistically expect there to be two variants, the RTX 4070 and the RTX 4070 Ti. The former will be part of the initial lineup while the latter would launch as a mid-cycle refresh. The most entry-level GeForce RTX 30 had 23% fewer cores compared to the full chip but this time, the '80-class' graphics cards are going to be powered by their own chip rather than relying on the AD102 GPU that the flagships use. As such, we can expect anywhere from 6144 to the full 7680 core configuration.
NVIDIA GeForce RTX 4070 'Expected' Specifications
The NVIDIA GeForce RTX 4070 is going to be a cut-down configuration with a slightly higher core count than the RTX 3070 TI (6144 cores) allowing some room for a 'Ti' variant in the future with the full-fat configuration. The core count is expected to be 7168 which means 56 of the 60 SMs will be enabled. The GPU will come packed with 48 MB of L2 cache and up to 160 ROPs which is simply insane.
The clock speeds are not confirmed yet but considering that the TSMC 4N process is being used, we are expecting clocks between the 2.0-3.0 GHz range. The higher than usual clock speed bump comes from the fact that NVIDIA is making a two-node jump considering the Ampere GPUs with Samsung 8nm node was in reality a 10nm process node with some optimizations. NVIDIA is skipping 7nm and going straight for a 5nm node and not even the vanilla variant but an optimized version of it. With Pascal on the TSMC 16nm node, NVIDIA delivered a huge frequency leap and we can expect a similar jump this time around too.
As for memory specs, the GeForce RTX 4070 is expected to rock 10 GB GDDR6 capacities that are said to be clocked at 18 Gbps speeds across a 160-bit bus interface. The card may rock a TBP of 300W and the leaker also mentions that the pricing of this card would not be lower than the RTX 3070 or 3070 Ti graphics cards. Simply put, we can expect a price bump in the '70' class graphics segment.
As for its feature set, the NVIDIA GeForce RTX 4070 graphics cards will rock all the modern NV feature sets such as the latest 4th Gen Tensor Cores, 3rd gen RT cores, and the latest NVENC Encoder, and NVCDEC Decoder, and support for the latest APIs. They will pack all the modern RTX features such as DLSS, Reflex, Broadcast, Resizable-BAR, Freestyle, Ansel, Highlights, Shadowplay, and G-SYNC support too.
NVIDIA GeForce RTX 4070 Series Preliminary Specs:
|Graphics Card Name||NVIDIA GeForce RTX 4070 Ti||NVIDIA GeForce RTX 4070||NVIDIA GeForce RTX 3070 Ti||NVIDIA GeForce RTX 3070|
|GPU Name||AD104-400?||AD104-300?||Ampere GA104-400||Ampere GA104-300|
|Process Node||TSMC 4N||TSMC 4N||Samsung 8nm||Samsung 8nm|
|Transistors||TBD||TBD||17.4 Billion||17.4 Billion|
|TMUs / ROPs||TBD / 160||TBD / 144||192/ 96||184 / 96|
|Tensor / RT Cores||TBD / TBD||TBD / TBD||192/ 48||184 / 46|
|Base Clock||TBD||TBD||1575 MHz||1500 MHz|
|Boost Clock||TBD||TBD||1770 MHz||1730 MHz|
|FP32 Compute||~38 TFLOPs||~36 TFLOPs||22 TFLOPs||20 TFLOPs|
|RT TFLOPs||TBD||TBD||42 TFLOPs||40 TFLOPs|
|Tensor-TOPs||TBD||TBD||174 TOPs||163 TOPs|
|Memory Capacity||12 GB GDDR6X?||12 GB GDDR6||8 GB GDDR6X||8 GB GDDR6|
|Memory Speed||21 Gbps||18 Gbps||19 Gbps||14 Gbps|
|Bandwidth||504 GB/s||432 GB/s||608 Gbps||448 Gbps|
|Price (MSRP / FE)||$599 US?||$499 US?||$599 US||$499 US|
|Launch (Availability)||2022||2022||10th June 2021||29th October 2020|
NVIDIA GeForce RTX 4070 Graphics Cards Performance
As for the performance of the gaming GPUs, we can only use theoretical numbers here since the launch is a bit far away but based on what we know, the RTX 40 series cards might be the first gaming cards to hit the 100 TFLOPs compute horsepower limit.
Just for comparison's sake:
- NVIDIA GeForce RTX 4090 Ti: ~103 TFLOPs (FP32) (Assuming 2.8 GHz clock)
- NVIDIA GeForce RTX 4090: ~90 TFLOPs (FP32) (Assuming 2.8 GHz clock)
- NVIDIA GeForce RTX 4080: ~50 TFLOPs (FP32) (Assuming 2.5 GHz clock)
- NVIDIA GeForce RTX 3090 Ti: 40 TFLOPs (FP32) (1.86 GHz Boost clock)
- NVIDIA GeForce RTX 4070 Ti: ~38 TFLOPs (FP32) (Assuming 2.5 GHz clock)
- NVIDIA GeForce RTX 4070: ~36 TFLOPs (FP32) (Assuming 2.5 GHz clock)
- NVIDIA GeForce RTX 3090: 36 TFLOPs (FP32) (1.69 GHz Boost clock)
- NVIDIA GeForce RTX 3080: 30 TFLOPs (FP32) (1.71 GHz Boost clock)
- NVIDIA GeForce RTX 3070 Ti: 22 TFLOPs (FP32) (1.77 GHz Boost clock)
- NVIDIA GeForce RTX 3070: 20 TFLOPs (FP32) (1.72 GHz Boost clock)
Based on a theoretical clock speed of 2.5 GHz, you get up to 36 TFLOPs of compute performance and the rumors are suggesting even higher boost clocks. Now, these are definitely sounding like peak clocks, similar to AMD's peak frequencies which are higher than the average 'Game' clock. A 36+ TFLOPs compute performance means more performance on a '70-class' GPU than an '80-class flagship' which will be a good bump. But one should keep in mind that compute performance doesn't necessarily indicate the overall gaming performance.
This will be around a 80% compute performance uplift for the GeForce RTX 4070 graphics card versus its predecessor and this is without even factoring in the RT and Tensor core performance which is expected to get major lifts too in their respective department. Now FLOPs aren't necessarily reflective of the graphics or gaming performance but they do provide a metric that can be used for comparison.
Gamers should expect 2K gaming to be buttery smooth on these graphics cards and with DLSS, we might even see playable 60 FPS at 4K resolution with DLSS enabled. But to be realistic, 1440p game titles would be the sweet spot for this graphics card.
NVIDIA GeForce RTX 4070 Graphics Cards Price & Availability
The NVIDIA GeForce RTX 3070 was launched at $499 US but its refreshes have really messed up the prices. The RTX 3070 Ti launched at $599 US, a $100 US premium over the standard model but not offering a big enough performance boost to make the extra $ worth it.
NVIDIA really needs to secure more gamers in the $500 US segment but they have to deliver a product that's worth the price. The RTX 2070 and RTX 3070 were a little underwhelming & while they did manage to close the gap with the previous-gen flagship, they still couldn't recapture the glory of the GTX 1070 which was a good bit ahead of the flagship Titan's that preceded it or the absolutely amazing price point of the GTX 970. With the performance leaps that are expected from the Ada Lovelace architecture, the RTX 4070 has the potential to become one stellar graphics card option in the $500 US price range. Also, if NVIDIA can simply ignore the refresh and give us a full-fat GA104 configuration at the $500 price point, that would be an insane value and one of the best graphics card in its class for some time to come.
The NVIDIA GeForce RTX 40 series graphics cards are rumored for a mid-July reveal and a launch is expected in Q3 2022 so we will know for sure what NVIDIA is up to in the next few months.
NVIDIA GeForce GPU Segment/Tier Prices
|Titan Tier||Titan X (Maxwell)||Titan X (Pascal)||Titan Xp (Pascal)||Titan V (Volta)||Titan RTX (Turing)||GeForce RTX 3090||GeForce RTX 3090 Ti
GeForce RTX 3090
|Price||$999 US||$1199 US||$1199 US||$2999 US||$2499 US||$1499 US||$1999 US
|Ultra Enthusiast Tier||GeForce GTX 980 Ti||GeForce GTX 980 Ti||GeForce GTX 1080 Ti||GeForce RTX 2080 Ti||GeForce RTX 2080 Ti||GeForce RTX 3080 Ti||GeForce RTX 3080 Ti|
|Price||$649 US||$649 US||$699 US||$999 US||$999 US||$1199 US||$1199 US|
|Enthusiast Tier||GeForce GTX 980||GeForce GTX 1080||GeForce GTX 1080||GeForce RTX 2080||GeForce RTX 2080 SUPER||GeForce RTX 3080 10 GB||GeForce RTX 3080 12 GB|
|Price||$549 US||$549 US||$549 US||$699 US||$699 US||$699 US||$999 US|
|High-End Tier||GeForce GTX 970||GeForce GTX 1070||GeForce GTX 1070||GeForce RTX 2070||GeForce RTX 2070 SUPER||GeForce RTX 3070 Ti|
GeForce RTX 3070
|GeForce RTX 3070 Ti 16 GB|
|Price||$329 US||$379 US||$379 US||$499 US||$499 US||$599|
|Mainstream Tier||GeForce GTX 960||GeForce GTX 1060||GeForce GTX 1060||GeForce GTX 1060||GeForce RTX 2060 SUPER|
GeForce RTX 2060
GeForce GTX 1660 Ti
GeForce GTX 1660 SUPER
GeForce GTX 1660
|GeForce RTX 3060 Ti|
GeForce RTX 3060 12 GB
|GeForce RTX 3060 Ti
GeForce RTX 3060 12 GB
|Price||$199 US||$249 US||$249 US||$249 US||$399 US|
|Entry Tier||GTX 750 Ti|
|GTX 950||GTX 1050 Ti|
|GTX 1050 Ti|
|GTX 1650 SUPER|
|GTX 1650 SUPER|
|$149 US||$139 US|