NVIDIA has just released the new GeForce RTX 2070 SUPER and RTX 2060 SUPER graphics cards, with the RTX 2080 SUPER to be released on a later date. If you’re updated with the latest news, this is no longer a surprise, since information has been leaked on the web regarding the SUPER graphics cards several days ago. The new GeForce RTX 20 Series SUPER cards doesn’t really offer anything new in the table. Except that they are merely a “SUPER” or bumped up specs version of their predecessors. TL;DR, the new RTX 2070 SUPER now performs somewhat on par with the RTX 2080, and the RTX 2060 SUPER performs almost on par with the RTX 2070. Why not call them Ti? Well, I don’t know. Perhaps “SUPER” sounds better than “Ti”? And besides, the “Ti” suffix has been around for many (many) years and it’s already an overused suffix. Go ahead and check out more details regarding their specs and some benchmarks below.
NVIDIA GeForce RTX 2070 SUPER and RTX 2060 SUPER Released
GeForce RTX 2060 SUPER in a Gist
Let’s talks about the GeForce RTX 2060 SUPER first. It’s based on the TU106 GPU, and comes with the same features that are available on the existing RTX 2060 like ray tracing and AI-enhance graphics. The core count has been increased in the RTX 2060 SUPER, memory bus has been expanded, and now has 8GB of GDDR6 memory.
According to NVIDIA (unfortunately we don’t have samples yet for day 1 review), the GeForce RTX 2060 SUPER delivers higher performance than the GeForce GTX 1080, and is also up to 22% faster (15% average) than the GeForce RTX 2060 GPU that was launched earlier this year at 2560×1440. The GeForce RTX 2060 SUPER is just 1% slower than the stock GeForce RTX 2070 on average overall at 1440p. Like I said, the new RTX 2060 SUPER is like the old RTX 2070.
GeForce RTX 2060 SUPER vs RTX 2060 vs GTX 1060 Specifications
Here is a specs comparison of the RTX 2060 SUPER (Turing) vs RTX 2060 (Turing) and GTX 1060 (Pascal).
| GPU | GeForce RTX 2060 SUPER (Turing) | GeForce RTX 2060 (Turing) | GeForce GTX 1060 (Pascal) |
|---|---|---|---|
| SMs | 34 | 30 | 10 |
| CUDA Cores | 2176 | 1920 | 1280 |
| Base Clock | 1470 MHz | 1365 MHz | 1506 MHz |
| GPU Boost Clock | 1650 MHz | 1680 MHz | 1708 MHz |
| FLOPS | 7+7 TOPS (7.2 TFLOPS FP32 / 7.2 TFLOPS INT32) | 6+6 TOPS (6.5 TFLOPS FP32 / 6.5 TFLOPS INT32) | 4.4 TFLOPS |
| Tensor FLOPS | 57.4 TFLOPS | 51.6 TFLOPS | N/A |
| Rays Cast | 6 Giga Rays | 5 Giga Rays | 0.4 Giga Rays |
| Texture Units | 136 | 120 | 80 |
| Texel fill-rate | 224.4 Gigatexels/sec | 201.6 Gigatexels/sec | 120.5 Gigatexels/sec |
| Memory Interface | 256-bit | 192-bit | 192-bit |
| Memory Data Rate | 14 Gbps | 14 Gbps | 8 Gbps |
| Memory Bandwidth | 448 GB/sec | 336.1 GB/sec | 192 GB/sec |
| Memory Size | 8 GB | 6 GB | 6 GB |
| Max L1 Cache Size | 2176 KB | 1920 KB | 480 KB |
| TGP | 175 Watts | 160 Watts | 120 Watts |
GeForce RTX 2070 SUPER in a Gist
Powering the GeForce RTX 2070 SUPER graphics card is a derivative of NVIDIA’s TU104 GPU. The RTX 2070 SUPER like a supercharged RTX 2070 that offers performance on par with an RTX 2080. According to NVIDIA, in terms of performance, the RTX 2070 SUPER is faster than NVIDIA’s former flagship GeForce GTX 1080 Ti GPU, and is also up to 24% faster (16% average) overall than the standard GeForce RTX 2070 at 2560×1440 resolution. The RTX 2080 is slightly faster than the GTX 1080 Ti; thus we can conclude that the performance of the RTX 2070 SUPER is similar to an GTX 1080 Ti (without ray tracing) and almost on par with the RTX 2080.
Like I said, we don’t have any samples yet, unfortunately. But theoretically, if NVIDIA’s claimed performance is true; that’s how I see things and I expect those kind of performance numbers from these new SUPER graphics cards.
GeForce RTX 2070 SUPER vs RTX 2070 vs GTX 1070 Specifications
Here is a specs comparison of the RTX 2070 SUPER (Turing) vs RTX 2070 (Turing) and GTX 1070 (Pascal).
| GPU | GeForce RTX 2070 SUPER (Turing) | GeForce RTX 2070 (Turing) | GeForce GTX 1070 (Pascal) |
|---|---|---|---|
| SMs | 40 | 36 | 15 |
| CUDA Cores | 2560 | 2304 | 1920 |
| Base Clock | 1605 MHz | 1410 MHz | 1506 MHz |
| GPU Boost Clock | 1770 MHz | 1620 MHz | 1683 MHz |
| FLOPS | 9+9 TOPS (9.1 TFLOPS FP32 / 9.1 TFLOPS INT32) | 7.5+7.5 TOPS (7.5 TFLOPS FP32 / 7.5 TFLOPS INT32) | 6.5 TFLOPS |
| Tensor FLOPS | 72 TFLOPS | 60 TFLOPS | N/A |
| Rays Cast | 7 Giga Rays | 6 Giga Rays | 0.6 Giga Rays |
| Texture Units | 184 | 144 | 120 |
| Texel fill-rate | 325.7 Gigatexels/sec | 233.3 Gigatexels/sec | 180.7 Gigatexels/sec |
| Memory Interface | 256-bit | 256-bit | 256-bit |
| Memory Data Rate | 14 Gbps | 14 Gbps | 8 Gbps |
| Memory Bandwidth | 448 GB/sec | 448 GB/sec | 256 GB/sec |
| Memory Size | 8 GB | 8 GB | 8 GB |
| Max L1 Cache Size | 2560 KB | 2304 KB | 720 KB |
| TGP | 215 Watts | 175 Watts | 150 Watts |
GeForce RTX 2070 SUPER vs RTX 2060 SUPER Specs Comparison
| RTX 2070 SUPER | RTX 2060 SUPER | |
|---|---|---|
| Graphics Processing Clusters | 5 or 6 | 3 |
| Texture Processing Clusters | 20 | 17 |
| Streaming Multiprocessors | 40 | 34 |
| CUDA Cores (single precision) | 2560 | 2176 |
| Tensor Cores | 320 | 272 |
| RT Cores | 40 | 34 |
| Texture Units | 160 | 136 |
| ROP Units | 64 | 64 |
| Rays Cast | 7 Giga Rays/second | 6 Giga Rays/second |
| Base Clock | 1605 MHz | 1470 MHz |
| Boost Clock | 1770 MHz | 1650 MHz |
| Memory Clock | 7000 MHz | 7000 MHz |
| Memory Data Rate | 14 Gbps | 14 Gbps |
| L2 Cache Size | 4096 K | 4096 K |
| Total Video Memory | 8192MB GDDR6 | 8192MB GDDR6 |
| Memory Interface | 256-bit | 256-bit |
| Total Memory Bandwidth | 448 GB/s | 448 GB/s |
| Texture Rate (Bilinear) | 283.2 GigaTexels/second | 246.2 GigaTexels/second |
| Fabrication Process | 12 nm FFN | 12 nm FFN |
| Transistor Count | 13.6 Billion | 10.8 Billion |
| Connectors | 3 x DisplayPort 1 x HDMI 1 x USB Type-C | 2 x DisplayPort 1 x Dual-Link DVI 1 x HDMI 1 x USB Type-C |
| Form Factor | Dual Slot | Dual Slot |
| Power Connectors | One 6-pin, One 8-pin | One 8-pin |
| Recommended Power Supply | 650 Watts | 550 Watts |
| Total Graphics Power (TGP) | 215 Watts | 175 Watts |
GeForce RTX 2070 SUPER and RTX 2060 SUPER Benchmark Comparison
RTX 2070 SUPER and RTX 2060 SUPER Pricing and Availability
The new GeForce RTX 2070 SUPER and RTX 2060 SUPER graphics cards will be available starting July 9, 2019. Custom SUPER cards from partners will become available starting July 9 as well or perhaps after the SUPER Founders Edition’s release. Meanwhile the RTX 2080 SUPER will be released later this month on July 23. MSRP are as follows:
- GeForce RTX 2080 SUPER starting at $699 USD
- GeForce RTX 2070 SUPER starting at $499 USD
- GeForce RTX 2060 SUPER starting at $399 USD
Clearly, NVIDIA is trying to disrupt AMD’s release of their NAVI graphics cards – the Radeon RX 5700 XT and RX 5700, due to be officially released this July 7. AMD doesn’t have anything to offer that will compete with anything above RTX 2080 level. The RX 5700 XT and RX 5700 were designed to compete with NVIDIA RTX 2070 and RTX 2060 graphics cards. It looks like it’s not looking good for the red team with the SUPER cards at this point. Let’s just wait and see, after all, most people will buy based on price / performance.
