NVIDIA’s GeForce RTX 3070 graphics card benchmarks have leaked within Ashes of The Singularity benchmark and spotted by Rogame. The GeForce RTX 3070 launches next Thursday and will come with a $499 US price tag with NVIDIA claiming that it offers better performance than the RTX 2080 Ti.
NVIDIA GeForce RTX 3070 Graphics Card Benchmarks Leak Out, An RTX 2080 Ti Replacement For $499 US But With 8 GB Memory
From what we have seen officially, the NVIDIA GeForce RTX 3070 seems to be quite the performer at its $499 US price point however there are several factors that will determine how successful the card really is. NVIDIA has claimed that the GeForce RTX 3070 is faster than the GeForce RTX 2080 Ti however we don’t know if that applies to a certain number of titles (with RTX specific features ON) or all games in general. There’s also the matter of availability if NVIDIA can provide the RTX 3070 in enough quantities to meet the crazy high demand for its GeForce RTX 30 series cards or would we see a similar dilemma as the RTX 3090/3080 launch.
There’s also the Radeon RX 6000 series which will be unveiled just a day before the retail launch for the RTX 3070 graphics cards. Now the reviews for the Founders Edition variant are planned for the 27th of October but if AMD can show that they are offering a better product with better specs and value, then consumers would obviously be more inclined to wait and see what the Radeon team is cooking rather than jumping on to buy the GeForce RTX 3070.
This might in fact be a blessing in disguise for NVIDIA as there will be less demand for the card at launch, giving NVIDIA enough time to produce more chips to send out to AIBs by the time the RX 6000 series actually launches in the retail segment. All of this remains to be seen in the coming weeks but for now, we have this one interesting benchmark that has popped up in a game that initially favored AMD’s GPUs but ever since Pascal, the tides have turned in NVIDIA’s favor. This specific title is none other than Ashes of The Singularity.
In the benchmarks, the NVIDIA GeForce RTX 3070 scores formidable results. At 1080p and Ultra-Wide resolutions, the GeForce RTX 3070 seems to perform really well and is actually faster than the GeForce RTX 2080 Ti, even if by a 1-2% difference. However, moving to 4K, the card nets poor GPU performance than the RTX 2080 Ti. The performance falls by 2.5% compared to the RTX 2080 Ti that could be due to the lower VRAM buffer of just 8 GB versus the 11 GB of the RTX 2080 Ti. However, as stated before, in titles that utilize RTX features and take advantage of Ampere’s dual-path uarch design, we might see bigger gains but that remains to be evaluated in our own review for the card which should be available for you to read when the embargo lifts.
NVIDIA GeForce RTX 3070 Graphics Card Specifications – GA104 GPU & 8 GB GDDR6 Memory
At the heart of the NVIDIA GeForce RTX 3070 graphics card lies the GA104 GPU. The GA104 is one of the many Ampere GPUs that we will be getting on the gaming segment. The GA104 GPU is the second-fastest Ampere chip in the stack. The GPU is based on Samsung’s 8nm (N8) process node. The GPU measures at 392.5mm2 and features 17.4 Billion transistors which are almost 93% of the transistors featured on the TU102 GPU. At the same time, the GA104 GPU is almost half the size of the TU102 GPU which is an insane amount of density.
For the GeForce RTX 3070, NVIDIA has enabled a total of 46 SM units on its flagship which results in a total of 5888 CUDA cores. In addition to the CUDA cores, NVIDIA’s GeForce RTX 3070 also comes packed with next-generation RT (Ray-Tracing) cores, Tensor cores, and brand new SM or streaming multi-processor units.
In terms of memory, the RTX 3070 features 8 GB of GDDR6 memory. The GeForce RTX 3070 comes with memory at speeds of 14 Gbps. That along with a full uncut bus interface of 256-bit will deliver a cumulative bandwidth of 448 Gbps.
NVIDIA GeForce RTX 30 Series ‘Ampere’ Graphics Card Specifications:
| Graphics Card Name | NVIDIA GeForce RTX 3060 Ti | NVIDIA GeForce RTX 3070 | NVIDIA GeForce RTX 3080 | NVIDIA GeForce RTX 3090 |
|---|---|---|---|---|
| GPU Name | Ampere GA104-200 | Ampere GA104-300 | Ampere GA102-200 | Ampere GA102-300 |
| Process Node | Samsung 8nm | Samsung 8nm | Samsung 8nm | Samsung 8nm |
| Die Size | 395.2mm2 | 395.2mm2 | 628.4mm2 | 628.4mm2 |
| Transistors | 17.4 Billion | 17.4 Billion | 28 Billion | 28 Billion |
| CUDA Cores | 4864 | 5888 | 8704 | 10496 |
| TMUs / ROPs | 152 / 80 | 184 / 96 | 272 / 96 | 328 / 112 |
| Tensor / RT Cores | 152 / 38 | 184 / 46 | 272 / 68 | 328 / 82 |
| Base Clock | 1410 MHz | 1500 MHz | 1440 MHz | 1400 MHz |
| Boost Clock | 1665 MHz | 1730 MHz | 1710 MHz | 1700 MHz |
| FP32 Compute | 16 TFLOPs | 20 TFLOPs | 30 TFLOPs | 36 TFLOPs |
| RT TFLOPs | 32 TFLOPs | 40 TFLOPs | 58 TFLOPs | 69 TFLOPs |
| Tensor-TOPs | 192 TOPs | 163 TOPs | 238 TOPs | 285 TOPs |
| Memory Capacity | 8 GB GDDR6 | 8 GB GDDR6 | 10 GB GDDR6X | 24 GB GDDR6X |
| Memory Bus | 256-bit | 256-bit | 320-bit | 384-bit |
| Memory Speed | 14 Gbps | 14 Gbps | 19 Gbps | 19.5 Gbps |
| Bandwidth | 448 Gbps | 448 Gbps | 760 Gbps | 936 Gbps |
| TGP | 175W | 220W | 320W | 350W |
| Price (MSRP / FE) | $399 US | $499 US | $699 US | $1499 US |
| Launch (Availability) | 2nd December 2020 | 29th October 2020 | 17th September 2020 | 24th September 2020 |
For now, it looks like the GeForce RTX 3070 would be a great option for anyone who doesn’t own either the GeForce RTX 2080 Ti or the GeForce RTX 1080 Ti. You would get lower temperatures, a lower power consumption rating, and an upgrade to a faster graphics card for a much lower price point. (This isn’t our final conclusion or recommendation for the RTX 3070 graphics card).