NVIDIA GeForce RTX 4070 SUPER vs NVIDIA GeForce RTX 5070 SUPER Comparison

Where Each One Wins

The benchmark database shows a decisive split: the NVIDIA GeForce RTX 4070 SUPER wins the only recorded head-to-head test, the 3DMark Steel Nomad DX12 benchmark. In that test, the RTX 4070 SUPER scores 4627 against 2690 for the RTX 5070 SUPER, a 72% advantage. Across the broader recorded data, the RTX 4070 SUPER has 10 benchmark entries covering DirectX 10, 11, 12, and 9, plus OpenCL, Vulkan, and compute workloads. The RTX 5070 SUPER has only one recorded benchmark result, the 3DMark Steel Nomad DX12 test.

The RTX 4070 SUPER holds an average benchmark score of 43223 across all recorded tests, placing it in the 83rd percentile of all GPUs tracked. Its nearest rivals in the database include the NVIDIA Quadro M6000 24 GB at 43262 (0.1% higher), the NVIDIA GeForce RTX 5050 Mobile at 43268 (0.1% higher), the NVIDIA Quadro M6000 at 43301 (0.2% higher), and the NVIDIA GeForce RTX 4090 Mobile at 43667 (1% higher). These are close margins, indicating the RTX 4070 SUPER sits in a tightly packed performance band.

The RTX 5070 SUPER, in contrast, has an average benchmark score of 2690 from its single recorded test, placing it in the 18th percentile of all GPUs. Its nearest rivals are the NVIDIA Quadro K1100M at 2664 (1% lower), the NVIDIA GeForce GT 1030 at 2662 (1.1% lower), the Intel Arc Pro B50 at 2660 (1.1% lower), and the NVIDIA GeForce GT 440 at 2645 (1.7% lower). The data places the RTX 5070 SUPER in a much lower performance tier, despite its more recent generation and larger memory pool.

The recorded results indicate a stark performance gap between the two cards. The RTX 4070 SUPER delivers substantially higher frame rates in the Steel Nomad test, and its multi-test benchmark history provides a fuller picture of consistent performance across legacy and modern APIs. The RTX 5070 SUPER, with only one recorded test, shows a single data point that puts it far behind.

Architecture Differences

The two cards use different silicon from different generations. The RTX 4070 SUPER uses the AD104 chip built on Ada Lovelace architecture, while the RTX 5070 SUPER uses the GB205 chip built on Blackwell 2.0 architecture. Both are manufactured by TSMC on a 5 nm process, but the transistor counts and die sizes differ. The RTX 4070 SUPER has 35,800 million transistors on a 294 mm² die, yielding a transistor density of 121.8M per mm². The RTX 5070 SUPER has 31,100 million transistors on a 263 mm² die, yielding 118.3M per mm². The RTX 4070 SUPER therefore packs more transistors and a larger die.

The shading unit counts differ: the RTX 4070 SUPER has 7168 shading units, 224 texture mapping units, and 80 raster operations pipelines. The RTX 5070 SUPER has 6400 shading units, 200 TMUs, and 80 ROPs. The RTX 4070 SUPER also has 56 ray tracing cores and 224 tensor cores, while the RTX 5070 SUPER has 50 RT cores and 200 tensor cores. These counts favor the RTX 4070 SUPER across the board.

Clock speeds show a different pattern. The RTX 5070 SUPER has a higher base clock at 2325 MHz versus 1980 MHz for the RTX 4070 SUPER, and a higher boost clock at 2512 MHz versus 2475 MHz. However, the RTX 4070 SUPER still produces higher throughput figures. Its pixel rate is 198.0 GPixel/s versus 201.0 GPixel/s for the RTX 5070 SUPER, a slight edge for the newer card. The texture rates are 554.4 GTexel/s for the RTX 4070 SUPER versus 502.4 GTexel/s for the RTX 5070 SUPER. FP32 compute measures 35.48 TFLOPS for the RTX 4070 SUPER versus 32.15 TFLOPS for the RTX 5070 SUPER, and FP16 is identical at 1:1 ratios for both.

Memory configurations differ substantially. The RTX 4070 SUPER uses 12 GB of GDDR6X on a 192-bit bus with 504.2 GB/s bandwidth. The RTX 5070 SUPER uses 18 GB of GDDR7 on a 192-bit bus with 672.0 GB/s bandwidth. The newer card has more memory and faster memory technology, delivering 33% more bandwidth. The memory clock rates are 1313 MHz (21 Gbps effective) for the RTX 4070 SUPER versus 1750 MHz (28 Gbps effective) for the RTX 5070 SUPER.

The cards also differ in power and connectivity. The RTX 4070 SUPER has a 220 W TDP with a suggested PSU of 550 W, while the RTX 5070 SUPER has a 275 W TDP and no suggested PSU recorded. Both use a single 16-pin power connector and are dual-slot designs. The RTX 4070 SUPER uses PCIe 4.0 x16 while the RTX 5070 SUPER uses PCIe 5.0 x16. Display outputs differ: the RTX 4070 SUPER has 1x HDMI 2.1 and 3x DisplayPort 1.4a, while the RTX 5070 SUPER has 1x HDMI 2.1b and 3x DisplayPort 2.1b.

The Verdict

The recorded data clearly favors the RTX 4070 SUPER in raw benchmark performance. Its 4627 score in 3DMark Steel Nomad DX12 is 72% higher than the 2690 score of the RTX 5070 SUPER. The RTX 4070 SUPER also has a far richer benchmark history, with 10 recorded tests spanning multiple API generations, and it sits in the 83rd percentile of all GPUs. The RTX 5070 SUPER sits in the 18th percentile based on its single test.

The RTX 5070 SUPER offers advantages in memory capacity (18 GB versus 12 GB), memory bandwidth (672.0 GB/s versus 504.2 GB/s), and memory type (GDDR7 versus GDDR6X). It also has a newer PCIe interface (5.0 versus 4.0), newer display outputs (DisplayPort 2.1b versus 1.4a), and a higher boost clock (2512 MHz versus 2475 MHz). Its power draw is higher at 275 W versus 220 W.

However, the performance data does not support the RTX 5070 SUPER as a faster card. The single recorded benchmark places it far behind, and its nearest rivals in the database are low-end GPUs like the GT 1030 and GT 440. The RTX 4070 SUPER, by contrast, competes with the RTX 4090 Mobile and Quadro M6000 variants, all within 1% of its average score.

For users prioritizing benchmark performance, the RTX 4070 SUPER is the clear choice based on recorded data. For users prioritizing memory capacity and bandwidth, the RTX 5070 SUPER offers 18 GB and 672.0 GB/s, but the performance deficit in the recorded tests is substantial. The production status also differs: the RTX 4070 SUPER is end-of-life while the RTX 5070 SUPER is active.

FAQ

Q: Which card scores higher in 3DMark Steel Nomad DX12?

A: The NVIDIA GeForce RTX 4070 SUPER scores 4627 versus 2690 for the RTX 5070 SUPER, a 72% advantage.

Q: How much memory does each card have?

A: The RTX 4070 SUPER has 12 GB of GDDR6X on a 192-bit bus, while the RTX 5070 SUPER has 18 GB of GDDR7 on a 192-bit bus.

Q: What is the memory bandwidth difference?

A: The RTX 5070 SUPER delivers 672.0 GB/s versus 504.2 GB/s for the RTX 4070 SUPER.

Q: Which card has more shading units?

A: The RTX 4070 SUPER has 7168 shading units, while the RTX 5070 SUPER has 6400.

Q: What are the power consumption figures?

A: The RTX 4070 SUPER has a 220 W TDP, and the RTX 5070 SUPER has a 275 W TDP.

Q: How do the cards compare in overall percentile ranking?

A: The RTX 4070 SUPER sits in the 83rd percentile of all GPUs with an average score of 43223, while the RTX 5070 SUPER sits in the 18th percentile with an average score of 2690.

Head-to-Head Benchmarks

The only direct comparison in the database is the 3DMark Steel Nomad DX12 test. The RTX 4070 SUPER scores 4627, and the RTX 5070 SUPER scores 2690. The RTX 4070 SUPER wins by 72%, a substantial margin that dominates the head-to-head comparison.

Beyond the single direct test, the RTX 4070 SUPER has nine additional benchmark entries that the RTX 5070 SUPER lacks. These include Geekbench OpenCL at 172795, Geekbench Vulkan at 205624, Passmark DirectX 10 at 167, Passmark DirectX 11 at 273, Passmark DirectX 12 at 110, Passmark DirectX 9 at 344, Passmark G2D at 1184, Passmark G3D at 29995, and Passmark GPU Compute at 17108. The RTX 5070 SUPER has no recorded results for any of these tests.

The average benchmark scores reinforce the gap. The RTX 4070 SUPER averages 43223 across its 10 tests, while the RTX 5070 SUPER averages 2690 across its single test. The RTX 4070 SUPER's nearest rivals are all within 1% of its average score, including the RTX 4090 Mobile at 43667. The RTX 5070 SUPER's nearest rivals are the Quadro K1100M at 2664, GT 1030 at 2662, Arc Pro B50 at 2660, and GT 440 at 2645, all within 1.7% of its average.

The data shows that the RTX 4070 SUPER delivers superior performance in the recorded test suite. Its wins extend across API generations, from DirectX 9 through DirectX 12, plus OpenCL and Vulkan. The RTX 5070 SUPER's single result does not provide evidence of competitive performance in any workload.

Specification Differences

The two cards differ in several key specifications. The RTX 4070 SUPER uses the AD104 chip on Ada Lovelace architecture, while the RTX 5070 SUPER uses the GB205 chip on Blackwell 2.0 architecture. Both use a 5 nm process from TSMC, but the RTX 4070 SUPER has 35,800 million transistors on a 294 mm² die, while the RTX 5070 SUPER has 31,100 million transistors on a 263 mm² die. Transistor density is 121.8M per mm² for the RTX 4070 SUPER and 118.3M per mm² for the RTX 5070 SUPER.

Clock speeds differ: the RTX 4070 SUPER runs at 1980 MHz base and 2475 MHz boost, while the RTX 5070 SUPER runs at 2325 MHz base and 2512 MHz boost. Memory clocks are 1313 MHz (21 Gbps effective) for the RTX 4070 SUPER and 1750 MHz (28 Gbps effective) for the RTX 5070 SUPER.

Core counts favor the RTX 4070 SUPER: 7168 shading units versus 6400, 224 TMUs versus 200, 80 ROPs versus 80, 56 RT cores versus 50, and 224 tensor cores versus 200. Throughput figures also favor the RTX 4070 SUPER in most measures: FP32 at 35.48 TFLOPS versus 32.15 TFLOPS, texture rate at 554.4 GTexel/s versus 502.4 GTexel/s. The pixel rate slightly favors the RTX 5070 SUPER at 201.0 GPixel/s versus 198.0 GPixel/s.

Memory differs in capacity and type: 12 GB GDDR6X for the RTX 4070 SUPER versus 18 GB GDDR7 for the RTX 5070 SUPER. Both use a 192-bit bus, but bandwidth is 504.2 GB/s versus 672.0 GB/s. Power draw is 220 W for the RTX 4070 SUPER and 275 W for the RTX 5070 SUPER. The RTX 4070 SUPER has a suggested PSU of 550 W, while the RTX 5070 SUPER has none recorded.

Physical dimensions vary: the RTX 4070 SUPER is 267 mm long, 112 mm tall, and 42 mm wide, while the RTX 5070 SUPER is 245 mm long, 115 mm tall, and 40 mm wide. Both are dual-slot with a single 16-pin power connector. The bus interface is PCIe 4.0 x16 for the RTX 4070 SUPER and PCIe 5.0 x16 for the RTX 5070 SUPER. Display outputs are HDMI 2.1 and DisplayPort 1.4a for the RTX 4070 SUPER, and HDMI 2.1b and DisplayPort 2.1b for the RTX 5070 SUPER. API support is identical: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The RTX 4070 SUPER launched with a 599 USD MSRP, while no launch MSRP is recorded for the RTX 5070 SUPER.