AMD Radeon RX 9070 vs NVIDIA GeForce RTX 2070 SUPER Comparison

AMD
RADEON

AMD Radeon RX 9070

CORE STATE Navi 48
VRAM 16 GB
CLOCK SPEED 2520 MHz
TDP 220 W
BUS WIDTH 256 bit
ARCHITECTURE RDNA 4.0
nm
PROCESS 4 nm
LAUNCH DATE 2025
VS
NVIDIA
GEFORCE

GeForce RTX 2070 SUPER

CORE STATE TU104
VRAM 8 GB
CLOCK SPEED 1770 MHz
TDP 215 W
BUS WIDTH 256 bit
ARCHITECTURE Turing
nm
PROCESS 12 nm
LAUNCH DATE 2019

PERFORMANCE BENCHMARKS

3dmark_3dmark_steel_nomad_dx12
6,290
1,651
geekbench_opencl
131,539
83,358
geekbench_vulkan
58,705
90,637
passmark_directx_10
141
132
passmark_directx_11
281
151
passmark_directx_12
74
67
passmark_directx_9
343
223
passmark_g2d
1,280
878
passmark_g3d
25,381
18,169
passmark_gpu_compute
14,737
7,557

Analysis: AMD Radeon RX 9070 vs NVIDIA GeForce RTX 2070 SUPER

Head-to-Head Benchmarks

The recorded data shows a dominant result for the AMD Radeon RX 9070 in nine of the ten head-to-head benchmark comparisons. The most striking margin appears in the 3DMark Steel Nomad DX12 test, where the RX 9070 scores 6290 against the RTX 2070 SUPER's 1651, a 281% advantage. This is not a marginal generational step; it is a complete overhaul in raw DX12 throughput.

The compute-oriented tests reinforce the same story. In Passmark GPU Compute, the RX 9070 posts 14737 versus 7557, a 95% lead. Geekbench OpenCL shows a 57.8% delta (131539 vs 83358). Passmark DirectX 11 follows with a 86.1% gap (281 vs 151), and DirectX 9 shows a 53.8% gap (343 vs 223). Even the older DirectX 10 path favors the RX 9070, albeit more modestly, at 6.8% (141 vs 132). The 2D graphics test also leans AMD, 1280 vs 878, a 45.8% margin. The Passmark G3D score, which aggregates overall 3D performance, lands at 25381 for the RX 9070 versus 18169 for the RTX 2070 SUPER, a 39.7% advantage.

The single reverse result occurs in Geekbench Vulkan, where the RTX 2070 SUPER scores 90637 against the RX 9070's 58705, a 35.2% lead for NVIDIA. This is a notable outlier given the pattern elsewhere. Every other benchmark, from DX9 to DX12 to compute, falls in favor of the AMD part. The RTX 2070 SUPER's Vulkan strength suggests a specific API-level optimization that does not translate to the other test suites.

When placed against the broader database, the RX 9070 sits at the 69th percentile of all GPUs, with an average benchmark score of 23877. Its nearest rivals include the RTX 2080 SUPER (24170, delta -1.2%) and the RTX 3080 Mobile (23628, delta 1.1%). The RTX 2070 SUPER, by contrast, sits at the 65th percentile with an average score of 20282, and its nearest rivals are the Intel Arc B570 (20556, delta -1.3%) and the Intel Arc A750 (20582, delta -1.5%). The gap between the two cards in average score is roughly 3595 points, which is consistent with the head-to-head margins seen in most individual tests.

Architecture Differences

The architectural gap between these two products is vast, and the benchmark data reflects that. The RX 9070 is built on TSMC's 4 nm process node, while the RTX 2070 SUPER uses TSMC's 12 nm node. The transistor counts differ enormously: the RX 9070 packs 53,900 million transistors on a 357 mm² die, while the RTX 2070 SUPER has 13,600 million on a 545 mm² die. This translates to a transistor density of 151.0M per mm² for AMD versus 25.0M per mm² for NVIDIA. The RX 9070 achieves roughly six times the density, which is a direct outcome of the newer fabrication process.

The RDNA 4.0 architecture on the RX 9070 uses a different design philosophy than the Turing architecture on the RTX 2070 SUPER. The RX 9070 features 3584 shading units, 224 texture mapping units, and 128 raster operation units. The RTX 2070 SUPER has 2560 shading units, 160 TMUs, and 64 ROPs. The RX 9070 also carries 56 ray tracing cores, whereas the RTX 2070 SUPER has 40. Notably, the RTX 2070 SUPER includes 320 tensor cores, a feature the RX 9070 does not list. The RX 9070's FP32 throughput is 36.13 TFLOPS, and its FP16 is identical at 36.13 TFLOPS (1:1 ratio). The RTX 2070 SUPER delivers 9.062 TFLOPS FP32 but 18.12 TFLOPS FP16 (2:1 ratio), meaning NVIDIA's card has a different compute balance between FP32 and FP16.

Memory configurations also diverge. The RX 9070 has 16 GB of GDDR6 on a 256-bit bus, achieving 644.6 GB/s bandwidth. The RTX 2070 SUPER has 8 GB of GDDR6 on the same 256-bit bus, but only 448.0 GB/s. The clock speeds tell a similar story: the RX 9070 boosts to 2520 MHz with a game clock of 2070 MHz, while the RTX 2070 SUPER boosts to only 1770 MHz. The RX 9070's memory runs at 2518 MHz (20.1 Gbps effective) versus 1750 MHz (14 Gbps effective) for the older card.

The bus interface differs as well: the RX 9070 uses PCIe 5.0 x16, while the RTX 2070 SUPER uses PCIe 3.0 x16. Display outputs also differ, with the RX 9070 offering one HDMI 2.1b and three DisplayPort 2.1a connectors, while the RTX 2070 SUPER offers one HDMI 2.0, three DisplayPort 1.4a, and one USB Type-C. Both cards are dual-slot, but the RX 9070 requires two 8-pin power connectors versus the RTX 2070 SUPER's single 6-pin plus single 8-pin. The RX 9070 has a 220 W TDP versus 215 W for the RTX 2070 SUPER, and both suggest a 550 W PSU.

Where Each One Wins

The RX 9070 wins in the vast majority of scenarios. For DX12 gaming at high resolutions, the Steel Nomad result (281% lead) suggests a massive advantage in modern, API-heavy workloads. The Passmark DX11 and DX9 wins (86.1% and 53.8%, respectively) indicate that older DirectX titles also run significantly faster on the AMD card. Compute workloads, as measured by Passmark GPU Compute and Geekbench OpenCL, are strongly in favor of the RX 9070, with margins of 95% and 57.8%. The 2D performance, often relevant for desktop compositing and multi-monitor setups, also favors the RX 9070 by 45.8%.

The RTX 2070 SUPER's only recorded win is in Geekbench Vulkan, where it leads by 35.2%. This suggests that for Vulkan-based applications, specifically those that heavily use the Vulkan API's low-level features, the NVIDIA card retains a measurable edge. The tensor cores on the RTX 2070 SUPER, while not directly benchmarked in the head-to-head list, may contribute to this Vulkan result in workloads that leverage those specialized units. However, the data does not show any other test where the NVIDIA card is competitive.

For users running a mix of DX9, DX10, DX11, DX12, and OpenCL workloads, the RX 9070 is the clear choice based on the recorded scores. For Vulkan-centric workloads, the RTX 2070 SUPER still holds a niche advantage. Given that the RX 9070 wins 9 out of 10 head-to-head tests, the overall performance profile is heavily skewed toward the AMD product.

The Verdict

The benchmark data is unambiguous: the AMD Radeon RX 9070 outperforms the NVIDIA GeForce RTX 2070 SUPER across nearly every measured category. The average benchmark score difference (23877 vs 20282) places the RX 9070 about 18% higher overall, and individual test deltas range from 6.8% to 281% in AMD's favor. The only exception is Vulkan, where the RTX 2070 SUPER leads by 35.2%.

For a user upgrading from an RTX 2070 SUPER, the RX 9070 offers a substantial performance uplift in DX12, DX11, DX9, OpenCL, and 2D workloads. The 16 GB memory capacity is double that of the RTX 2070 SUPER, and the bandwidth is roughly 44% higher, which matters for texture-heavy scenes and high-resolution rendering. The newer PCIe 5.0 interface and DisplayPort 2.1a outputs provide modern connectivity that the older card lacks.

The RTX 2070 SUPER remains relevant only for Vulkan-specific applications, where its 90637 score in Geekbench Vulkan beats the RX 9070's 58705. This could matter for users running specific Vulkan games or compute workloads that are optimized for NVIDIA's architecture. However, the rest of the data suggests that this is a narrow exception rather than a general trend.

The production status also differs: the RX 9070 is listed as Active, while the RTX 2070 SUPER is End-of-life. The release dates are far apart, with the RTX 2070 SUPER launching in July 2019 and the RX 9070 in March 2025. The RTX 2070 SUPER's successor is the GeForce 30 series, while the RX 9070 has no successor listed. For most buyers, the RX 9070 is the data-supported choice.

FAQ

Q: Which card wins in DirectX 12 performance?

A: The AMD Radeon RX 9070 wins decisively. In the 3DMark Steel Nomad DX12 test, it scores 6290 versus 1651 for the RTX 2070 SUPER, a 281% advantage. The Passmark DirectX 12 test also favors the RX 9070, 74 vs 67, a 10.4% lead.

Q: Is there any benchmark where the RTX 2070 SUPER beats the RX 9070?

A: Yes, in Geekbench Vulkan. The RTX 2070 SUPER scores 90637, while the RX 9070 scores 58705, giving the NVIDIA card a 35.2% lead. This is the only win for the RTX 2070 SUPER out of ten head-to-head tests.

Q: How do the memory capacities compare?

A: The RX 9070 has 16 GB of GDDR6, while the RTX 2070 SUPER has 8 GB. Both use a 256-bit bus, but the RX 9070 achieves a bandwidth of 644.6 GB/s versus 448.0 GB/s for the RTX 2070 SUPER.

Q: What is the average benchmark score for each card?

A: The RX 9070 has an average benchmark score of 23877, placing it at the 69th percentile of all GPUs. The RTX 2070 SUPER has an average score of 20282, placing it at the 65th percentile.

Q: Which card has a higher transistor density?

A: The RX 9070 has a transistor density of 151.0M per mm², while the RTX 2070 SUPER has 25.0M per mm². This is due to the RX 9070's 4 nm process node compared to the RTX 2070 SUPER's 12 nm node.

Q: What are the power requirements for each card?

A: The RX 9070 has a 220 W TDP and requires two 8-pin power connectors. The RTX 2070 SUPER has a 215 W TDP and requires one 6-pin plus one 8-pin connector. Both have a suggested PSU of 550 W.

Specification Differences

| Specification | AMD Radeon RX 9070 | NVIDIA GeForce RTX 2070 SUPER |

|---|---|---|

| Architecture | RDNA 4.0 | Turing |

| Process Node | 4 nm | 12 nm |

| Transistors | 53,900 million | 13,600 million |

| Die Size | 357 mm² | 545 mm² |

| Transistor Density | 151.0M / mm² | 25.0M / mm² |

| Base Clock | 1330 MHz | 1605 MHz |

| Boost Clock | 2520 MHz | 1770 MHz |

| Game Clock | 2070 MHz | None |

| Memory Clock | 2518 MHz (20.1 Gbps effective) | 1750 MHz (14 Gbps effective) |

| Memory Size | 16 GB | 8 GB |

| Memory Bandwidth | 644.6 GB/s | 448.0 GB/s |

| Shading Units | 3584 | 2560 |

| TMUs | 224 | 160 |

| ROPs | 128 | 64 |

| RT Cores | 56 | 40 |

| Tensor Cores | None listed | 320 |

| Pixel Rate | 322.6 GPixel/s | 113.3 GPixel/s |

| Texture Rate | 564.5 GTexel/s | 283.2 GTexel/s |

| FP32 Performance | 36.13 TFLOPS | 9.062 TFLOPS |

| FP16 Performance | 36.13 TFLOPS (1:1) | 18.12 TFLOPS (2:1) |

| TDP | 220 W | 215 W |

| Power Connectors | 2x 8-pin | 1x 6-pin + 1x 8-pin |

| Bus Interface | PCIe 5.0 x16 | PCIe 3.0 x16 |

| Display Outputs | 1x HDMI 2.1b, 3x DisplayPort 2.1a | 1x HDMI 2.0, 3x DisplayPort 1.4a, 1x USB Type-C |

| Production Status | Active | End-of-life |

| Release Date | 2025-03-05 | 2019-07-08 |

| Launch MSRP | 549 USD | 499 USD |

DETAILED SPECIFICATIONS

SPECIFICATION
RX 9070
RTX 2070 SUPER
Core Specs
Shading Units
3,584
2,560 -28.6%
Shaders
3,584
2,560 -28.6%
TMUs
224
160 -28.6%
ROPs
128
64 -50.0%
Compute Units
56
—
SM Count
—
40
Clocks
Base Clock
1330 MHz
1605 MHz
Boost Clock
2520 MHz
1770 MHz
Game Clock
2070 MHz
—
Memory Clock
2518 MHz 20.1 Gbps effective
1750 MHz 14 Gbps effective
Memory
Memory Size
16 GB
8 GB
VRAM (MB)
16,384
8,192 -50.0%
Memory Type
GDDR6
GDDR6
Memory Bus
256 bit
256 bit
Bandwidth
644.6 GB/s
448.0 GB/s
Cache
L1 Cache
—
64 KB (per SM)
L2 Cache
8 MB
4 MB
L3 Cache
64 MB
—
L0 Cache
32 KB per WGP
—
Performance
Pixel Rate
322.6 GPixel/s
113.3 GPixel/s
Texture Rate
564.5 GTexel/s
283.2 GTexel/s
FP32 (TFLOPS)
36.13 TFLOPS
9.062 TFLOPS
FP64 (TFLOPS)
1,129.0 GFLOPS (1:32)
283.2 GFLOPS (1:32)
FP16 (TFLOPS)
36.13 TFLOPS (1:1)
18.12 TFLOPS (2:1)
AI/RT
RT Cores
56
40 -28.6%
Tensor Cores
—
320
Matrix Cores
112
—
Power
TDP
220 W
215 W
TDP (W)
220
215 -2.3%
Suggested PSU
550 W
550 W
Power Connectors
2x 8-pin
1x 6-pin + 1x 8-pin
Architecture
Architecture
RDNA 4.0
Turing
GPU Name
Navi 48
TU104
Generation
Navi IV (RX 9000)
GeForce 20
Process Size
4 nm
12 nm
Transistors
53,900 million
13,600 million
Die Size
357 mm²
545 mm²
Foundry
TSMC
TSMC
Density
151.0M / mm²
25.0M / mm²
API Support
DirectX
12 Ultimate (12_2)
12 Ultimate (12_2)
OpenGL
4.6
4.6
Vulkan
1.4
1.4
OpenCL
2.2
3.0
CUDA
—
7.5
Shader Model
6.9
6.8
Physical
Slot Width
Dual-slot
Dual-slot
Length
—
267 mm 10.5 inches
Height
—
116 mm 4.6 inches
Outputs
1x HDMI 2.1b3x DisplayPort 2.1a
1x HDMI 2.03x DisplayPort 1.4a1x USB Type-C
Bus Interface
PCIe 5.0 x16
PCIe 3.0 x16
Other
Launch Price
549 USD
499 USD
Production
Active
End-of-life
Predecessor
Navi III
GeForce 10
Successor
—
GeForce 30
View Radeon RX 9070 Details View GeForce RTX 2070 SUPER Details