AMD Radeon RX 7700 vs NVIDIA GeForce RTX 5090 D V2 Comparison

AMD
RADEON

AMD Radeon RX 7700

CORE STATE Navi 32
VRAM 16 GB
CLOCK SPEED 2459 MHz
TDP 200 W
BUS WIDTH 256 bit
ARCHITECTURE RDNA 3.0
nm
PROCESS 5 nm
LAUNCH DATE 2025
VS
NVIDIA
GEFORCE

GeForce RTX 5090 D V2

CORE STATE GB202
VRAM 24 GB
CLOCK SPEED 2407 MHz
TDP 575 W
BUS WIDTH 384 bit
ARCHITECTURE Blackwell 2.0
nm
PROCESS 5 nm
LAUNCH DATE 2025

PERFORMANCE BENCHMARKS

3dmark_3dmark_steel_nomad_dx12
2,865
16,504
geekbench_opencl
106,028
N/A
passmark_directx_10
92
N/A
passmark_directx_11
186
N/A
passmark_directx_12
96
N/A
passmark_directx_9
261
N/A
passmark_g2d
1,206
N/A
passmark_g3d
20,974
N/A
passmark_gpu_compute
10,963
N/A

Analysis: AMD Radeon RX 7700 vs NVIDIA GeForce RTX 5090 D V2

The GeForce RTX 5090 D V2 and the Radeon RX 7700 occupy vastly different tiers of the graphics card market, and the benchmark data reflects a decisive performance gap. In the single shared 3DMark Steel Nomad DX12 test, the NVIDIA card posts a score of 16504 against AMD’s 2865, a difference of 476.1% in favor of the RTX 5090 D V2. This is not a close contest; it is a generational and segment-defining margin that places the two cards in entirely separate performance classes.

Head-to-Head Benchmarks

The only direct comparison available in the data is the 3DMark Steel Nomad DX12 benchmark, and it is overwhelmingly one-sided. The NVIDIA GeForce RTX 5090 D V2 achieves a score of 16504, while the AMD Radeon RX 7700 manages 2865. The deltaPct of 476.1% means the NVIDIA card delivers nearly five times the performance of the AMD card in this specific test. This is the largest possible win in the dataset, and it sets the tone for every other consideration.

Beyond the direct head-to-head, the average benchmark scores reinforce the gap. The RTX 5090 D V2 has an average benchmark score of 16504, identical to its single 3DMark result, placing it in the 59th percentile of all GPUs. The RX 7700, despite having nine benchmark entries, averages 15852, which puts it in the 58th percentile — just one percentile point lower. This is a crucial nuance: while the Steel Nomad score difference is enormous, the overall percentile rankings are nearly identical, suggesting the AMD card performs competitively in other workloads not covered by the single shared test.

The nearest rivals for the RTX 5090 D V2 are the NVIDIA T400 (avg score 16508, 0% delta), the AMD Radeon PRO W7500 (16415, 0.5% behind), and the NVIDIA RTX PRO 6000 Blackwell (16408, 0.6% behind). The RX 7700’s rivals include the AMD Radeon R9 370X (15862, 0.1% ahead), the AMD Radeon RX 9060 (16014, 1% behind), and the NVIDIA GeForce RTX 3060 Ti (16129, 1.7% behind). These clusters show that each card is competitive within its own performance neighborhood, but those neighborhoods are far apart.

Where Each One Wins

The RTX 5090 D V2 wins the only benchmark where both cards appear, and it does so by a margin that leaves no room for interpretation. In 3DMark Steel Nomad DX12, the NVIDIA card’s 16504 score versus 2865 for the RX 7700 means it is the clear choice for any workload that stresses DirectX 12 rendering at high resolutions or with heavy effects. The data shows a 476.1% advantage, which translates to a card that can handle scenes the AMD card cannot approach.

The RX 7700, however, shows strength in the broader benchmark suite. Its average score of 15852 across nine tests — including PassMark DirectX 9 (261), DirectX 10 (92), DirectX 11 (186), DirectX 12 (96), G2D (1206), G3D (20974), and GPU Compute (10963) — suggests it is a more versatile card for legacy and compute-oriented tasks. The Geekbench OpenCL score of 106028 is particularly notable, indicating strong general-purpose compute performance that the single 3DMark test does not capture. While the RX 7700 loses the head-to-head, it wins on the breadth of its benchmark coverage, making it a better pick for mixed workloads that include older APIs or OpenCL compute.

The wins are asymmetric: NVIDIA wins the modern DX12 test decisively, while AMD wins the diversity of tests. For a user prioritizing a single, demanding modern game or renderer, the RTX 5090 D V2 is the only logical choice. For someone running a variety of applications, including legacy DirectX titles or compute tasks, the RX 7700’s multiple high scores in PassMark and Geekbench make it a more rounded option — albeit at a far lower absolute performance ceiling.

Architecture Differences

The architectural divide between these two cards is stark. The RTX 5090 D V2 uses NVIDIA’s Blackwell 2.0 architecture on the GB202 chip, built on a 5 nm process at TSMC with 92,200 million transistors on a 750 mm² die. The RX 7700 uses AMD’s RDNA 3.0 architecture on the Navi 32 chip, also on a 5 nm TSMC process, but with just 28,100 million transistors on a 346 mm² die. The transistor density tells the story: NVIDIA packs 122.9 million transistors per mm², while AMD manages 81.2 million. This density advantage translates directly into the RTX 5090 D V2’s massive 21760 shading units, 680 TMUs, and 176 ROPs, compared to the RX 7700’s 2560 shading units, 160 TMUs, and 96 ROPs.

Memory configurations diverge sharply. The RTX 5090 D V2 comes with 24 GB of GDDR7 on a 384-bit bus, delivering 1.34 TB/s of bandwidth. The RX 7700 offers 16 GB of GDDR6 on a 256-bit bus, with 624.1 GB/s. The NVIDIA card’s memory bandwidth is more than double, which is critical for high-resolution textures and ray tracing workloads. The RX 7700’s 16 GB is still ample for most games, but the bus width and memory type put it at a clear disadvantage in bandwidth-bound scenarios.

Compute capabilities follow the same pattern. The RTX 5090 D V2 delivers 104.8 TFLOPS of FP32 and FP16 performance, along with 170 RT cores and 680 tensor cores. The RX 7700 offers 25.18 TFLOPS of FP32 and FP16, with 40 RT cores and no tensor cores listed. The NVIDIA card’s tensor cores are essential for AI features like DLSS, while the RX 7700 lacks that hardware entirely. Clock speeds are closer: the RTX 5090 D V2 boosts to 2407 MHz, while the RX 7700 boosts to 2459 MHz, with a game clock of 2041 MHz. Pixel rate (423.6 GPixel/s vs 236.1) and texture rate (1,636.8 GTexel/s vs 393.4) both favor NVIDIA by roughly 2x to 4x.

Power and interface differences are equally pronounced. The RTX 5090 D V2 has a TDP of 575 W, requires a 950 W PSU, and uses a single 16-pin connector. The RX 7700 has a 200 W TDP, needs a 550 W PSU, and uses two 8-pin connectors. The NVIDIA card is also larger at 304 mm in length versus 267 mm for AMD, and uses PCIe 5.0 x16 versus PCIe 4.0 x16. Both cards support DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, but the RTX 5090 D V2 offers three DisplayPort 2.1b outputs and one HDMI 2.1b, while the RX 7700 has two DisplayPort 2.1, one HDMI 2.1a, and a USB Type-C port.

FAQ

Q: Which card has a higher 3DMark Steel Nomad DX12 score?

A: The NVIDIA GeForce RTX 5090 D V2 scores 16504, which is 476.1% higher than the AMD Radeon RX 7700’s 2865 in the same test.

Q: What are the average benchmark scores for each card?

A: The RTX 5090 D V2 has an average benchmark score of 16504, while the RX 7700 averages 15852 across nine tests.

Q: How much memory does each card have, and what type?

A: The RTX 5090 D V2 has 24 GB of GDDR7 on a 384-bit bus, while the RX 7700 has 16 GB of GDDR6 on a 256-bit bus.

Q: Are there any tests where the RX 7700 scores higher than the RTX 5090 D V2?

A: No. The only shared benchmark is 3DMark Steel Nomad DX12, where the RTX 5090 D V2 wins. The RX 7700 has additional scores in PassMark and Geekbench that are not available for the NVIDIA card.

Q: What is the TDP difference between the two cards?

A: The RTX 5090 D V2 has a TDP of 575 W, while the RX 7700 has a TDP of 200 W.

Q: Do both cards support DirectX 12 Ultimate?

A: Yes, both the RTX 5090 D V2 and the RX 7700 list DirectX 12 Ultimate (12_2) in their API support.

The Verdict

The data is unambiguous for anyone building a high-end system: the NVIDIA GeForce RTX 5090 D V2 is the superior performer in the modern DirectX 12 workload measured, with a 476.1% lead over the RX 7700 in 3DMark Steel Nomad. Its 24 GB of GDDR7 memory, 104.8 TFLOPS of compute, and 170 RT cores put it in a different league entirely. The launch MSRP of 2,299 USD reflects that positioning, and the 575 W TDP with a 950 W PSU requirement makes it a serious commitment for any build.

The AMD Radeon RX 7700, however, is not without merit. Its average benchmark score of 15852 is within 4% of the RTX 5090 D V2’s 16504, and its percentile rank of 58 versus 59 shows it holds its own in the broader GPU landscape. The additional benchmark data — including a 106028 Geekbench OpenCL score and strong PassMark results across DirectX 9 through 12 — indicates a card that excels in compute and legacy workloads. Its 200 W TDP and 550 W PSU requirement make it far more accessible for mainstream systems.

The choice comes down to the workload. If the priority is maximum performance in a single modern benchmark or game, the RTX 5090 D V2 is the only option supported by the data. If the priority is versatility across a range of applications, particularly OpenCL compute or older DirectX titles, the RX 7700 offers a more balanced profile at a fraction of the power draw. The RTX 5090 D V2 wins on raw power; the RX 7700 wins on efficiency and breadth. The data does not support any other conclusion.

DETAILED SPECIFICATIONS

SPECIFICATION
RX 7700
RTX 5090 D V2
Core Specs
Shading Units
2,560
21,760 +750.0%
Shaders
2,560
21,760 +750.0%
TMUs
160
680 +325.0%
ROPs
96
176 +83.3%
Compute Units
40
—
SM Count
—
170
Clocks
Base Clock
1900 MHz
2017 MHz
Boost Clock
2459 MHz
2407 MHz
Game Clock
2041 MHz
—
Shader Clock
2041 MHz
—
Memory Clock
2438 MHz 19.5 Gbps effective
1750 MHz 28 Gbps effective
Memory
Memory Size
16 GB
24 GB
VRAM (MB)
16,384
24,576 +50.0%
Memory Type
GDDR6
GDDR7
Memory Bus
256 bit
384 bit
Bandwidth
624.1 GB/s
1.34 TB/s
Cache
L1 Cache
128 KB per Array
128 KB (per SM)
L2 Cache
2 MB
96 MB
L3 Cache
64 MB
—
L0 Cache
32 KB per WGP
—
Performance
Pixel Rate
236.1 GPixel/s
423.6 GPixel/s
Texture Rate
393.4 GTexel/s
1,636.8 GTexel/s
FP32 (TFLOPS)
25.18 TFLOPS
104.8 TFLOPS
FP64 (TFLOPS)
786.9 GFLOPS (1:32)
1.637 TFLOPS (1:64)
FP16 (TFLOPS)
25.18 TFLOPS (1:1)
104.8 TFLOPS (1:1)
AI/RT
RT Cores
40
170 +325.0%
Tensor Cores
—
680
Matrix Cores
80
—
Power
TDP
200 W
575 W
TDP (W)
200
575 +187.5%
Suggested PSU
550 W
950 W
Power Connectors
2x 8-pin
1x 16-pin
Architecture
Architecture
RDNA 3.0
Blackwell 2.0
GPU Name
Navi 32
GB202
Codename
Wheat Nas
—
Generation
Navi III (RX 7000)
GeForce 50
Process Size
5 nm
5 nm
Transistors
28,100 million
92,200 million
Die Size
346 mm²
750 mm²
Foundry
TSMC
TSMC
Density
81.2M / mm²
122.9M / 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
—
12.0
Shader Model
6.9
6.9
Physical
Slot Width
Dual-slot
Dual-slot
Length
267 mm 10.5 inches
304 mm 12 inches
Height
135 mm 5.3 inches
137 mm 5.4 inches
Outputs
1x HDMI 2.1a2x DisplayPort 2.11x USB Type-C
1x HDMI 2.1b3x DisplayPort 2.1b
Bus Interface
PCIe 4.0 x16
PCIe 5.0 x16
Other
Launch Price
—
2,299 USD
Production
Active
Active
Predecessor
Navi II
GeForce 40
Successor
Navi IV
GeForce 60
View Radeon RX 7700 Details View GeForce RTX 5090 D V2 Details