AMD Radeon RX 7650 GRE vs NVIDIA GeForce RTX 5090 Comparison

AMD
RADEON

AMD Radeon RX 7650 GRE

CORE STATE Navi 33
VRAM 8 GB
CLOCK SPEED 2695 MHz
TDP 170 W
BUS WIDTH 128 bit
ARCHITECTURE RDNA 3.0
nm
PROCESS 6 nm
LAUNCH DATE 2025
VS
NVIDIA
GEFORCE

GeForce RTX 5090

CORE STATE GB202
VRAM 32 GB
CLOCK SPEED 2407 MHz
TDP 575 W
BUS WIDTH 512 bit
ARCHITECTURE Blackwell 2.0
nm
PROCESS 5 nm
LAUNCH DATE 2025

PERFORMANCE BENCHMARKS

3dmark_3dmark_steel_nomad_dx12
2,336
18,355
geekbench_opencl
83,109
334,370
geekbench_vulkan
N/A
376,728
passmark_directx_10
N/A
226
passmark_directx_11
N/A
341
passmark_directx_12
N/A
185
passmark_directx_9
N/A
395
passmark_g2d
N/A
1,413
passmark_g3d
N/A
39,650
passmark_gpu_compute
N/A
26,756

Analysis: AMD Radeon RX 7650 GRE vs NVIDIA GeForce RTX 5090

Head-to-Head Benchmarks

The recorded database contains two direct benchmark comparisons between the AMD Radeon RX 7650 GRE and the NVIDIA GeForce RTX 5090, and the results are unambiguous. In the 3DMark Steel Nomad DX12 test, the RX 7650 GRE scores 2,336 points, while the RTX 5090 reaches 18,355 points. That represents a 87.3% deficit for the AMD card, meaning the NVIDIA GPU delivers roughly seven times the raw performance in this particular DirectX 12 workload. The gap is so large that the RX 7650 GRE's score is closer to a rounding error on the RTX 5090's output than to a competitive result.

The second recorded test, Geekbench OpenCL, shows a similar pattern. The AMD card scores 83,109, while the NVIDIA card scores 334,370, a 75.1% difference. This compute-oriented benchmark measures general-purpose GPU throughput, and the RTX 5090's advantage here is substantial but slightly less pronounced than in the 3DMark test. The RTX 5090 wins both head-to-head comparisons, giving it a 2-0 record in the direct matchups.

To contextualize these numbers, the database places the RX 7650 GRE in the 83rd percentile of all GPUs, while the RTX 5090 sits in the 92nd percentile. The average benchmark score tells an even clearer story: the RX 7650 GRE averages 42,723 points across all recorded tests, while the RTX 5090 averages 79,842 points, nearly double. This average includes multiple additional benchmarks recorded for the RTX 5090 that the AMD card does not have, such as PassMark DirectX 9 (395 points), DirectX 10 (226 points), DirectX 11 (341 points), DirectX 12 (185 points), G2D (1,413 points), G3D (39,650 points), and GPU Compute (26,756 points), plus a Geekbench Vulkan score of 376,728.

The nearest rivals for each card further illustrate the positioning. The RX 7650 GRE's closest competitors include the NVIDIA GeForce RTX 4070 SUPER with an average score of 43,223 (1.2% ahead), the NVIDIA Quadro M6000 24 GB at 43,262 (1.2% ahead), the NVIDIA GeForce RTX 5050 Mobile at 43,268 (1.3% ahead), and the NVIDIA Quadro M6000 at 43,301 (1.3% ahead). The RTX 5090's nearest rivals are the NVIDIA Tesla P100 PCIe 16 GB at 79,605 (0.3% behind), the Tesla P100 PCIe 12 GB at 79,396 (0.6% behind), the AMD Radeon RX 6850M XT at 78,940 (1.1% behind), and the AMD Radeon Pro Vega 64X at 80,959 (1.4% ahead). These groupings show that the RX 7650 GRE competes in the mid-range segment, while the RTX 5090 sits at the very top of the performance hierarchy, trading places only with workstation-grade accelerators.

FAQ

Q: Which card has the higher average benchmark score?

A: The NVIDIA GeForce RTX 5090 has an average benchmark score of 79,842, while the AMD Radeon RX 7650 GRE averages 42,723. The RTX 5090's average is roughly 87% higher.

Q: How do the two cards compare in 3DMark Steel Nomad DX12?

A: The RTX 5090 scores 18,355 points compared to the RX 7650 GRE's 2,336 points. The RTX 5090 wins by 87.3%, a margin that places the AMD card far behind in this DirectX 12 workload.

Q: What is the percentile ranking of each card among all GPUs?

A: The RX 7650 GRE is in the 83rd percentile, while the RTX 5090 is in the 92nd percentile. The RTX 5090 ranks higher, meaning it outperforms a larger share of all recorded GPUs.

Q: Which card has more graphics memory?

A: The RTX 5090 has 32 GB of GDDR7 memory on a 512-bit bus with 1.79 TB/s bandwidth. The RX 7650 GRE has 8 GB of GDDR6 memory on a 128-bit bus with 288.0 GB/s bandwidth.

Q: What are the nearest rivals for each card in the database?

A: The RX 7650 GRE's closest rival is the NVIDIA GeForce RTX 4070 SUPER with a 1.2% higher average score. The RTX 5090's closest rival is the NVIDIA Tesla P100 PCIe 16 GB, which scores 0.3% lower.

Q: Which card has a higher transistor density?

A: The RTX 5090 has a transistor density of 122.9 million transistors per square millimeter, while the RX 7650 GRE has 65.2 million per square millimeter.

Architecture Differences

The two cards come from fundamentally different design philosophies. The AMD Radeon RX 7650 GRE uses the Navi 33 chip built on RDNA 3.0 architecture, codenamed Hotpink Bonefish, and belongs to the Navi III (RX 7000) generation. The NVIDIA GeForce RTX 5090 uses the GB202 chip built on Blackwell 2.0 architecture and belongs to the GeForce 50 generation. Both are manufactured by TSMC, but at different process nodes: the RX 7650 GRE uses a 6 nm process, while the RTX 5090 uses a 5 nm process. This node difference contributes to the RTX 5090's significantly higher transistor count of 92,200 million, compared to the RX 7650 GRE's 13,300 million. The die sizes also differ dramatically: 750 mm² for the RTX 5090 versus 204 mm² for the RX 7650 GRE. The transistor density reflects this: 122.9 million per square millimeter for NVIDIA versus 65.2 million for AMD.

The compute architectures diverge in their handling of specialized workloads. The RX 7650 GRE has 32 ray tracing cores and no tensor cores, relying on RDNA 3.0's unified shader design. The RTX 5090 has 170 ray tracing cores and 680 tensor cores, the latter dedicated to AI acceleration and deep learning tasks. The shading unit counts reflect the scale difference: the RX 7650 GRE has 2,048 shading units, while the RTX 5090 has 21,760, more than ten times as many. Texture mapping units also differ, with 128 on the AMD card versus 680 on the NVIDIA card, and render output units show 64 versus 176.

The memory architecture is another major separator. The RX 7650 GRE uses GDDR6 memory at 18 Gbps effective speed, while the RTX 5090 uses GDDR7 at 28 Gbps effective speed. The bus widths are 128-bit versus 512-bit, and the resulting bandwidth is 288.0 GB/s versus 1.79 TB/s. This bandwidth advantage allows the RTX 5090 to feed its much larger compute pipeline without bottlenecks. The RX 7650 GRE's memory clock is 2250 MHz, while the RTX 5090's memory clock is 1750 MHz, but the effective speed is higher on the NVIDIA card due to the GDDR7 standard's data rate capabilities.

The API support is identical for both cards: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The bus interfaces differ, with the RX 7650 GRE using PCIe 4.0 x8 and the RTX 5090 using PCIe 5.0 x16, giving the NVIDIA card twice the lanes at twice the transfer standard.

Specification Differences

The clock speeds reveal an interesting inversion. The RX 7650 GRE has a base clock of 1720 MHz and a boost clock of 2695 MHz, with a game clock of 2350 MHz. The RTX 5090 has a base clock of 2017 MHz and a boost clock of 2407 MHz. The AMD card actually boosts higher than the NVIDIA card, but this does not compensate for the massive difference in compute units. The RX 7650 GRE's higher boost clock is a consequence of its smaller, more power-efficient design.

The power draw differs substantially. The RX 7650 GRE has a TDP of 170 W and uses a single 8-pin power connector, with a suggested PSU of 450 W. The RTX 5090 has a TDP of 575 W, uses a 16-pin power connector, and requires a suggested PSU of 950 W. The physical dimensions also differ: the RX 7650 GRE is 204 mm long and 115 mm tall, while the RTX 5090 is 304 mm long, 137 mm tall, and 40 mm wide. Both are dual-slot cards.

The display outputs are similar but not identical. The RX 7650 GRE offers 1x HDMI 2.1a and 3x DisplayPort 2.1. The RTX 5090 offers 1x HDMI 2.1b and 3x DisplayPort 2.1b. The small version number differences indicate minor specification updates on the NVIDIA side. The pixel rates are 172.5 GPixel/s for the AMD card versus 423.6 GPixel/s for the NVIDIA card. Texture rates are 345.0 GTexel/s versus 1,636.8 GTexel/s. The FP32 and FP16 throughput are both 22.08 TFLOPS on the RX 7650 GRE and 104.8 TFLOPS on the RTX 5090, with both cards showing a 1:1 ratio between FP32 and FP16.

The production status for both cards is Active. The RX 7650 GRE released on 2025-02-06, a week after the RTX 5090's release on 2025-01-29. The AMD card's predecessor is Navi II and its successor is Navi IV, while the NVIDIA card's predecessor is GeForce 40 and its successor is GeForce 60. The launch MSRP of the RX 7650 GRE is 279 USD, while the RTX 5090 has a launch MSRP of 1,999 USD.

The Verdict

The benchmark data is clear: the NVIDIA GeForce RTX 5090 dominates the AMD Radeon RX 7650 GRE in every recorded head-to-head test. The 3DMark Steel Nomad DX12 result shows an 87.3% advantage for NVIDIA, and the Geekbench OpenCL result shows a 75.1% advantage. The average benchmark scores confirm this hierarchy, with the RTX 5090 at 79,842 versus 42,723 for the RX 7650 GRE. The RTX 5090 wins both direct comparisons, giving it a perfect 2-0 record.

The RTX 5090's nearest rivals are enterprise-class accelerators like the Tesla P100 series, while the RX 7650 GRE's nearest rivals are mainstream consumer cards like the RTX 4070 SUPER and the RTX 5050 Mobile. This positioning within the database's rival lists reinforces the conclusion that these two cards serve entirely different market tiers. The RX 7650 GRE is a capable entry point into the Radeon RX 7000 series, while the RTX 5090 represents the top of NVIDIA's GeForce 50 stack.

The architecture differences explain the performance gap. The RTX 5090 has 21,760 shading units versus 2,048, 680 tensor cores versus none, 170 ray tracing cores versus 32, and 32 GB of GDDR7 memory versus 8 GB of GDDR6. The process node advantage (5 nm versus 6 nm), the larger die (750 mm² versus 204 mm²), and the higher transistor count (92,200 million versus 13,300 million) all contribute to the NVIDIA card's superior throughput. The RTX 5090's 104.8 TFLOPS FP32 performance is nearly five times the RX 7650 GRE's 22.08 TFLOPS.

The power requirements reflect the performance gap. The RTX 5090 draws 575 W versus the RX 7650 GRE's 170 W, and the suggested PSU jumps from 450 W to 950 W. The physical size also increases, from a 204 mm card to a 304 mm card. These are practical considerations for system builders, but they do not change the fundamental performance relationship.

Where Each One Wins

The RX 7650 GRE wins in efficiency-oriented scenarios. Its 170 W TDP and 450 W suggested PSU make it suitable for systems with modest power delivery. Its 204 mm length fits into smaller cases, and its single 8-pin power connector is widely compatible. The 6 nm process node and 204 mm² die size indicate a design focused on cost-effective manufacturing. The card's 83rd percentile ranking shows it outperforms most GPUs in the database, even if it falls far short of the RTX 5090.

The RTX 5090 wins in every performance-oriented scenario recorded. The 3DMark Steel Nomad DX12 score of 18,355 versus 2,336 demonstrates a massive advantage in modern DirectX 12 gaming workloads. The Geekbench OpenCL score of 334,370 versus 83,109 shows superiority in general-purpose compute. The additional benchmarks recorded for the RTX 5090, including PassMark G3D at 39,650 and Geekbench Vulkan at 376,728, indicate strong performance across multiple API and workload types. The 32 GB memory capacity and 1.79 TB/s bandwidth make it suitable for large datasets and high-resolution textures.

The database's nearest rival data confirms the segmentation. The RX 7650 GRE trades places with the RTX 4070 SUPER, the Quadro M6000 variants, and the RTX 5050 Mobile, all within 1.3% of its average score. The RTX 5090 trades places with the Tesla P100 variants and the RX 6850M XT, all within 1.4%. Neither card's closest competitors overlap with the other's, showing that they occupy distinct performance tiers. The RTX 5090's 92nd percentile ranking versus the RX 7650 GRE's 83rd percentile places them nine percentile points apart, a significant gap in a database containing thousands of GPUs.

The RX 7650 GRE's strengths lie in its accessibility and efficiency, while the RTX 5090's strengths lie in raw performance and feature set. The recorded data does not indicate any benchmark where the AMD card outperforms the NVIDIA card. The wins column shows 0 for the RX 7650 GRE and 2 for the RTX 5090. For users requiring maximum compute throughput, ray tracing performance, or memory bandwidth, the RTX 5090 is the clear choice. For users prioritizing lower power draw, smaller physical footprint, and compatibility with standard power supplies, the RX 7650 GRE presents a reasonable option, but it cannot match the RTX 5090 in any measured performance category.

DETAILED SPECIFICATIONS

SPECIFICATION
RX 7650 GRE
RTX 5090
Core Specs
Shading Units
2,048
21,760 +962.5%
Shaders
2,048
21,760 +962.5%
TMUs
128
680 +431.3%
ROPs
64
176 +175.0%
Compute Units
32
SM Count
170
Clocks
Base Clock
1720 MHz
2017 MHz
Boost Clock
2695 MHz
2407 MHz
Game Clock
2350 MHz
Shader Clock
2350 MHz
Memory Clock
2250 MHz 18 Gbps effective
1750 MHz 28 Gbps effective
Memory
Memory Size
8 GB
32 GB
VRAM (MB)
8,192
32,768 +300.0%
Memory Type
GDDR6
GDDR7
Memory Bus
128 bit
512 bit
Bandwidth
288.0 GB/s
1.79 TB/s
Cache
L1 Cache
128 KB per Array
128 KB (per SM)
L2 Cache
2 MB
96 MB
L3 Cache
32 MB
L0 Cache
32 KB per WGP
Performance
Pixel Rate
172.5 GPixel/s
423.6 GPixel/s
Texture Rate
345.0 GTexel/s
1,636.8 GTexel/s
FP32 (TFLOPS)
22.08 TFLOPS
104.8 TFLOPS
FP64 (TFLOPS)
689.9 GFLOPS (1:32)
1.637 TFLOPS (1:64)
FP16 (TFLOPS)
22.08 TFLOPS (1:1)
104.8 TFLOPS (1:1)
AI/RT
RT Cores
32
170 +431.3%
Tensor Cores
680
Matrix Cores
64
Power
TDP
170 W
575 W
TDP (W)
170
575 +238.2%
Suggested PSU
450 W
950 W
Power Connectors
1x 8-pin
1x 16-pin
Architecture
Architecture
RDNA 3.0
Blackwell 2.0
GPU Name
Navi 33
GB202
Codename
Hotpink Bonefish
Generation
Navi III (RX 7000)
GeForce 50
Process Size
6 nm
5 nm
Transistors
13,300 million
92,200 million
Die Size
204 mm²
750 mm²
Foundry
TSMC
TSMC
Density
65.2M / 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
204 mm 8 inches
304 mm 12 inches
Height
115 mm 4.5 inches
137 mm 5.4 inches
Outputs
1x HDMI 2.1a3x DisplayPort 2.1
1x HDMI 2.1b3x DisplayPort 2.1b
Bus Interface
PCIe 4.0 x8
PCIe 5.0 x16
Other
Launch Price
279 USD
1,999 USD
Production
Active
Active
Predecessor
Navi II
GeForce 40
Successor
Navi IV
GeForce 60
View Radeon RX 7650 GRE Details View GeForce RTX 5090 Details