AMD Instinct MI455X vs NVIDIA GeForce RTX 4090 Mobile Comparison

AMD
RADEON

AMD Instinct MI455X

CORE STATE MI450 256CU
VRAM 432 GB
CLOCK SPEED 2400 MHz
TDP 2300 W
BUS WIDTH 24576 bit
ARCHITECTURE CDNA 5.0
nm
PROCESS 2 nm
LAUNCH DATE 2026
VS
NVIDIA
GEFORCE

GeForce RTX 4090 Mobile

CORE STATE AD103
VRAM 16 GB
CLOCK SPEED 1695 MHz
TDP 120 W
BUS WIDTH 256 bit
ARCHITECTURE Ada Lovelace
nm
PROCESS 5 nm
LAUNCH DATE 2023

PERFORMANCE BENCHMARKS

geekbench_opencl
N/A
180,831
geekbench_vulkan
N/A
170,774
passmark_directx_10
N/A
173
passmark_directx_11
N/A
262
passmark_directx_12
N/A
107
passmark_directx_9
N/A
310
passmark_g2d
N/A
984
passmark_g3d
N/A
27,212
passmark_gpu_compute
N/A
12,347

Analysis: AMD Instinct MI455X vs NVIDIA GeForce RTX 4090 Mobile

The AMD Instinct MI455X and NVIDIA GeForce RTX 4090 Mobile occupy entirely different segments of the hardware spectrum. The MI455X is an accelerator module with a 2300 W power envelope, built for compute density, while the RTX 4090 Mobile is a 120 W integrated graphics processor for laptops. The data shows that the RTX 4090 Mobile has a recorded benchmark presence, sitting in the 84th percentile of all GPUs, with an average benchmark score of 43667. The MI455X has no recorded benchmark scores in the database, placing it at the 50th percentile with an average score of zero. This difference in data availability shapes the analysis, as the comparison relies on architectural specifications rather than direct performance measurements.

Where Each One Wins

The NVIDIA GeForce RTX 4090 Mobile wins in every measurable benchmark category, simply because it is the only one of the two with recorded test results. Its strongest showing comes in the geekbench_opencl test, where it scores 180831. The geekbench_vulkan test also delivers a strong result at 170774. In the passmark suite, the RTX 4090 Mobile scores 27212 in passmark_g3d, 12347 in passmark_gpu_compute, 984 in passmark_g2d, 310 in passmark_directx_9, 262 in passmark_directx_11, 173 in passmark_directx_10, and 107 in passmark_directx_12. These numbers establish that the mobile GPU has a fully realized software stack and driver support across a range of APIs.

The AMD Instinct MI455X does not have any benchmark entries, so its wins cannot be expressed through recorded scores. However, the specification sheet indicates that its advantage lies in raw compute throughput and memory capacity. The MI455X has 32768 shading units, 1024 TMUs, and a texture rate of 2457.6 GTexel/s. Its FP32 and FP16 performance are both rated at 157.3 TFLOPS. These specifications dwarf the RTX 4090 Mobile, which has 9728 shading units, 304 TMUs, and an FP32 output of 32.98 TFLOPS. The MI455X also carries 432 GB of HBM4 memory on a 24576 bit bus, delivering a bandwidth of 23.3 TB/s. The RTX 4090 Mobile, by contrast, has 16 GB of GDDR6 memory on a 256 bit bus, with a bandwidth of 576.0 GB/s. In terms of memory capacity and bandwidth, the MI455X is the clear winner, though these are theoretical figures rather than measured results.

Architecture Differences

The two GPUs are built on different architectures and process nodes. The MI455X uses the CDNA 5.0 architecture with a chip designated as MI450 256CU, manufactured on a 2 nm process at TSMC. The RTX 4090 Mobile uses the Ada Lovelace architecture with an AD103 chip, also from TSMC, but on a 5 nm process. The MI455X has a large die size of 2990 mm², which is considerably larger than the RTX 4090 Mobile's 379 mm². The MI455X contains 320,000 million transistors, while the RTX 4090 Mobile contains 45,900 million. Interestingly, the transistor density of the RTX 4090 Mobile is higher at 121.1M per mm², compared to the MI455X's 107.0M per mm², despite the older process node.

Memory technology is a major divergence. The MI455X uses HBM4 memory with a 24576 bit bus, while the RTX 4090 Mobile uses GDDR6 with a 256 bit bus. The MI455X has no display outputs, no render output units, and no support for DirectX, OpenGL, or Vulkan, which indicates that it is not designed for graphics rendering. The RTX 4090 Mobile supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, and it has 76 RT cores and 304 tensor cores, which the MI455X lacks entirely. The RTX 4090 Mobile has 112 ROPs and a pixel rate of 189.8 GPixel/s, while the MI455X lists a pixel rate of 0 MPixel/s.

The MI455X uses a PCIe 6.0 x16 bus interface, while the RTX 4090 Mobile uses PCIe 4.0 x16. The MI455X is an EAM Module with no power connectors and a suggested PSU of 2700 W. The RTX 4090 Mobile is an IGP with no power connectors and no suggested PSU listed. The MI455X has a base clock of 1000 MHz and a boost clock of 2400 MHz, while the RTX 4090 Mobile has a base clock of 1335 MHz and a boost clock of 1695 MHz. The MI455X has a memory clock of 1900 MHz, which translates to 7.6 Gbps effective, while the RTX 4090 Mobile has a memory clock of 2250 MHz, or 18 Gbps effective.

Head-to-Head Benchmarks

The database contains no head-to-head benchmark entries for these two products, so there are no direct comparison scores to walk through. The only recorded measurements belong to the RTX 4090 Mobile, and they show a spread of results across different test types. The highest score is 180831 in geekbench_opencl, which is about 5.9% higher than the geekbench_vulkan score of 170774. In the passmark suite, passmark_g3d leads with 27212, followed by passmark_gpu_compute at 12347. The legacy DirectX tests show lower figures, with passmark_directx_9 at 310, passmark_directx_11 at 262, passmark_directx_10 at 173, and passmark_directx_12 at 107. The passmark_g2d score of 984 is a separate metric that measures 2D performance rather than 3D rendering.

The nearest rivals for the RTX 4090 Mobile, based on average benchmark score, are all NVIDIA professional cards. The NVIDIA RTX A6000 leads with an average score of 44075, which is 0.9% higher than the RTX 4090 Mobile. The NVIDIA Quadro M6000 sits at 43301, which is 0.8% lower. The NVIDIA GeForce RTX 5050 Mobile is at 43268, which is 0.9% lower, and the NVIDIA Quadro M6000 24 GB is at 43262, also 0.9% lower. These deltas are small, indicating that the RTX 4090 Mobile sits in a dense cluster of similarly scoring GPUs in the database. The MI455X has no nearest rivals listed, which further confirms that its performance profile has not been measured.

FAQ

Q: What is the average benchmark score for the AMD Instinct MI455X?

A: The MI455X has an average benchmark score of 0, with no recorded benchmark entries in the database.

Q: What is the average benchmark score for the NVIDIA GeForce RTX 4090 Mobile?

A: The RTX 4090 Mobile has an average benchmark score of 43667, and it sits in the 84th percentile of all GPUs.

Q: Which GPU has more memory bandwidth?

A: The AMD Instinct MI455X has a memory bandwidth of 23.3 TB/s, while the NVIDIA GeForce RTX 4090 Mobile has a memory bandwidth of 576.0 GB/s.

Q: Does the AMD Instinct MI455X support DirectX?

A: No, the MI455X lists DirectX support as N/A, along with OpenGL and Vulkan, and it has no display outputs.

Q: Which GPU has more shading units?

A: The AMD Instinct MI455X has 32768 shading units, while the NVIDIA GeForce RTX 4090 Mobile has 9728 shading units.

Q: What is the release date for each GPU?

A: The AMD Instinct MI455X has a release date of 2026-07-22, and the NVIDIA GeForce RTX 4090 Mobile has a release date of 2023-01-02.

Specification Differences

The two GPUs differ across nearly every specification field. The MI455X uses a 2 nm process node, while the RTX 4090 Mobile uses a 5 nm node, both from TSMC. The MI455X has 320,000 million transistors on a 2990 mm² die, while the RTX 4090 Mobile has 45,900 million transistors on a 379 mm² die. Transistor density is 107.0M per mm² for the MI455X and 121.1M per mm² for the RTX 4090 Mobile. The base clock is 1000 MHz for the MI455X and 1335 MHz for the RTX 4090 Mobile, while the boost clock is 2400 MHz for the MI455X and 1695 MHz for the RTX 4090 Mobile.

Memory configurations are starkly different. The MI455X has 432 GB of HBM4 on a 24576 bit bus with 23.3 TB/s bandwidth. The RTX 4090 Mobile has 16 GB of GDDR6 on a 256 bit bus with 576.0 GB/s bandwidth. The memory clock is 1900 MHz, or 7.6 Gbps effective, for the MI455X, and 2250 MHz, or 18 Gbps effective, for the RTX 4090 Mobile. The MI455X has 32768 shading units, 1024 TMUs, and 0 ROPs, while the RTX 4090 Mobile has 9728 shading units, 304 TMUs, and 112 ROPs. The MI455X has no RT cores or tensor cores, while the RTX 4090 Mobile has 76 RT cores and 304 tensor cores.

Compute rates are also far apart. The MI455X delivers 157.3 TFLOPS for both FP32 and FP16, with a texture rate of 2457.6 GTexel/s and a pixel rate of 0 MPixel/s. The RTX 4090 Mobile delivers 32.98 TFLOPS for both FP32 and FP16, with a texture rate of 515.3 GTexel/s and a pixel rate of 189.8 GPixel/s. The MI455X has a TDP of 2300 W and a suggested PSU of 2700 W, while the RTX 4090 Mobile has a TDP of 120 W and no suggested PSU listed. The MI455X is an EAM Module and the RTX 4090 Mobile is an IGP, and both use a PCIe x16 interface, with the MI455X on PCIe 6.0 and the RTX 4090 Mobile on PCIe 4.0. The MI455X has no display outputs, while the RTX 4090 Mobile has display outputs that are portable device dependent. The MI455X supports no graphics APIs, while the RTX 4090 Mobile supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The MI455X has no launch MSRP listed, and neither does the RTX 4090 Mobile.

DETAILED SPECIFICATIONS

SPECIFICATION
Instinct MI455X
RTX 4090 Mobile
Core Specs
Shading Units
32,768
9,728 -70.3%
Shaders
32,768
9,728 -70.3%
TMUs
1,024
304 -70.3%
ROPs
0
112 +∞%
Compute Units
256
—
SM Count
—
76
Clocks
Base Clock
1000 MHz
1335 MHz
Boost Clock
2400 MHz
1695 MHz
Memory Clock
1900 MHz 7.6 Gbps effective
2250 MHz 18 Gbps effective
Memory
Memory Size
432 GB
16 GB
VRAM (MB)
442,368
16,384 -96.3%
Memory Type
HBM4
GDDR6
Memory Bus
24576 bit
256 bit
Bandwidth
23.3 TB/s
576.0 GB/s
Cache
L1 Cache
32 KB (per CU)
128 KB (per SM)
L2 Cache
192 MB
64 MB
Performance
Pixel Rate
0 MPixel/s
189.8 GPixel/s
Texture Rate
2,457.6 GTexel/s
515.3 GTexel/s
FP32 (TFLOPS)
157.3 TFLOPS
32.98 TFLOPS
FP64 (TFLOPS)
2.458 TFLOPS (1:64)
515.3 GFLOPS (1:64)
FP16 (TFLOPS)
157.3 TFLOPS (1:1)
32.98 TFLOPS (1:1)
AI/RT
RT Cores
—
76
Tensor Cores
—
304
Matrix Cores
1,024
—
Power
TDP
2300 W
120 W
TDP (W)
2,300
120 -94.8%
Suggested PSU
2700 W
—
Power Connectors
None
None
Architecture
Architecture
CDNA 5.0
Ada Lovelace
GPU Name
MI450 256CU
AD103
Generation
Instinct (MIx)
GeForce 40 Mobile
Process Size
2 nm
5 nm
Transistors
320,000 million
45,900 million
Die Size
2990 mm²
379 mm²
Foundry
TSMC
TSMC
Density
107.0M / mm²
121.1M / mm²
API Support
DirectX
—
12 Ultimate (12_2)
OpenGL
—
4.6
Vulkan
—
1.4
OpenCL
3.0
3.0
CUDA
—
8.9
Shader Model
—
6.8
Physical
Slot Width
EAM Module
IGP
Outputs
No outputs
Portable Device Dependent
Bus Interface
PCIe 6.0 x16
PCIe 4.0 x16
Other
Production
—
Active
Predecessor
Radeon Instinct
GeForce 30 Mobile
Successor
—
GeForce 50 Mobile
View Instinct MI455X Details View GeForce RTX 4090 Mobile Details