AMD Instinct MI350X vs NVIDIA GeForce RTX 5080 Comparison

The Verdict

The AMD Instinct MI350X and NVIDIA GeForce RTX 5080 address entirely different workloads, and the recorded data makes that separation explicit. The RTX 5080 carries a full set of benchmark results across DirectX and compute tests, placing it in the 87th percentile of all GPUs in the database with an average benchmark score of 56,083. The MI350X has no recorded benchmark scores and sits at the 50th percentile, which reflects its absence from the consumer-oriented test suite rather than a performance deficiency.

The RTX 5080 is the only one of the two with measurable graphics performance data, and its nearest rivals confirm its positioning. It leads the AMD Radeon 8060S by 0.6%, the AMD Radeon RX 6750 GRE 12 GB by 0.7%, and the AMD Radeon Pro W5700X by 2.3%. The AMD Radeon RX 9070 GRE is 2.2% ahead of the RTX 5080. These margins are narrow, indicating that the RTX 5080 competes in a tightly packed segment of the database's graphics rankings.

The MI350X, by contrast, is an OAM module with no display outputs, no graphics API support, and a 1000 W TDP. Its data profile points to a compute accelerator with a 288 GB HBM3e memory pool, an 8192-bit bus, and 8.19 TB/s of bandwidth. The RTX 5080 uses 16 GB of GDDR7 on a 256-bit bus with 960.0 GB/s. These are not comparable products in any meaningful benchmark sense. The verdict from the data: the RTX 5080 serves graphics and general compute workloads with verified scores, while the MI350X targets memory-capacity-bound and bandwidth-bound acceleration tasks where its specifications are the only available metrics.

Architecture Differences

The two chips come from different foundry nodes and design philosophies. The MI350X uses a 3 nm process at TSMC with 185,000 million transistors on a 2380 mm² die, yielding a transistor density of 77.7M per mm². The RTX 5080 uses a 5 nm process at TSMC with 45,600 million transistors on a 378 mm² die, yielding a transistor density of 120.6M per mm². The MI350X is physically far larger and packs over four times the transistor count, but the RTX 5080 achieves a higher density per square millimeter.

The architecture names differ completely. The MI350X is built on CDNA 4.0, AMD's compute-focused instruction set architecture, while the RTX 5080 uses Blackwell 2.0, NVIDIA's graphics and compute hybrid. The MI350X has no graphics API support: DirectX, OpenGL, and Vulkan are all listed as N/A. The RTX 5080 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. This is the clearest architectural divergence; one chip is a pure compute accelerator, the other is a full-featured graphics processor.

The MI350X uses a chip design labeled "MI350 256CU" with 16,384 shading units and 1,024 texture mapping units. It has zero ROPs, no RT cores, no tensor cores, and a pixel rate of 0 MPixel/s. The texture rate is 2,252.8 GTexel/s. The RTX 5080 uses the GB203 chip with 10,752 shading units, 336 TMUs, 112 ROPs, 84 RT cores, and 336 tensor cores. Its pixel rate is 293.1 GPixel/s and its texture rate is 879.3 GTexel/s. The MI350X has more shading units and higher texture throughput, but the RTX 5080 has dedicated ray tracing and tensor hardware that the MI350X lacks entirely.

Clock behavior also separates the two. The MI350X runs at a 1000 MHz base and 2200 MHz boost, with memory at 2000 MHz (8 Gbps effective). The RTX 5080 runs at 2295 MHz base and 2617 MHz boost, with memory at 1875 MHz (30 Gbps effective). The RTX 5080's higher clocks compensate for its smaller shading unit count in some throughput calculations, but the MI350X's raw unit count drives its FP32 and FP16 figures.

Head-to-Head Benchmarks

The head-to-head benchmark table is empty, and the win counts are zero for both sides. The database records no direct comparisons between the MI350X and the RTX 5080. This absence is itself informative: the MI350X has no benchmark entries at all, while the RTX 5080 has ten recorded scores across multiple test suites.

The RTX 5080's benchmark results show a wide spread across different workloads. In 3DMark Steel Nomad DX12, it scores 8,637. In Geekbench OpenCL, it scores 235,901, and in Geekbench Vulkan, it scores 255,450. PassMark results range from 151 in DirectX 12 to 389 in DirectX 9, with a G3D score of 36,565 and a GPU compute score of 21,789. The PassMark G2D score of 1,415 and the DirectX 10 score of 208 complete the set. Its average benchmark score is 56,083.

The nearest rival data provides context for the RTX 5080's standing. The AMD Radeon RX 9070 GRE is 2.2% ahead with an average score of 57,367. The AMD Radeon 8060S trails by 0.6% at 55,757, the AMD Radeon RX 6750 GRE 12 GB trails by 0.7% at 55,698, and the AMD Radeon Pro W5700X trails by 2.3% at 54,828. These deltas are small, placing the RTX 5080 in a cluster where a few percentage points separate adjacent products.

For the MI350X, no such comparisons exist. The database records no scores, no rivals, and no percentile movement beyond its static 50th percentile placement. The only quantitative signals are its specification-derived figures: 72.09 TFLOPS FP32, 72.09 TFLOPS FP16, 2,252.8 GTexel/s texture rate, and 0 MPixel/s pixel rate. These numbers indicate a compute-oriented design with no rasterization pipeline, which explains the absence of graphics benchmarks.

Specification Differences

The two products differ across nearly every specification field. Starting with the chip and process: the MI350X uses the MI350 256CU chip on a 3 nm node, while the RTX 5080 uses the GB203 chip on a 5 nm node. Transistor counts are 185,000 million versus 45,600 million, and die sizes are 2380 mm² versus 378 mm². Transistor density favors the RTX 5080 at 120.6M per mm² versus 77.7M per mm².

Clocks diverge substantially. The MI350X has a 1000 MHz base and 2200 MHz boost. The RTX 5080 has a 2295 MHz base and 2617 MHz boost. Memory clocks are 2000 MHz (8 Gbps effective) for the MI350X and 1875 MHz (30 Gbps effective) for the RTX 5080. Memory capacity is the largest gap: 288 GB of HBM3e for the MI350X against 16 GB of GDDR7 for the RTX 5080. Bus widths are 8192 bit versus 256 bit, and bandwidth is 8.19 TB/s versus 960.0 GB/s.

Compute unit counts differ as well. The MI350X has 16,384 shading units, 1,024 TMUs, 0 ROPs, no RT cores, and no tensor cores. The RTX 5080 has 10,752 shading units, 336 TMUs, 112 ROPs, 84 RT cores, and 336 tensor cores. Pixel rates are 0 MPixel/s versus 293.1 GPixel/s. Texture rates are 2,252.8 GTexel/s versus 879.3 GTexel/s. FP32 and FP16 are both 72.09 TFLOPS for the MI350X and 56.28 TFLOPS for the RTX 5080.

Power and physical design also separate the products. The MI350X has a TDP of 1000 W and requires a 1400 W suggested PSU. The RTX 5080 has a TDP of 360 W and a 750 W suggested PSU. The MI350X is an OAM Module with no power connectors and no display outputs. The RTX 5080 is Dual-slot with one 16-pin connector and outputs including 1x HDMI 2.1b and 3x DisplayPort 2.1b. The RTX 5080 measures 304 mm by 137 mm by 40 mm, while the MI350X measures 102 mm by 165 mm with no height listed.

Release dates and generations also differ. The MI350X was released on 2025-06-11, predated by Radeon Instinct. The RTX 5080 was released on 2025-01-29, predated by GeForce 40 and succeeded by GeForce 60. The RTX 5080 is listed as Active in production status, while the MI350X has no production status recorded. The RTX 5080 has a launch MSRP of 999 USD; the MI350X has none recorded.

FAQ

Q: Which product has higher FP32 compute throughput?

A: The AMD Instinct MI350X lists 72.09 TFLOPS FP32, while the NVIDIA GeForce RTX 5080 lists 56.28 TFLOPS FP32. Both products also list identical FP16 figures at 72.09 TFLOPS and 56.28 TFLOPS respectively, with a 1:1 ratio.

Q: Does the MI350X support DirectX or Vulkan?

A: No. The MI350X lists DirectX, OpenGL, and Vulkan all as N/A. The RTX 5080 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Q: How does the RTX 5080 compare to its nearest rivals in the database?

A: The RTX 5080 has an average benchmark score of 56,083. It is 2.2% ahead of the AMD Radeon Pro W5700X, 0.7% ahead of the AMD Radeon RX 6750 GRE 12 GB, and 0.6% ahead of the AMD Radeon 8060S. The AMD Radeon RX 9070 GRE is 2.2% ahead of the RTX 5080.

Q: What memory configurations do the two products use?

A: The MI350X uses 288 GB of HBM3e on an 8192-bit bus with 8.19 TB/s bandwidth. The RTX 5080 uses 16 GB of GDDR7 on a 256-bit bus with 960.0 GB/s bandwidth.

Q: Which product has dedicated ray tracing hardware?

A: The RTX 5080 has 84 RT cores and 336 tensor cores. The MI350X has no RT cores and no tensor cores listed. The MI350X also has 0 ROPs and a pixel rate of 0 MPixel/s.

Q: What are the power requirements for each product?

A: The MI350X has a TDP of 1000 W and a suggested PSU of 1400 W. The RTX 5080 has a TDP of 360 W and a suggested PSU of 750 W. The MI350X uses no power connectors, while the RTX 5080 uses one 16-pin connector.