AMD Instinct MI355X vs AMD Radeon 890M Comparison

Head-to-Head Benchmarks

The recorded database contains no direct head-to-head benchmark comparisons between the AMD Instinct MI355X and the AMD Radeon 890M. The MI355X has no benchmark entries, an average benchmark score of 0, and no nearest rivals listed. The Radeon 890M, by contrast, has a substantial set of measured results across ten tests, with an average benchmark score of 9210 and a percentile ranking of 45 among all GPUs.

The Radeon 890M's strongest recorded result appears in Geekbench Vulkan, where it scores 40808, followed by 37254 in Geekbench OpenCL. Its Passmark G3D score of 8076 indicates general 3D performance, while the Passmark GPU Compute score of 4157 reflects compute workloads. The 890M's weakest measured results are in the older DirectX API tests: Passmark DirectX 9 scores 97, DirectX 11 scores 73, and DirectX 10 scores 33. The DirectX 12 test shows 37, and the 3DMark Steel Nomad DX12 test records 590.

Because the MI355X lacks any benchmark scores, the data cannot show a single numeric comparison between the two. The MI355X does hold a 50th percentile ranking versus all GPUs, which matches the 890M's 45th percentile closely, but this percentile is based on an average score of zero, making it a placeholder rather than a measured outcome.

Architecture Differences

The two accelerators diverge fundamentally in architecture. The MI355X uses the CDNA 4.0 architecture on the MI350 256CU chip, built on a 3 nm process at TSMC. The Radeon 890M uses RDNA 3.5 on the Strix Point chip, manufactured on a 4 nm node, also at TSMC. The MI355X belongs to the Instinct (MIx) generation, while the 890M is part of the Navi III IGP (Strix Point Mobile) generation.

Transistor counts differ dramatically. The MI355X packs 185,000 million transistors on a 2380 mm² die, yielding a density of 77.7 million transistors per mm². The 890M contains 34,000 million transistors on a 233 mm² die, with a higher density of 145.9 million per mm². The smaller node and larger die give the MI355X far more raw silicon, while the 890M's denser layout reflects its integrated design.

Clock behavior separates the two clearly. The MI355X runs a base clock of 1000 MHz with a boost of 2400 MHz. The 890M starts at 400 MHz base and boosts to 2900 MHz. The 890M's higher boost clock indicates a design tuned for burst workloads, while the MI355X relies on massive parallelism rather than clock speed.

Memory configurations are entirely different classes. The MI355X uses 288 GB of HBM3e across an 8192-bit bus, delivering 8.19 TB/s of bandwidth. Memory clock is 2000 MHz with 8 Gbps effective. The 890M uses system shared memory, with the size, type, bus width, and bandwidth all listed as system dependent. This makes the 890M's memory performance a function of the host platform, not a fixed specification.

The MI355X has 16384 shading units and 1024 texture mapping units, but zero ROPs and a pixel rate of 0 MPixel/s. Its texture rate reaches 2457.6 GTexel/s. The 890M has 1024 shading units, 64 TMUs, 32 ROPs, and 16 ray tracing cores. Its pixel rate is 92.80 GPixel/s and texture rate is 185.6 GTexel/s. The MI355X has no ray tracing cores listed, while the 890M includes them.

Compute throughput shows the scale gap. The MI355X delivers 78.64 TFLOPS for both FP32 and FP16, with a 1:1 ratio. The 890M delivers 5.939 TFLOPS for both FP32 and FP16, also 1:1. The MI355X is roughly 13 times higher in raw FP32 throughput, a decisive margin.

Power and physical specifications differ sharply. The MI355X has a TDP of 1400 W with a suggested PSU of 1800 W. It is an OAM module with no power connectors listed, no display outputs, and dimensions of 102 mm length and 165 mm width. The 890M has a 15 W TDP, is an IGP with no power connectors, display outputs described as portable device dependent, and no dimensions recorded.

API support separates them completely. The MI355X lists DirectX, OpenGL, and Vulkan as N/A. The 890M supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The MI355X is not designed for graphics rendering APIs, while the 890M is fully featured for consumer graphics.

The Verdict

The data supports a straightforward split. The MI355X is a compute accelerator with no graphics API support, no display outputs, and no measured benchmarks. It targets workloads that require enormous memory bandwidth, 288 GB of HBM3e, and massive FP32 throughput at 78.64 TFLOPS. Its 1400 W TDP and OAM form factor indicate a data center or server installation, not a desktop or mobile part.

The Radeon 890M is an integrated GPU for portable devices. It has full DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4 support, 16 ray tracing cores, and a 15 W TDP. Its measured benchmarks place it in the 45th percentile among all GPUs, with an average score of 9210 and nearest rivals like the AMD Radeon Vega 8 at 9221 (-0.1% delta) and the NVIDIA GeForce GTX 960 at 9273 (-0.7% delta). These rivals are within 1% of the 890M's average score, indicating the 890M sits in a competitive mid-range integrated segment.

The MI355X's 50th percentile is misleading because it has no benchmark data. The database records no wins for either part in head-to-head tests, as none exist. The choice between them is not a performance question but a workload question. The MI355X serves compute-heavy, graphics-free environments. The 890M serves graphics and general computing in power-constrained mobile systems.

FAQ

Q: Which GPU has higher FP32 compute performance?

A: The MI355X delivers 78.64 TFLOPS FP32, while the 890M delivers 5.939 TFLOPS FP32. The MI355X is approximately 13 times higher.

Q: Does the MI355X support DirectX 12 Ultimate?

A: No. The MI355X lists DirectX as N/A. The 890M supports DirectX 12 Ultimate (12_2).

Q: How much memory does the Radeon 890M have?

A: The 890M uses system shared memory. Size, type, bus width, and bandwidth are all system dependent, meaning they vary by host platform.

Q: What is the TDP of each GPU?

A: The MI355X has a TDP of 1400 W with a suggested PSU of 1800 W. The 890M has a TDP of 15 W.

Q: Which GPU has ray tracing cores?

A: The 890M has 16 ray tracing cores. The MI355X lists no ray tracing cores.

Q: How does the 890M compare to its nearest rivals?

A: The 890M's average benchmark score is 9210. The AMD Radeon Vega 8 scores 9221 (-0.1%), the NVIDIA GeForce GTX 960 scores 9273 (-0.7%), the NVIDIA GeForce GTX 465 scores 9294 (-0.9%), and the NVIDIA GeForce GTX 850M scores 9302 (-1%). All are within one percent.

Where Each One Wins

The MI355X wins on every raw compute and memory metric recorded. Its 288 GB HBM3e memory with 8.19 TB/s bandwidth dwarfs the 890M's system shared memory. Its 78.64 TFLOPS FP32 throughput is an order of magnitude higher. Its 2457.6 GTexel/s texture rate exceeds the 890M's 185.6 GTexel/s by more than 13 times. The MI355X also uses a smaller 3 nm process node versus the 890M's 4 nm, and carries 185,000 million transistors versus 34,000 million.

The 890M wins on graphics features and practical deployment. It supports DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, while the MI355X has no API support. The 890M has 32 ROPs and a 92.80 GPixel/s pixel rate, while the MI355X has zero ROPs and 0 MPixel/s. The 890M's 15 W TDP allows use in portable devices, while the MI355X requires a 1400 W power envelope and an OAM module form factor. The 890M's 2900 MHz boost clock exceeds the MI355X's 2400 MHz boost.

Specification Differences

| Specification | AMD Instinct MI355X | AMD Radeon 890M |

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

| Architecture | CDNA 4.0 | RDNA 3.5 |

| Chip | MI350 256CU | Strix Point |

| Generation | Instinct (MIx) | Navi III IGP (Strix Point Mobile) |

| Process Node | 3 nm | 4 nm |

| Transistors | 185,000 million | 34,000 million |

| Die Size | 2380 mm² | 233 mm² |

| Transistor Density | 77.7M / mm² | 145.9M / mm² |

| Base Clock | 1000 MHz | 400 MHz |

| Boost Clock | 2400 MHz | 2900 MHz |

| Memory Size | 288 GB | System Shared |

| Memory Type | HBM3e | System Shared |

| Memory Bus Width | 8192 bit | System Shared |

| Memory Bandwidth | 8.19 TB/s | System Dependent |

| Shading Units | 16384 | 1024 |

| TMUs | 1024 | 64 |

| ROPs | 0 | 32 |

| Ray Tracing Cores | None listed | 16 |

| Pixel Rate | 0 MPixel/s | 92.80 GPixel/s |

| Texture Rate | 2457.6 GTexel/s | 185.6 GTexel/s |

| FP32 | 78.64 TFLOPS | 5.939 TFLOPS |

| FP16 | 78.64 TFLOPS (1:1) | 5.939 TFLOPS (1:1) |

| TDP | 1400 W | 15 W |

| Slot Width | OAM Module | IGP |

| Power Connectors | None | None |

| Suggested PSU | 1800 W | Not listed |

| Bus Interface | PCIe 5.0 x16 | PCIe 4.0 x8 |

| Display Outputs | No outputs | Portable Device Dependent |

| DirectX | N/A | 12 Ultimate (12_2) |

| OpenGL | N/A | 4.6 |

| Vulkan | N/A | 1.4 |

| Release Date | 2025-06-11 | 2024-07-14 |

| Production Status | Not listed | Active |

| Predecessor | Radeon Instinct | Navi II IGP |

| Launch MSRP | Not listed | Not listed |