AMD Instinct MI350X vs Intel Arc 140T Mobile Comparison

Head-to-Head Benchmarks

The database contains no head-to-head benchmark entries for the AMD Instinct MI350X versus the Intel Arc 140T Mobile. Both products show zero recorded benchmark scores, zero wins for either side, and no nearest rival comparisons. The average benchmark score for each is zero, and both sit at the 50th percentile among all GPUs in the database, which reflects the absence of measured performance data rather than parity in capability.

Without direct measurements, the raw specifications must carry the analytical weight. The AMD Instinct MI350X delivers 72.09 TFLOPS of FP32 compute, while the Intel Arc 140T Mobile delivers 4.813 TFLOPS. That is a 14.98x difference in raw single-precision throughput. In FP16, the MI350X maintains 72.09 TFLOPS at a 1:1 ratio, while the Arc 140T reaches 9.626 TFLOPS at a 2:1 ratio, meaning the MI350X leads by 7.49x in half-precision work.

Texture rate follows a similar pattern. The MI350X processes 2,252.8 GTexel/s against the Arc 140T's 150.4 GTexel/s, a 14.98x gap that mirrors the FP32 ratio. Pixel rate is the one metric where the Intel part posts a nonzero figure: 75.20 GPixel/s, while the MI350X shows 0 MPixel/s because it has no ROPs configured for rasterization. The MI350X is not designed to output pixels; it is a compute accelerator with no display outputs.

Memory bandwidth separates the two by an even wider margin. The MI350X accesses 288 GB of HBM3e across an 8192-bit bus, yielding 8.19 TB/s. The Arc 140T uses system shared memory with bandwidth labeled "System Dependent," which means its effective throughput varies with the host platform and cannot be expressed as a fixed number. In practice, shared memory on a mobile IGP will operate far below dedicated HBM3e, but the database does not provide a concrete figure for the Intel part.

Clock behavior also differs fundamentally. The MI350X runs at a 1000 MHz base and 2200 MHz boost, while the Arc 140T runs at 300 MHz base and 2350 MHz boost. The Intel part has a higher boost clock by 150 MHz, but it achieves that with 1024 shading units against 16,384 on the AMD part. The MI350X has 16x the shading units, 16x the TMUs, and the Arc 140T's 32 ROPs have no counterpart on the AMD accelerator.

The verdict from the recorded data is clear: no benchmarks exist to compare, and the specification deltas are so large that any direct performance contest would be lopsided. The MI350X is a 1000 W OAM module; the Arc 140T is a 35 W integrated GPU. The data confirms they occupy entirely different segments.

Architecture Differences

The MI350X uses the CDNA 4.0 architecture on AMD's Instinct (MIx) generation, built on a 3 nm TSMC process. The Arc 140T uses Xe-LPG+ architecture on Intel's Arc Graphics-M (Arrow Lake) generation, built on a 5 nm TSMC process. Both come from TSMC, but the node difference matters: 3 nm versus 5 nm.

Transistor counts and die sizes are only recorded for the MI350X. It packs 185,000 million transistors on a 2380 mm² die, giving a transistor density of 77.7 million per square millimeter. The Arc 140T's transistor count and die size are listed as unknown, so no density comparison is possible from the data.

The MI350X uses the MI350 256CU chip, which implies 256 compute units. The Arc 140T uses the Arrow Lake-H chip, an integrated solution. The MI350X has 16,384 shading units, 1,024 TMUs, and zero ROPs. The Arc 140T has 1,024 shading units, 64 TMUs, and 32 ROPs. The AMD part has no ray tracing cores listed, while the Intel part includes 8 RT cores.

Memory architecture diverges sharply. The MI350X uses 288 GB of HBM3e with an 8192-bit bus and 8.19 TB/s bandwidth. Memory clock is 2000 MHz with 8 Gbps effective. The Arc 140T uses system shared memory, meaning size, type, bus width, and bandwidth are all platform-dependent. The Intel part has no dedicated VRAM.

Power delivery reflects the intended use. The MI350X draws 1000 W TDP with a suggested PSU of 1400 W and no power connectors on the card itself, since it mounts as an OAM Module. The Arc 140T draws 35 W TDP and is an IGP with no separate power connector requirement. The MI350X connects via PCIe 5.0 x16; the Arc 140T uses the IGP bus interface.

API support differs completely. The MI350X lists DirectX, OpenGL, and Vulkan as N/A, confirming it is a compute-only accelerator. The Arc 140T supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, making it a full graphics solution. Display outputs on the MI350X are listed as "No outputs," while the Arc 140T's display outputs are "Portable Device Dependent," meaning they exist but vary by laptop design.

Physical dimensions are only recorded for the MI350X: 102 mm in length and 165 mm in width. The Arc 140T has no recorded dimensions because it is integrated into a processor package. Release dates show the Arc 140T launched on 2025-01-12, and the MI350X launched on 2025-06-11.

Where Each One Wins

The MI350X wins in every compute-oriented metric recorded in the database. FP32 throughput, FP16 throughput, texture rate, memory capacity, memory bandwidth, and memory bus width all favor the AMD accelerator by overwhelming margins. Its 288 GB of HBM3e and 8.19 TB/s bandwidth are designed for large model inference and training workloads that require massive data movement. The 16,384 shading units and 1,024 TMUs process dense matrix operations at scale.

The Arc 140T wins in every graphics-oriented metric. It has 32 ROPs and 8 ray tracing cores, both of which are absent on the MI350X. It outputs pixels at 75.20 GPixel/s, while the MI350X outputs none. It supports DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, while the MI350X supports none of these APIs. The Arc 140T also wins on power efficiency by a wide margin: 35 W versus 1000 W TDP, meaning it consumes 96.5% less power.

The clock speed comparison is nuanced. The Arc 140T boosts to 2350 MHz, which is 150 MHz higher than the MI350X's 2200 MHz boost. However, the MI350X has 16x the shading units, so the higher clock on the Intel part does not translate into higher aggregate throughput. The base clock difference is starker: 1000 MHz on the MI350X versus 300 MHz on the Arc 140T, a 3.33x gap.

Form factor determines usage. The MI350X is an OAM Module with no display outputs, designed for server racks and accelerator bays. The Arc 140T is an IGP integrated into a mobile processor, designed for laptops where space and power are constrained. The production status field lists the Arc 140T as "Active," while the MI350X has no production status recorded.

The data indicates that the MI350X is for data center compute tasks: AI training, scientific simulation, and high-performance computing. The Arc 140T is for consumer mobile graphics: gaming, media playback, and general GPU-accelerated applications. Neither product is a substitute for the other.

FAQ

Q: Which GPU has higher FP32 compute performance?

A: The AMD Instinct MI350X delivers 72.09 TFLOPS of FP32, while the Intel Arc 140T Mobile delivers 4.813 TFLOPS. The MI350X leads by 14.98x.

Q: Does the Intel Arc 140T support ray tracing?

A: Yes, the Arc 140T includes 8 ray tracing cores and supports DirectX 12 Ultimate (12_2), which includes ray tracing features. The MI350X has no ray tracing cores listed.

Q: How much memory does each GPU have?

A: The MI350X has 288 GB of HBM3e memory with an 8192-bit bus and 8.19 TB/s bandwidth. The Arc 140T uses system shared memory, so its capacity and bandwidth depend on the host system.

Q: What is the power consumption difference?

A: The MI350X has a TDP of 1000 W and requires a suggested PSU of 1400 W. The Arc 140T has a TDP of 35 W. The Arc 140T consumes 96.5% less power.

Q: Can the MI350X output video to a display?

A: No, the MI350X has no display outputs and lists DirectX, OpenGL, and Vulkan support as N/A. It is a compute-only accelerator. The Arc 140T supports DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4 with display outputs that vary by portable device.

Q: Which GPU has more shading units?

A: The MI350X has 16,384 shading units, while the Arc 140T has 1,024. The MI350X has 16x more shading units, along with 1,024 TMUs versus 64 on the Arc 140T.

The Verdict

The data supports only one conclusion: these products serve different markets and should be selected based on workload type, not direct comparison. The AMD Instinct MI350X is the choice for compute-heavy tasks that demand massive memory capacity and bandwidth. Its 288 GB of HBM3e and 8.19 TB/s bandwidth are unmatched by the Arc 140T's system shared memory. Its 72.09 TFLOPS of FP32 and FP16 compute make it suitable for AI and scientific workloads that can utilize a 1000 W OAM module in a server environment.

The Intel Arc 140T Mobile is the choice for mobile graphics workloads. Its 32 ROPs, 8 ray tracing cores, and full API support for DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4 enable gaming and general GPU acceleration in laptops. Its 35 W TDP fits within mobile power budgets, and its higher boost clock of 2350 MHz helps compensate for its smaller shading unit count.

The MI350X cannot display output, so it is useless for interactive graphics. The Arc 140T cannot match the MI350X's compute throughput by any recorded metric, so it is unsuitable for large-scale parallel compute. The absence of benchmark scores in the database means no real-world performance comparison exists, but the specification gaps are so large that any benchmark would likely confirm the segmentation.

Users with data center compute needs should select the MI350X. Users with mobile graphics needs should select the Arc 140T. The two products do not compete; they complement different system designs.

Specification Differences

| Specification | AMD Instinct MI350X | Intel Arc 140T Mobile |

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

| Architecture | CDNA 4.0 | Xe-LPG+ |

| Generation | Instinct (MIx) | Arc Graphics-M (Arrow Lake) |

| Process Node | 3 nm | 5 nm |

| Foundry | TSMC | TSMC |

| Transistors | 185,000 million | unknown |

| Die Size | 2380 mm² | unknown |

| Transistor Density | 77.7M / mm² | null |

| Base Clock | 1000 MHz | 300 MHz |

| Boost Clock | 2200 MHz | 2350 MHz |

| Memory Clock | 2000 MHz 8 Gbps effective | System Shared |

| 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 |

| RT Cores | null | 8 |

| Pixel Rate | 0 MPixel/s | 75.20 GPixel/s |

| Texture Rate | 2,252.8 GTexel/s | 150.4 GTexel/s |

| FP32 | 72.09 TFLOPS | 4.813 TFLOPS |

| FP16 | 72.09 TFLOPS (1:1) | 9.626 TFLOPS (2:1) |

| TDP | 1000 W | 35 W |

| Slot Width | OAM Module | IGP |

| Power Connectors | None | null |

| Suggested PSU | 1400 W | null |

| Bus Interface | PCIe 5.0 x16 | IGP |

| Display Outputs | No outputs | Portable Device Dependent |

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

| OpenGL | N/A | 4.6 |

| Vulkan | N/A | 1.4 |

| Dimensions | 102 mm 4 inches x 165 mm 6.5 inches | null |

| Production Status | null | Active |

| Release Date | 2025-06-11 | 2025-01-12 |

| Predecessor | Radeon Instinct | HD Graphics-M |