Intel Arc Graphics 128EU Mobile

The Intel Arc Graphics 128EU Mobile is an integrated GPU built on the Meteor Lake chip and the Xe-LPG architecture. It belongs to the Arc Graphics-M (Meteor Lake) generation, is manufactured by Intel on a 10 nm process, and is listed as Active in production. The release date is 2023-12-13, with HD Graphics-M listed as its predecessor and no successor recorded.

Memory Subsystem

The memory configuration is entirely system-shared. Memory size, memory type, bus width, and memory clock are all listed as "System Shared," and bandwidth is recorded as "System Dependent." There is no dedicated VRAM capacity and no fixed memory bus width in the data. The GPU relies on the host platform's system memory, so the effective bandwidth available to the graphics engine is determined by the platform rather than by the GPU itself.

This has direct implications for high-resolution workloads. A shared memory subsystem means frame-buffer data is allocated from system RAM, competing with CPU memory traffic. There is no fixed frame-buffer size, and the data does not specify whether a reserved portion of system memory is carved out for graphics. The reported bus width is also system-shared, meaning the memory interface is whatever the host implements. For high resolutions, where memory capacity and bandwidth demands rise, performance becomes dependent on the host's memory configuration and cannot be characterized from this GPU record alone. The bus interface is listed as Ring Bus, which ties the GPU to the processor's internal interconnect and reinforces the integrated, shared-memory design.

Power and Cooling

The TDP is 28 W, and the slot width is listed as IGP. No power connectors are recorded, and no suggested PSU value is listed. The data therefore provides no discrete power-supply requirement or auxiliary connector specification. As an integrated part, the GPU falls under the host platform's power delivery and thermal design rather than a separate add-in card regime.

The Ring Bus interface is the only bus information in the record. The absence of a discrete power connector list means the platform is expected to supply power through the processor socket and motherboard design. The data does not specify a cooler or a thermal solution, so cooling responsibilities rest with the host system. The 28 W TDP is the sole power-related number in the fact pack, and it is the figure that system designers can use as the GPU's thermal/power envelope.

Benchmark Performance

The benchmark array in the fact pack is empty, and the average benchmark score field is 0. As a result, there are no measured game scores, application scores, or workload results to analyze. The only relative rank present is percentileVsAllGpus, which is 50, placing this GPU at the midpoint of the database's all-GPU distribution. That percentile is not accompanied by any raw score, so it cannot be used to derive performance gaps to other parts.

The quantitative performance data that does exist consists of peak throughput and clock numbers. The base clock is 300 MHz, and the boost clock is 2250 MHz. The GPU contains 1024 shading units, 64 TMUs, and 32 ROPs. Peak FP32 throughput is 4.608 TFLOPS, while FP16 throughput is 9.216 TFLOPS, a 2:1 ratio relative to FP32. Texture rate is 144.0 GTexel/s, and pixel rate is 72.00 GPixel/s. These are theoretical peak rates, not actual benchmark results, and they should be read as upper bounds rather than expected application performance. The spread between the 300 MHz base clock and the 2250 MHz boost clock is substantial, but without benchmark entries the data cannot show how sustained clock behavior affects real workloads.

How It Compares

The nearestRivals list in the fact pack is empty. There are no rival GPU names, no rival scores, and no deltaPct values. Because of that, no exact percentage-difference comparison can be written, and no "30% ahead" or "25% behind" statements can be made from this record. The only comparative data point is the 50th percentile versus all GPUs, which indicates a median overall position in the database ranking, but it does not identify neighboring products or the distance between them.

No rival paragraphs can be constructed because there are no nearest rivals to analyze. The predecessor HD Graphics-M is listed, but the fact pack provides no benchmark scores for either that part or this one, so generational improvement cannot be quantified. The product is still Active in production, and no successor is listed. The empty benchmark and rival fields mean the database entry provides a specification record rather than a measured competitive evaluation.

Ray Tracing and Feature Set

The fact pack reports no RT core count and no tensor core count; both fields are null. Consequently, ray tracing acceleration and tensor-based or AI workload throughput cannot be quantified from this data. The record does not state whether dedicated ray tracing hardware exists.

The feature set is defined on the API side. DirectX 12 (12_1) is supported, along with OpenGL 4.6 and Vulkan 1.4. These are the API compatibility levels listed for the part. The underlying architecture is Xe-LPG, and the generation is Arc Graphics-M (Meteor Lake). Display outputs are described as "Portable Device Dependent," meaning the number and type of display connections are determined by the host portable device rather than by the GPU itself. Without RT core or tensor core entries, the ray tracing and machine-learning feature set remains unspecified in the available data.