Intel Iris Plus Graphics 640 Mobile

Intel Iris Plus Graphics 640 Mobile is an integrated graphics processor from Intel, built on the Kaby Lake GT3e chip and Generation 9.5 architecture. Fabricated on Intel's 14 nm++ process, this end-of-life mobile GPU occupies the 50th percentile among all GPUs in the database, placing it squarely in the middle of the performance distribution. The following analysis details its benchmark performance, memory characteristics, target usage scenarios, power requirements, and competitive positioning.

Benchmark Performance

The Intel Iris Plus Graphics 640 Mobile's average benchmark score is 0, which places it at the 50th percentile of all GPUs tracked. This positioning is unusual in that the 50th percentile typically represents the median performer, yet the raw score of 0 indicates that the database has no recorded synthetic benchmarks for this specific iGPU. Consequently, performance analysis relies on architectural specifications and clock-derived compute rates rather than direct comparison scores.

The GPU's raw compute throughput is derived from its 384 shading units operating at a boost clock of 1100 MHz, yielding a peak FP32 performance of 844.8 GFLOPS. This figure is modest by discrete GPU standards but respectable for an integrated solution from the Kaby Lake era. The texture rate of 52.80 GTexel/s, calculated from 48 TMUs, and the pixel rate of 6.600 GPixel/s, derived from 6 ROPs, further define its fillrate capabilities. These numbers suggest that the Iris Plus 640 Mobile is best suited for light to moderate graphical workloads rather than demanding gaming or compute tasks.

Without nearest rivals listed in the data, comparative percentage deltas cannot be calculated. However, the 50th percentile ranking implies that this GPU outperforms half of all GPUs in the database, which includes a wide range of both integrated and discrete parts. The FP16 performance of 1.690 TFLOPS (2:1 ratio) indicates support for reduced-precision compute, which can be leveraged in certain applications for improved throughput. The boost clock of 1100 MHz, up from a 300 MHz base, shows a significant thermal headroom for burst workloads, though sustained performance will depend on the host system's cooling solution.

Memory Subsystem

The memory configuration of the Intel Iris Plus Graphics 640 Mobile is entirely system-dependent. The VRAM size, type, and bus width are all listed as "System Shared," meaning the iGPU allocates from the host's main memory rather than having dedicated video memory. The bandwidth is similarly "System Dependent," which means actual performance will vary significantly based on the laptop's RAM configuration—dual-channel high-speed DDR4 will yield markedly better results than single-channel lower-speed modules.

This architecture imposes inherent limitations for high-resolution gaming. Because the GPU must share memory bandwidth with the CPU and other system processes, texture streaming and frame buffer operations compete for the same resources. For 1080p gaming, the shared memory approach can be adequate for older titles or esports games with reduced settings, but higher resolutions such as 1440p or 4K will quickly exhaust both bandwidth and capacity, leading to stuttering and reduced frame rates. The system's memory bandwidth becomes the primary bottleneck rather than the GPU's compute capacity.

The ROP count of 6 is notably low, which directly limits pixel throughput at high resolutions. This means that even if system memory bandwidth were sufficient, the pixel fillrate would cap performance in resolution-heavy scenarios. The shared memory architecture also means that the effective VRAM available to the GPU is dynamic, changing with the operating system's memory management and other application demands. Users should ensure their system has ample RAM—ideally 16 GB or more—to provide the iGPU with sufficient headroom.

Who Should Consider It

Given the benchmark data, the Intel Iris Plus Graphics 640 Mobile targets a narrow use case: lightweight, portable computing with occasional light gaming. The 844.8 GFLOPS of FP32 performance places it in the field of integrated graphics that can handle esports titles at 720p or 1080p with low to medium settings. Games with simple geometry and limited texture demands, such as older or indie titles, will be playable, while modern AAA releases will likely struggle to maintain playable frame rates even at 720p.

For productivity workloads, the GPU can accelerate video encoding and decoding, as well as basic image processing, but its 6 ROPs and shared memory limit its utility for professional creative work. Users who primarily need web browsing, office applications, and media consumption will find the performance adequate. The FP16 capability of 1.690 TFLOPS (2:1) may benefit machine learning inference tasks that support reduced precision, but this is a niche use case.

The 50th percentile ranking suggests that for the full spectrum of GPUs, this part sits in the middle—but within the integrated-only category, it is likely above average due to the GT3e configuration. However, the lack of dedicated VRAM and the system-dependent bandwidth mean that real-world performance is heavily influenced by the host laptop's memory configuration and thermal design. Users should not expect to run modern games at high detail levels; the sweet spot is 720p with low settings for moderately demanding titles.

Power and Cooling

The Thermal Design Power (TDP) is listed at 15 W, which is typical for a low-power integrated GPU designed for ultraportable laptops. This modest power envelope means that the iGPU can be cooled by passive solutions or small fans, contributing to thin and light laptop designs. The 300 MHz base clock ensures minimal power draw during idle or light tasks, while the 1100 MHz boost clock is available when performance is needed, subject to the system's thermal and power limits.

The slot width is specified as "IGP" (Integrated Graphics Processor), confirming that this is not a discrete card but rather embedded within the CPU package. There are no power connectors listed, as the GPU draws power from the motherboard's regulated supply designed for the entire processor. The data does not include a suggested PSU rating, which is consistent with an integrated solution that does not require a separate power supply unit.

Cooling requirements are inherently system-dependent, as the GPU shares thermal dissipation with the CPU on the same die. Laptop manufacturers must design cooling solutions that handle the combined TDP of both components. The 14 nm++ process node helps reduce leakage power, but sustained boost clocks will generate heat that requires adequate ventilation. Users should be aware that in thin-and-light chassis, sustained gaming sessions may trigger thermal throttling, reducing clocks below the 1100 MHz boost to maintain safe temperatures.

How It Compares

The FACT PACK lists no nearest rivals for the Intel Iris Plus Graphics 640 Mobile, which means a direct comparative analysis against specific competing GPUs is not possible from the provided data. The average benchmark score of 0 and the empty nearestRivals array indicate that the database has not yet populated comparative metrics for this part.

Given the 50th percentile placement, however, the iGPU can be positioned qualitatively. Compared to lower-tier integrated graphics with fewer execution units and lower clock speeds, the Iris Plus 640 Mobile's 384 shading units and 1100 MHz boost provide a clear advantage in shader throughput. The FP32 performance of 844.8 GFLOPS exceeds that of older integrated solutions but falls far short of entry-level discrete GPUs, which typically offer several times this compute capacity.

The 15 W TDP is a defining characteristic that separates this part from higher-performance mobile GPUs. Discrete solutions in the 35-50 W range will offer significantly higher performance but require more cooling and battery capacity. The Iris Plus 640 Mobile's advantage lies in its minimal power footprint, enabling fanless or near-silent operation in ultraportable devices. Its 6 ROPs and system-shared memory are its most limiting factors, particularly for high-resolution workloads. The 52.80 GTexel/s texture rate is adequate for 1080p but will show limitations with heavily textured scenes. In the absence of direct rival scores, the 50th percentile serves as a benchmark-neutral indicator of its position relative to the broader GPU landscape.