AMD Radeon HD 7590M

The AMD Radeon HD 7590M, built on the Thames chip, is a mobile graphics processor from the London generation (HD 7500M). It uses the TeraScale 2 architecture and is fabricated by TSMC on a 40 nm process. The die integrates 716 million transistors across 118 mm², yielding a density of 6.1M / mm². Released on 2012-01-06, this part has reached end-of-life status. The database places it at the 50th percentile of all GPUs, with an average benchmark score of 0 and an empty benchmark array, meaning no synthetic scores are recorded. Positioned between the Vancouver and Solar System generations, the HD 7590M represents a specific point in AMD's mobile lineup.

Benchmark Performance

The HD 7590M's performance profile is defined by its theoretical throughput figures, as no direct benchmark scores are available. The FP32 compute rate is 576.0 GFLOPS, a modest figure for a mobile part. The pixel rate of 9.600 GPixel/s and texture rate of 14.40 GTexel/s further delineate its capabilities. With a 50th percentile ranking, the data indicates that this card sits exactly at the median of all GPUs in the database. This means half of all recorded parts are faster, and half are slower, though the absence of rival deltas prevents precise comparisons. The shading unit count of 480, along with 24 texture mapping units and 16 render output units, provides the architectural basis for these rates. The memory subsystem operates at 800 MHz with an effective data rate of 3.2 Gbps, delivering a bandwidth of 25.60 GB/s across a 64-bit bus. This bandwidth is a limiting factor, as the FP32 throughput of 576.0 GFLOPS is not matched by memory bandwidth. The 1024 MB GDDR5 frame buffer is adequate for older titles but constrained by the narrow bus. The 50th percentile ranking suggests that while not a top performer, it is not at the bottom either, making it a middle-of-the-road option in the historical GPU landscape. The theoretical fill rates, when compared to the compute output, indicate a balanced but low-end design, typical of an entry-level mobile part from its era.

Ray Tracing and Feature Set

The HD 7590M does not include dedicated ray tracing cores or tensor cores, as these are absent from the specification. The API support is limited to DirectX 11.2 (11_0) and OpenGL 4.4, with no Vulkan support listed. This places it firmly in the pre-ray-tracing era, where fixed-function pipelines and traditional rasterization dominated. The display outputs are 'Portable Device Dependent', indicating that the card is designed for laptops and other mobile devices, where the display connection is determined by the host system. The lack of Vulkan support means that modern titles relying on Vulkan for lower-level access will not run on this hardware. The TeraScale 2 architecture is the foundation, and the absence of tensor cores means no AI-accelerated features are available. The memory type is GDDR5, which was a high-bandwidth standard at the time, but the 64-bit bus width limits the effective bandwidth to 25.60 GB/s. The card's feature set is therefore limited to the DirectX 11.2 and OpenGL 4.4 feature levels, which were contemporary for its release date. No hardware-accelerated ray tracing or deep learning super sampling is present, so any such workload is entirely unsupported.

Power and Cooling

The thermal design power of the HD 7590M is rated at 18 W, an exceptionally low figure that implies minimal cooling requirements. No power connector specifications are provided, and no suggested PSU is listed in the data. This suggests that the card draws its power entirely from the motherboard slot, which is consistent with its mobile orientation. The bus interface is PCIe 2.0 x16, a standard that provides sufficient bandwidth for the card's capabilities. The low TDP of 18 W means that a simple cooling solution, such as a small heat sink or fan, is sufficient. The absence of a power connector requirement further simplifies integration into portable devices. The 40 nm process node contributes to the low power draw, as does the relatively small die size of 118 mm². The transistor density of 6.1M / mm² indicates a mature manufacturing process. Overall, the power and cooling profile is one of efficiency and simplicity, tailored for thin-and-light notebooks. The data shows no separate power connector or PSU recommendation, reinforcing the idea that this is a low-consumption part that can be powered by the system's existing infrastructure.

Who Should Consider It

The HD 7590M is positioned for users with modest graphics demands. Given its 50th percentile ranking and theoretical throughput of 576.0 GFLOPS, it is suitable for basic productivity, video playback, and legacy gaming at low resolutions. The 1024 MB GDDR5 memory, while limited, is enough for older titles that do not require large texture pools. The pixel rate of 9.600 GPixel/s and texture rate of 14.40 GTexel/s indicate that it can handle simple 3D scenes. However, the 64-bit memory bus and 25.60 GB/s bandwidth will bottleneck any attempt at high-detail or high-resolution gaming. The lack of Vulkan support further restricts compatibility with modern game engines. Users who expect to run contemporary AAA titles at high settings will be disappointed. Instead, this card is best suited for office applications, web browsing, and playing games from the early 2010s era. The end-of-life status means that driver updates are likely no longer provided, so users should be aware of potential compatibility issues with newer operating systems. The 50th percentile ranking suggests that it is an average performer, but the absence of benchmark scores means that real-world performance must be inferred from the theoretical numbers. The memory bandwidth of 25.60 GB/s is a hard ceiling for texture-heavy workloads, making it a poor fit for modern game design.

How It Compares

The nearestRivals list for the HD 7590M is empty, meaning the database provides no direct rival names, scores, or percentage deltas. This limits the comparison to its overall percentile rank of 50, which places it exactly in the middle of all GPUs. Without rival data, it is impossible to state specific performance deltas against competing parts. The predecessor is the Vancouver generation, and the successor is the Solar System generation, indicating a clear lineage in AMD's mobile product stack. The card's 480 shading units and 16 ROPs are the architectural building blocks, but without rival specifications, no direct comparison can be made. The FP32 output of 576.0 GFLOPS is a useful baseline, but it cannot be contextualized against specific competitors. The 50th percentile rank, however, provides a general sense of position: this card is neither a high-end performer nor a low-end outlier. The empty benchmark array suggests that no standardized tests were run on this part, or that the results were not recorded. Consequently, any comparison must rely on the absolute specifications and the percentile rank, rather than on direct head-to-head data. Users looking for a historical reference point can use the 50th percentile as a guide, but for precise rival comparisons, this dataset offers no support. The 18 W TDP and 25.60 GB/s bandwidth are the defining characteristics, but without rival figures, they remain isolated metrics in the broader hardware landscape.