ATI Mobility Radeon HD 5870

The ATI Mobility Radeon HD 5870 is a mobile graphics processor from AMD, built on the TeraScale 2 architecture and manufactured by TSMC on a 40 nm process. The chip, codenamed Broadway, integrates 1,040 million transistors into a 166 mm² die, achieving a transistor density of 6.3M per square millimeter. Released on 2010-01-06, this part is now end-of-life, and the database places it at the 50th percentile of all GPUs tracked. The data shows a mid-pack performer for its generation, defined by a specific balance of compute throughput, fillrate, and memory bandwidth.

Benchmark Performance

The compute capabilities of the ATI Mobility Radeon HD 5870 are anchored by its 800 shading units, 40 texture mapping units, and 16 raster output units. The data shows a floating-point throughput of 1,120.0 GFLOPS (FP32), which represents the raw shader processing power available for vertex and pixel workloads. The pixel rate is 11.20 GPixel/s, and the texture rate is 28.00 GTexel/s, indicating a GPU designed for balanced throughput rather than extreme specialization in any single direction. With an average benchmark score of 0 in the database, no standardized run has been recorded for this specific part, so the percentile ranking is derived from its aggregate characteristics. The 50th percentile standing places it exactly at the median of the entire GPU landscape, meaning it outperforms half of all GPUs tracked and is outperformed by the other half. The 800 shading units deliver a theoretical compute peak that would have been competitive for mobile gaming in its release period, but the 40 TMUs and 16 ROPs are a relatively modest combination. This suggests that texture-heavy scenes and high pixel fillrate requirements could become bottlenecks, while shader complexity is handled with relative ease. The 28.00 GTexel/s texture rate defines how quickly textures can be applied, and the 11.20 GPixel/s pixel rate governs how fast fragments are written to the frame buffer. These figures collectively indicate a GPU that excels in compute-heavy shader workloads but may struggle under extreme geometry or high-resolution pixel processing. The 50th percentile is a critical data point, confirming a mainstream tier placement without needing rival names.

Memory Subsystem

The memory subsystem of the ATI Mobility Radeon HD 5870 consists of 1024 MB of GDDR5 memory connected via a 128-bit bus. The memory clock runs at 1000 MHz, yielding an effective data rate of 4 Gbps per pin. The total bandwidth available is 64.00 GB/s, a relatively narrow figure for a GPU of this era. The 128-bit bus width is the primary constraint, as it halves the potential throughput compared to wider implementations. The data shows that this bandwidth is sufficient for common gaming scenarios at standard resolutions, but it becomes a limiting factor at higher resolutions or when extreme texture filtering and anti-aliasing are enabled. The 1024 MB frame buffer is a standard capacity for the time, but the narrow bus means the memory controller must be efficient to keep the 800 shading units fed with data. The 4 Gbps effective memory speed is a high data rate for GDDR5, yet the 64.00 GB/s total bandwidth is the definitive number for data transfer, directly impacting how fast textures and geometry are streamed from memory. In scenarios where the frame buffer is exhausted, performance will degrade, but within its 1 GB capacity, the bandwidth is a measured characteristic of this mobile part. The data indicates that the memory subsystem is balanced for the GPU's compute capabilities, but the bandwidth is the primary constraint when pushing resolution or texture quality, making high-resolution gaming a challenge for sustained frame rates.

Power and Cooling

The ATI Mobility Radeon HD 5870 is rated with a thermal design power (TDP) of 50 W. This is a relatively low power draw for a GPU with 800 shading units and a 1,120.0 GFLOPS compute rating, making it suitable for a mobile MXM module form factor. The slot width is specified as "MXM Module," and the bus interface is MXM-B (3.0), indicating that the card is designed for laptops and small-form-factor systems that use the MXM standard. The data does not specify any power connectors, suggesting that power is delivered entirely through the MXM slot itself. The 50 W TDP means that a capable air cooler is sufficient to manage thermals, as the power envelope is modest. There is no suggested PSU listed, which is typical for a mobile component, as the system's power delivery is integrated into the laptop design. The 40 nm process from TSMC contributes to this efficiency, as the 1,040 million transistors are packed into a 166 mm² die, allowing for a relatively high density of 6.3M transistors per square millimeter while maintaining a low power draw. The data shows that the 50 W TDP is a defining characteristic, enabling the MXM-B (3.0) interface to provide all necessary power without auxiliary connectors. This makes the card a viable option for thin-and-light gaming laptops of its era, where thermal and power constraints are paramount. The low power draw also implies reduced heat output, which is critical for the compact cooling solutions found in mobile chassis.

How It Compares

The data places the ATI Mobility Radeon HD 5870 at the 50th percentile of all GPUs tracked in the database, indicating a median performance level. Without specific rival scores or delta percentages available in the nearestRivals field, the comparison must rely on its position relative to the entire GPU population. The 50th percentile means that it outperforms exactly half of all GPUs and is outperformed by the other half, establishing it as a mainstream or mid-range part in the global context. Its predecessor, the M9x series, and its successor, the Vancouver series, bracket this product in the generational timeline. The transition from M9x to Vancouver represents an architectural progression, with the HD 5870 sitting between them as a distinct step. The data shows that the HD 5870 is an end-of-life product, meaning it has been superseded in the market by newer generations. In the broader landscape, its 800 shading units and 64.00 GB/s memory bandwidth define its capabilities, and the 50th percentile ranking confirms that it is neither a high-end nor a low-end part. The lack of a benchmark score (0) in the database suggests that no standardized benchmark runs have been recorded for this specific part, so the percentile ranking is derived from the aggregate characteristics of the GPU. Compared to the overall database, this GPU's compute density (1,120.0 GFLOPS) and fillrate (11.20 GPixel/s) place it in a specific performance tier that is exactly average. The 50th percentile is a strong indicator that it was a capable mobile solution, but not a flagship, and it serves as a baseline for understanding its standing in the absence of direct rival comparisons.

Ray Tracing and Feature Set

The ATI Mobility Radeon HD 5870 does not include dedicated ray tracing cores or tensor cores, as these fields are null in the data. This means the GPU relies entirely on its 800 shading units for all rendering tasks, including any ray-traced effects, which would be processed in a non-accelerated manner. The feature set is defined by its TeraScale 2 architecture, which supports DirectX 11.2 (11_0) and OpenGL 4.4. The DirectX 11.2 support indicates compatibility with modern gaming APIs of its time, including tessellation and compute shaders. OpenGL 4.4 support provides a broad compatibility layer for cross-platform applications. Vulkan support is not listed (null), which means the GPU lacks native Vulkan drivers, limiting its use in Vulkan-based titles. The absence of tensor cores means no AI-accelerated features, and the absence of RT cores means no hardware-accelerated ray tracing. The data shows that the feature set is purely rasterization-focused, with the API support being the primary interface for developers. The DirectX 11.2 (11_0) API is a specific version that enables feature level 11_0, which is the baseline for DirectX 11. The OpenGL 4.4 support is a mature version of that API. This combination makes the GPU suitable for games and applications that rely on these APIs, but it will not support newer features like ray tracing or variable rate shading that require dedicated hardware. The 40 TMUs and 16 ROPs are the fixed-function units that handle texture filtering and pixel output, respectively, and they are the backbone of the rasterization pipeline. The data confirms that this is a conventional GPU design from the TeraScale 2 era, with no dedicated acceleration for modern graphics features, making it a product of its time with clear limitations for contemporary workloads.