ATI Radeon X800 PRO AGP

ATI Radeon X800 PRO AGP is an AGP 8x graphics card from AMD, built on the R420 chip and the R400 architecture. It launched on 2004-04-30 and is now listed as End-of-life. The GPU is manufactured on a 130 nm process at TSMC with 160 million transistors, a 281 mm² die, and a transistor density of 569.4K / mm². In the database its percentile versus all GPUs is 50, while its average benchmark score is 0 — meaning the percentile is not backed by a measured performance entry.

Benchmark Performance

Benchmark data for this card is effectively absent. The avgBenchmarkScore field is 0, and the nearestRivals array is empty, so there are no scores, no deltaPct values, and no direct rivals to compare against. The only positional data is the 50th percentile across all GPUs. A 50th percentile suggests a middle-of-the-pack placement, but because the average score is 0, that placement cannot be treated as a validated result.

Instead of measured frame rates, the database provides architectural throughput limits. The X800 PRO AGP is equipped with 12 TMUs and 12 ROPs, giving a texture rate of 5.700 GTexel/s and a pixel rate of 5.700 GPixel/s. These two figures are identical, which indicates a balanced raster pipeline: the card can push texture fetches and pixel writes at the same rate. For a card released on 2004-04-30, that symmetry matters because texture-heavy scenes will not be bottlenecked by one side of the pipeline unless memory or other pipeline work intervenes.

The memory clock is listed as 450 MHz with 900 Mbps effective, but the core clock is not present in the database. Without a core clock and without benchmark scores, the effective compute throughput cannot be derived from the data. The R400 architecture’s feature set, not its clock speed, is the only architectural marker available.

Because the nearestRivals array is empty, there are no percentage deltas to cite. Any claim that this card is faster or slower than a specific competitor would be unsupported by the supplied data. The practical reading of the benchmark section is that the card has no measured standing in this database beyond a mid-list percentile.

Memory Subsystem

Memory capacity is 256 MB of GDDR3, connected across a 256-bit bus. The memory clock is 450 MHz, with a 900 Mbps effective data rate, and this combination produces 28.80 GB/s of bandwidth. The 256-bit bus is the key structural feature: it lets a relatively moderate memory clock move a substantial amount of data per cycle. On paper, 28.80 GB/s is a reasonable figure for an AGP 8x card of this generation, though the database does not list a direct comparison.

For high-resolution workloads, the limiting factors are both capacity and bandwidth. The listed capacity is 256 MB and the listed bandwidth is 28.80 GB/s. At high resolutions, the capacity constraint becomes visible before the bandwidth ceiling is reached, because the memory size is fixed and cannot be expanded. The 256-bit bus and 28.80 GB/s bandwidth help once the data is resident, but they cannot compensate for a small memory size.

The effective memory rate of 900 Mbps is the figure used for data transfer signaling. The 450 MHz base memory clock is the physical clock. Combined with the 256-bit bus, this yields the 28.80 GB/s bandwidth listed in the database. No other memory configurations are listed, so users should treat the 256 MB GDDR3 arrangement as the reference specification.

Ray Tracing and Feature Set

The X800 PRO AGP has no dedicated ray tracing cores and no tensor cores listed. It is a rasterization-oriented chip from the R400 generation, and its API support is DirectX 9.0b (9_2) and OpenGL 2.1. The database lists no Vulkan support. The API list defines the feature level: where those APIs stop, the card’s feature support stops.

The absence of RT and tensor core fields means those units are not part of the specification. In practice, any workload that depends on hardware ray tracing or tensor-accelerated processing will not run on this card. The display output set is 1x DVI, 1x VGA, and 1x S-Video, which is consistent with a card designed for conventional monitors and TV-out rather than modern multi-display environments.

Feature-level expectations should be set by the API list: DirectX 9.0b (9_2) and OpenGL 2.1. Without Vulkan support, the card cannot participate in Vulkan-based titles. The 12 TMUs and 12 ROPs are the listed rasterization resources; there is no separate RT or tensor hardware listed to offload those tasks.

Power and Cooling

Power requirements are modest. The card’s TDP is 48 W, and the suggested PSU rating is 200 W. It draws power through a single 1x Molex connector, and the cooler is a single-slot design. For an AGP 8x system, this means the electrical load is low enough that most AGP-era power supplies can handle it, provided they have a Molex lead available.

Single-slot cooling also means the card does not occupy a neighboring slot, which is a relevant mechanical detail for systems with limited slot spacing. The 48 W TDP is the thermal design power listed in the database; it is not a measured draw. Still, the suggested 200 W PSU gives a clear target for system builders. No additional power connector beyond the 1x Molex is required.

Because there are no clock or load measurements in the database, actual power draw under 3D load is not available. The listed TDP and PSU recommendation are the only power figures. Builders should use the 48 W TDP and 200 W suggested PSU as planning numbers.

How It Compares

The nearestRivals list for the ATI Radeon X800 PRO AGP is empty. As a result, there are no rival names, scores, or deltaPct values to present. The card’s only positional reference is the 50th percentile across all GPUs in the database, and that percentile is not accompanied by a measured average score (avgBenchmarkScore is 0).

Within the product family, the X800 PRO AGP sits after the Radeon R300 and before the Radeon R500 AGP. The database lists these as predecessor and successor, but does not provide comparative performance data between them. Without nearestRivals, no direct competitor positioning is possible from the available data.

Because the percentile is 50, a neutral reader might expect this card to be in the middle of the database’s GPU population. However, with a zero average benchmark score, the percentile is not anchored to any actual benchmark run. The comparison section contains no rival entries: this card has no listed rivals in the data.

Who Should Consider It

This is a card for systems that require AGP 8x and a DirectX 9.0b feature level. The 256 MB memory capacity and 28.80 GB/s bandwidth define its practical range. Users who intend to run older games at moderate resolutions and texture settings are the natural audience. The card has 12 TMUs and 12 ROPs, so it has no listed compute-oriented hardware; it is built for conventional pixel rendering.

High-resolution gaming is outside the design intention. The 256 MB capacity is fixed, and there is no Vulkan support in the database. Games that need a more current API or a larger memory capacity will not fit this card. For an AGP-era build, however, the X800 PRO AGP offers a single-slot, 48 W alternative that only requires a 200 W PSU and 1x Molex connector.

Because the database contains no measured benchmark scores, the recommendation cannot be framed as a specific frame-rate result. The recommendation is based on the architectural limits: 256 MB of GDDR3, 28.80 GB/s of bandwidth, a 256-bit bus, and a DirectX 9.0b (9_2) feature set. Those are enough to say the card is best suited to earlier, lighter rendering workloads rather than demanding modern titles.

Given the 50th percentile, a mid-list position among all GPUs is plausible, but readers should not infer a specific competitive margin from that single number. The card is end-of-life, which means it is a legacy part. Buyers with AGP motherboards and low power budgets are the target. Anyone expecting ray tracing or tensor workloads should look elsewhere, because this card has neither RT cores nor tensor cores listed.