ATI Radeon 9700 PRO X4

The ATI Radeon 9700 PRO X4 is an end-of-life graphics card from AMD, built on the R300 architecture and fabricated by TSMC on a 150 nm process. The die contains 110 million transistors across 215 mm², yielding a transistor density of 511.6K per mm². The database places the card at the 50th percentile among all GPUs, though no individual benchmark scores are recorded for this SKU. Released on 2002-08-18, the card succeeded the Radeon R200 and was followed by the Radeon R400 AGP.

Memory Subsystem

The memory subsystem is defined by a 128 MB DDR frame buffer on a 256-bit bus, with a memory clock of 310 MHz and an effective data rate of 620 Mbps. The resulting bandwidth is 19.84 GB/s. The 256-bit bus is a wide interface, but the 128 MB capacity is the primary constraint for high-resolution workloads. At a pixel rate of 2.600 GPixel/s, the card can fill frames quickly, but the frame buffer must hold color, depth, and texture data for the entire scene. Textures that exceed 128 MB will spill to system memory, and the PCI bus interface cannot service those spills at the speed of the local frame buffer. The bandwidth of 19.84 GB/s, when divided by the pixel rate, provides roughly 7.6 bytes per pixel of memory throughput per pixel — enough for typical color and depth data, but high-resolution depth buffers and multi-sampled render targets will consume the buffer rapidly. The 8 ROPs and 8 TMUs are balanced against the memory throughput; the data shows a design where the 256-bit bus width compensates for the 310 MHz memory clock. For high resolutions, the 128 MB capacity is the hard ceiling, not the bandwidth. The 256-bit bus width is the most striking aspect of the memory design; at 310 MHz, the effective 620 Mbps data rate produces 19.84 GB/s, which is a balanced figure for the 2.600 GPixel/s pixel rate.

Who Should Consider It

With no benchmark scores recorded, the case for this card rests on its memory configuration and fill-rate metrics. The 128 MB frame buffer and 19.84 GB/s bandwidth target moderate resolutions and settings that fit within the buffer. Users running DirectX 9 titles with modest texture sizes will find the card adequate; those who push high resolutions with large texture packs will hit the memory ceiling quickly. The 8 TMUs deliver a texture rate of 2.600 GTexel/s, and the 8 ROPs deliver 2.600 GPixel/s — a symmetric design that indicates the card is fill-rate balanced. The 50th percentile standing in the database places it at the median of all GPUs, which is a neutral position: not a high-end part, not a low-end part. The 50th percentile is a median position, meaning half of the GPUs in the database rank above and half below, though the lack of recorded scores for this card makes that percentile an abstract reference point. The dual-slot cooler and 2x Molex power requirement indicate a card that drew enough power to need a substantial cooler, but the suggested PSU of 200 W is a modest system-level recommendation. The PCI bus interface is a defining limitation; the card cannot take advantage of faster host interconnects, which will constrain performance in texture streaming scenarios.

Benchmark Performance

The database contains no benchmark scores for the ATI Radeon 9700 PRO X4. The benchmarks array is empty, and the average benchmark score is 0. The percentileVsAllGpus field is 50, which indicates a median position in the overall distribution of GPUs in the database, but this percentile is not tied to any recorded test. Without nearestRivals data, there are no deltaPct values to report, and no direct percentage comparisons against other cards are possible. The only performance-related metrics in the specifications are the pixel rate of 2.600 GPixel/s and the texture rate of 2.600 GTexel/s. These rates are identical, which suggests that the card's pixel and texture pipelines are clocked and configured symmetrically — a design choice that simplifies driver scheduling and avoids bottlenecks between the two stages. The 8 TMUs and 8 ROPs each operate at the same rate, producing the matched output. In the absence of recorded scores, these fill rates are the only quantitative performance indicators, and they describe a card that can process 2.6 billion pixels per second and 2.6 billion texels per second. The memory bandwidth of 19.84 GB/s is the third leg of the performance triangle; with the 256-bit bus, the card can sustain the fill rates without starving the pixel pipeline in typical workloads. The absence of recorded scores is itself a data point; it means the card has not been exercised by the benchmark suite, so the fill rates and memory bandwidth are the only quantitative anchors.

How It Compares

The nearestRivals field is empty, so this card cannot be positioned against specific competitors with percentage deltas. The broader context comes from its predecessor and successor: the Radeon R200 and the Radeon R400 AGP. The Radeon R200 is the previous generation, and the Radeon R400 AGP is the next, which indicates that the 9700 PRO X4 sits in the R300 generation. The predecessor and successor designations bracket the card within the R300 generation, but the absence of rival data prevents any assessment of how it fared against contemporary parts. The 50th percentile standing places it at the median of the entire database, but without rival scores, that standing cannot be decomposed. The memory configuration — 128 MB, DDR, 256-bit, 19.84 GB/s — is the most distinctive feature; the 256-bit bus is a wide interface, and the 19.84 GB/s bandwidth is the product of the 310 MHz clock and the effective 620 Mbps data rate. The dual-slot cooler and 2x Molex power connectors are physical indicators of a card designed for a desktop configuration with substantial cooling. The 4x DVI display outputs are notable: four DVI ports suggest a multi-display orientation rather than a single-output consumer card. The PCI bus interface is the host connection, and the card is end-of-life.

FAQ

Q: What memory type and bus width does the Radeon 9700 PRO X4 use?

A: It uses 128 MB of DDR memory on a 256-bit bus, with a memory clock of 310 MHz and an effective data rate of 620 Mbps.

Q: What is the memory bandwidth?

A: The bandwidth is 19.84 GB/s.

Q: What DirectX and OpenGL versions are supported?

A: The card supports DirectX 9.0 (9_0) and OpenGL 2.0. Vulkan support is not listed.

Q: What power connectors does the card require?

A: It requires two Molex connectors, and the suggested PSU is 200 W.

Q: How many display outputs does it have?

A: It has four DVI outputs.

Q: What is the production status and release date?

A: The card is end-of-life and was released on 2002-08-18.

Power and Cooling

The card is dual-slot, which indicates a substantial cooler. Power is delivered through two Molex connectors, and the suggested PSU is 200 W. A TDP figure is not provided in the specifications. The 150 nm process with 110 million transistors on a 215 mm² die gives a transistor density of 511.6K per mm² — a figure that reflects the manufacturing technology of the time. The dual-slot design and two power connectors suggest that the card's power draw was significant enough to require a dedicated cooler and an auxiliary power connection, but without a TDP number, the exact draw is unquantified. The 200 W suggested PSU is a system-level recommendation, not a card-level draw; it accounts for the rest of the system's components. The dual-slot cooler is a physical footprint that occupies two expansion slots, which is a consideration for chassis with limited slot spacing.

Ray Tracing and Feature Set

The specifications list no ray tracing cores and no tensor cores. The feature set is anchored to the DirectX 9.0 (9_0) API, with OpenGL 2.0 support. Vulkan is not listed, which is consistent with the card's 2002 release date. The absence of ray tracing and tensor cores means the card relies entirely on the traditional rasterization pipeline. The 8 TMUs and 8 ROPs handle texturing and pixel output, and the 2.600 GTexel/s and 2.600 GPixel/s rates define the throughput. The DirectX 9.0 (9_0) support indicates the card targets the ninth generation of Microsoft's graphics API, and OpenGL 2.0 provides an alternative API path. The 4x DVI outputs are the display feature set, allowing up to four digital displays. The PCI bus interface is the host connection, and the card is end-of-life.