ATI Radeon HD 2400 PRO PCI

The ATI Radeon HD 2400 PRO PCI is an end-of-life, low-power graphics card from AMD, built on the TeraScale architecture with the RV610 chip. It was manufactured by TSMC on a 65 nm process, with 180 million transistors packed into an 85 mm² die, giving a transistor density of 2.1M / mm². The card belongs to the Radeon R600 (HD 2400) generation, was released on 2007-06-27, and is listed in the database at the 50th percentile against all GPUs, though its average benchmark score is recorded as 0 and its benchmark array is empty.

Power and Cooling — TDP, PSU recommendation, connector requirements

The data lists a TDP of 20 W, which is a very low power envelope. Accordingly, the suggested PSU is only 200 W, and no power connectors are listed. The card is single-slot, and no length, height, or width dimensions are provided in the fact pack. With no external power connectors, the card’s power delivery description in the data is limited to the PCI bus interface and the 20 W TDP figure.

The 65 nm process node, 180 million transistor count, 85 mm² die size, and 2.1M / mm² transistor density all sit behind that 20 W figure. The memory clock is 400 MHz, or 800 Mbps effective, but the fact pack does not provide a separate memory power draw. The bus interface is PCI, not PCIe, which is more relevant to compatibility than to cooling. The low TDP also means a simple cooling solution is plausible, although the fact pack gives no explicit cooler specification beyond the single-slot form factor.

Ray Tracing and Feature Set — RT/tensor cores, API support from facts

The fact pack contains no RT core count and no tensor core count. That means ray tracing acceleration and tensor-oriented operations are not represented anywhere in the available specifications. The API entries are DirectX 10.0 (10_0), OpenGL 3.3, and a null Vulkan field. With Vulkan absent from the data, the card’s supported API paths are the older DirectX and OpenGL interfaces.

The feature set is otherwise defined by the TeraScale architecture and the RV610 chip. The card has 40 shading units, 4 texture mapping units, and 4 raster output units. These units are the core rendering blocks listed in the fact pack. The generation label is Radeon R600 (HD 2400), and the display output is listed as 1x DVI, which is a minimal output configuration. The absence of Vulkan support is a notable limitation, and the empty RT/tensor core fields make it clear that the card is not positioned as a hardware-accelerated ray tracing or tensor product.

Who Should Consider It

Because the benchmarks array is empty and the average benchmark score is 0, the data cannot support a conventional resolution-and-settings recommendation based on measured frame rates. What the data does support is a profile: a PCI-bus card with 512 MB of DDR2 memory, a 64-bit bus, 6.400 GB/s of bandwidth, a pixel rate of 2.100 GPixel/s, a texture rate of 2.100 GTexel/s, and 42.00 GFLOPS of FP32 compute. That profile is not suited to high-resolution, high-texture workloads, because the memory capacity and bandwidth are small and the pixel/texture rates are modest.

The card is best understood as a candidate for systems that require a low-power PCI display solution. The 20 W TDP, 200 W suggested PSU, and lack of power connectors make installation straightforward in the systems that can accept a PCI card. The single DVI output reinforces that this is a basic display card. The 50th-percentile field could look like a mid-pack position, but with no benchmark scores behind it, the percentile is not a meaningful performance endorsement. Any user considering the card for demanding 3D work would be going beyond what the data actually shows.

FAQ

Q: What DirectX version does the ATI Radeon HD 2400 PRO PCI support?

A: The fact pack lists DirectX 10.0 (10_0).

Q: Does the card support Vulkan?

A: No Vulkan support is listed; the Vulkan field is null.

Q: How much memory does it have, and on what bus width?

A: It has 512 MB of DDR2 memory on a 64-bit bus, with 6.400 GB/s of bandwidth.

Q: What PSU is recommended for this card?

A: The suggested PSU is 200 W, and the card’s TDP is 20 W with no power connectors listed.

Q: Does it have ray tracing or tensor cores?

A: The data lists no RT cores and no tensor cores.

Q: Is the card still in production?

A: No; its production status is end-of-life, and its release date is 2007-06-27.

Benchmark Performance

The benchmark section of the fact pack is empty. The average benchmark score is 0, and the nearestRivals array is also empty. As a result, there are no exact delta percentages to analyze, and no comparison of the form “X% faster than Y” can be produced from the data. The only quantitative performance-related fields available are raw throughput limits: 2.100 GPixel/s pixel rate, 2.100 GTexel/s texture rate, and 42.00 GFLOPS FP32 compute.

These figures are ceilings, not measured application results. The percentileVsAllGpus field is 50, but it coexists with an average benchmark score of 0 and an empty benchmarks array, so the percentile has no supporting benchmark average in the record. Without nearest rivals, there is no relative position to quantify and no percentage-based ranking to interpret. The empty benchmark data is itself the most important finding: the database does not provide any workload-based evidence for this card’s performance.

Memory Subsystem

The memory subsystem consists of 512 MB of DDR2 memory on a 64-bit bus. The memory clock is 400 MHz, which the fact pack expresses as 800 Mbps effective. Total bandwidth is 6.400 GB/s. For high resolutions, this is a restrictive configuration: 512 MB limits the amount of texture and geometry data that can be held locally, while 6.400 GB/s bounds how quickly data can move into the rendering units.

The 64-bit bus width is a structural reason for that bandwidth figure. The 2.100 GPixel/s pixel rate and 2.100 GTexel/s texture rate are modest, and both depend on the same memory channel. The interaction between memory size, bus width, and bandwidth is the clearest constraint in the fact pack. The card’s PCI bus interface is separate from the memory bus, but the memory subsystem itself would still be a limiting factor for higher-resolution work. The data shows no larger memory variant and no wider bus option within this entry.

How It Compares

The nearestRivals list is empty, so there are no rival entries to profile. The fact pack does name a predecessor, Radeon R500 PCIe, and a successor, Radeon R700, but it provides no scores, specifications, or deltas for either one. Therefore no percentage-based comparison can be drawn against those generations or against any other GPU.

The only comparative-looking value in the record is percentileVsAllGpus at 50, but because it appears alongside an average benchmark score of 0 and an empty benchmarks array, it cannot anchor a comparison. In the absence of nearest rivals, the card’s position is undefined in the database’s own comparison framework. There is no measured rival data to cite, no deltaPct field to analyze, and no relative ranking to evaluate. The database simply does not place this card against any measured competitor.