NVIDIA GeForce4 MX 460

The NVIDIA GeForce4 MX 460 is a legacy AGP 4x graphics card built on the Celsius architecture with the NV17 chip. Fabricated by TSMC on a 150 nm process, it packs 29 million transistors onto a 65 mm² die, yielding a transistor density of 446.2K per square millimeter. The card offers 4 texture mapping units and 2 ROPs, with a pixel rate of 600.0 MPixel/s and a texture rate of 1.200 GTexel/s. It carries 64 MB of DDR memory on a 128-bit bus, providing 7.200 GB/s of bandwidth. It supports DirectX 7.0 and OpenGL 1.5, and its production status is end-of-life, with a release date of February 5, 2002.

Benchmark Performance

The FACT PACK lists an average benchmark score of 0 and an empty benchmarks array, meaning no standardized performance scores are recorded for this card in the database. This absence of data indicates that the card was never subjected to the database's standard test suite. Instead, the available performance metrics are the pixel fillrate and texture fillrate. The pixel rate of 600.0 MPixel/s and texture rate of 1.200 GTexel/s define the card's raw fillrate capacity. With only 2 ROPs, the pixel throughput is inherently limited, as each ROP can only process a certain number of pixels per second. The 4 TMUs allow for a texture rate that is double the pixel rate, indicating a design balanced for texture-heavy workloads. Because no rival scores are present, direct percentage deltas cannot be computed. The card's performance profile is therefore best understood through its fillrate figures: 600.0 MPixel/s for pixel processing and 1.200 GTexel/s for texture sampling. These numbers suggest a card that can handle early 3D workloads at moderate settings, but the lack of benchmark data means no quantitative comparison to other GPUs is possible from this pack. The 150 nm process node and 29 million transistor count suggest a relatively simple design, which aligns with the modest fillrate figures. The pixel rate of 600.0 MPixel/s is a hard ceiling for how many pixels can be written per second, and the texture rate of 1.200 GTexel/s indicates the rate at which textures can be sampled. The 2 ROPs are the primary bottleneck for pixel output, while the 4 TMUs provide a balanced texture fetch capability.

How It Compares

The nearestRivals field in the data pack is empty, so no rival names, scores, or deltaPct values are available for comparison. This is a significant limitation, as it prevents any direct percentage-based positioning against competing GPUs. The FACT PACK identifies the GeForce 3 as the predecessor and the GeForce 4 Ti as the successor. The GeForce4 MX 460 sits in the middle of this lineage, inheriting the Celsius architecture. Without rival data, the comparison must be qualitative: the MX 460 is positioned below the GeForce 4 Ti in the product stack, as the Ti series typically commands higher performance. The predecessor GeForce 3 uses a different architecture, but the pack does not provide its specifications. Therefore, the only concrete comparison points are the card's own metrics. The empty nearestRivals list means that no direct competitor data exists in this pack, so any positioning must rely on the card's absolute fillrate and memory figures. In the absence of rival scores, the card's performance is best judged by its 600.0 MPixel/s pixel rate and 1.200 GTexel/s texture rate, which are modest but functional for its era. The card's single-slot design and lack of power connectors further suggest it was intended for mainstream, low-power systems, differentiating it from higher-end cards that require external power. The 64 MB memory capacity is another distinguishing factor, as higher-end cards of the time often featured larger frame buffers.

Memory Subsystem

The GeForce4 MX 460 comes with 64 MB of DDR memory. The memory bus is 128 bits wide, and the memory clock is 225 MHz, with an effective data rate of 450 Mbps. This yields a memory bandwidth of 7.200 GB/s. For high-resolution gaming, this bandwidth is a limiting factor. A 128-bit bus is relatively narrow, and the 7.200 GB/s figure is modest, but for its era it provided sufficient throughput for moderate resolutions. The 64 MB frame buffer constrains texture detail and anti-aliasing, as larger textures will exceed the available memory. The effective 450 Mbps data rate indicates DDR's double-pumped nature, effectively doubling the 225 MHz clock. This means that for each clock cycle, the memory can transfer data on both the rising and falling edges, doubling the theoretical throughput. With only 2 ROPs, the pixel fillrate of 600.0 MPixel/s will likely bottleneck before memory bandwidth becomes the primary constraint at lower resolutions. At higher resolutions, the 7.200 GB/s bandwidth will become the ceiling, limiting texture streaming and frame buffer writes. The 128-bit bus width is a key factor: a wider bus would allow more data per clock, but at 225 MHz, the bandwidth is fixed at 7.200 GB/s. This means that for high-resolution textures, the card will struggle, as the memory subsystem cannot feed the 4 TMUs fast enough to sustain 1.200 GTexel/s under heavy load. The 64 MB capacity is also a constraint, as it limits the number of textures that can be resident in memory simultaneously, directly impacting detail levels in games.

Who Should Consider It

Given the fillrate and memory figures, the GeForce4 MX 460 is suited for legacy systems running DirectX 7.0 titles. The 600.0 MPixel/s pixel rate and 1.200 GTexel/s texture rate indicate it can handle early 3D games at moderate resolutions and detail settings. The 64 MB frame buffer is adequate for lower resolutions, but not for high-resolution textures or heavy anti-aliasing. The card's AGP 4x interface is a legacy bus, so it is intended for older motherboards. Users with such systems who want to play DirectX 7.0-era games will find this card capable, but it is not suited for modern workloads. The lack of benchmark scores means precise resolution/settings recommendations cannot be derived from data, but the pixel rate of 600.0 MPixel/s suggests a practical ceiling at moderate resolutions with reduced detail. The 2 ROPs limit pixel fill, so higher resolutions will see significant frame rate drops. For users with a 200 W power supply and an AGP 4x slot, this card is a drop-in solution for retro gaming. The 64 MB memory is sufficient for the era's games, but not for texture-heavy mods or higher resolutions. The card's single-slot design and lack of power connectors make it easy to install in compact cases. However, its DirectX 7.0 support means it cannot run games that require a higher DirectX version, limiting its software compatibility to a narrow window of titles.

Ray Tracing and Feature Set

The FACT PACK lists no ray tracing cores (rtCores is null) and no tensor cores (tensorCores is null). This means the card has no dedicated hardware for ray tracing or AI acceleration. The API support is limited to DirectX 7.0 and OpenGL 1.5; Vulkan is not supported (null). The absence of RT and tensor cores means no hardware-accelerated ray tracing or DLSS-like features. The feature set is therefore purely fixed-function, relying on the 4 TMUs and 2 ROPs for rendering. The display outputs include 1x DVI, 1x VGA, and 1x S-Video, allowing connectivity to a range of monitors and TVs. The lack of Vulkan support means it cannot run modern Vulkan-based titles, and DirectX 7.0 limits it to very old games. The card's Celsius architecture is fixed-function, meaning it has no programmable shader units. This limits it to fixed-function pipeline operations, which were standard for DirectX 7.0-era games. The 4 TMUs handle all texture sampling, and the 2 ROPs manage pixel output, with no additional processing stages. For users interested in modern rendering techniques, this card is entirely unsuitable, as it lacks the hardware and API support for any advanced features.

Power and Cooling

The FACT PACK does not specify a TDP (thermal design power) for this card. However, the suggested PSU is 200 W, indicating a low-power design. The card uses no power connectors (None), meaning it draws all power from the AGP 4x slot. The slot width is Single-slot, so it occupies one expansion slot. With a 200 W PSU recommendation, this card is suitable for older systems with modest power supplies. The absence of external power connectors simplifies installation, as no additional cables are required. The 150 nm process node and 29 million transistors on a 65 mm² die suggest a relatively low thermal output, but the exact TDP is not recorded. The single-slot design allows for standard cooling solutions, and the card's end-of-life production status means it is only available on the used market. The lack of power connectors also means that the card's power draw is entirely dependent on the AGP slot's capacity, which is typically limited. The 200 W PSU recommendation is a guideline for the entire system, not just the card, indicating that the card itself is not power-hungry. The single-slot cooler is adequate for the card's thermal output, given the low transistor count and 150 nm process. Users with a 200 W power supply can confidently install this card without worrying about power delivery, as it draws everything from the slot.