NVIDIA GeForce2 MX DH Pro TV PCI

The NVIDIA GeForce2 MX DH Pro TV PCI is an end-of-life PCI graphics card built on the 180 nm TSMC process, featuring the NV11 chip from the Celsius architecture. It carries 20 million transistors on a 64 mm² die, resulting in a transistor density of 312.5K per mm². This card targets legacy systems, offering 32 MB of SDR memory on a 128-bit bus, with a memory clock of 166 MHz and a bandwidth of 2.656 GB/s. Its pixel rate is 350.0 MPixel/s, and the texture rate reaches 700.0 MTexel/s, while its overall percentile ranking against all GPUs stands at 50.

Benchmark Performance

The GeForce2 MX DH Pro TV PCI holds a neutral position in the hardware landscape, sitting exactly at the 50th percentile among all GPUs tracked in this database. This percentile ranking indicates that the card performs at the median level of every graphics processor ever benchmarked — a remarkable feat for a product from this era, but also a clear sign that its capabilities are firmly rooted in its release period. The average benchmark score is 0, which means there are no direct synthetic results to quote; however, the architectural metrics provide a reliable proxy for expected performance.

The card’s fill-rate figures tell the story. With a pixel rate of 350.0 MPixel/s, the GeForce2 MX DH Pro TV PCI can push roughly one-third of a gigapixel per second. Meanwhile, the texture rate of 700.0 MTexel/s doubles that number, indicating that the 4 texture mapping units (TMUs) are working in parallel to feed the 2 raster output units (ROPs). This 2:1 ratio between texture and pixel throughput is typical for this generation, allowing the card to handle texture-heavy scenes more efficiently than pure 2D or simple 3D workloads. The 2.656 GB/s of memory bandwidth, derived from the 128-bit bus and 166 MHz SDR memory, is the limiting factor in most scenarios — a common bottleneck for mid-range cards of this vintage.

Because the nearest rivals list is empty, there are no direct percentage deltas to cite against competing products. Instead, the 50th percentile placement serves as the benchmark context. This means half of all GPUs in the database score higher, and half score lower. For a PCI-based card with no auxiliary power connectors, achieving the median is a testament to the efficiency of the NV11 chip. The 180 nm process node, while large by modern standards, was competitive at launch, and the 20 million transistor count is modest but sufficient for the DirectX 7.0 feature set the card supports.

Ray Tracing and Feature Set

This card does not include dedicated ray tracing cores or tensor cores, as those technologies did not exist in the Celsius architecture. The GeForce2 MX DH Pro TV PCI relies entirely on traditional rasterization through its fixed-function pipeline. The absence of RT and tensor hardware means no hardware-accelerated ray tracing, no DLSS, and no AI-based upscaling — features that are ubiquitous in modern GPUs but were irrelevant in the year 2000. The card’s API support is limited to DirectX 7.0 and OpenGL 1.2, with no Vulkan support listed. This places it in the era of transform and lighting (T&L) being handled by the CPU or, at best, a rudimentary vertex pipeline.

The feature set is defined by what it lacks in modern terms but also by what it offers for its time. The 32 MB of SDR memory is adequate for early 3D games, and the 128-bit memory bus ensures that the 2.656 GB/s bandwidth is used efficiently. Display outputs include 2x VGA and 1x S-Video, making it suitable for multi-monitor setups or connecting to a television — hence the "TV" in its name. The PCI bus interface is a critical constraint; it offers far lower bandwidth than the AGP slots that were becoming standard, but it ensures compatibility with older motherboards.

The DirectX 7.0 support means the card can handle hardware T&L, a feature that was a differentiator for the GeForce 256 predecessor. However, the GeForce2 MX DH Pro TV PCI is a cut-down version of the full GeForce 2, with fewer TMUs and ROPs than the flagship. The 4 TMUs and 2 ROPs are half of what the full GeForce 2 offered, which explains its mid-range positioning. The lack of Vulkan support is irrelevant for its target software library, as OpenGL 1.2 was the standard for professional and early gaming applications.

How It Compares

Without any nearest rivals listed, a direct comparison must rely on the card’s own metrics and its position in the broader GPU hierarchy. The 50th percentile is a useful anchor: it means the GeForce2 MX DH Pro TV PCI is neither a high-end part nor a low-end one. It sits precisely in the middle of every GPU ever released, which is a surprising outcome for a product from 2000. This is likely because the database includes many older and low-power cards that this NVIDIA product outperforms, while modern GPUs vastly exceed it, pulling the average down.

The predecessor, GeForce 256, and the successor, GeForce 2, bracket this card in the product line. The GeForce2 MX DH Pro TV PCI is a variant of the GeForce 2 series, specifically designed for the PCI bus and budget-oriented systems. Compared to the full GeForce 2, it sacrifices texture and pixel throughput — the 700.0 MTexel/s texture rate is roughly half of what the full chip could achieve. The memory bandwidth of 2.656 GB/s is also lower, as the MX used SDR memory instead of faster DDR or SGRAM.

In the context of its era, the card’s 32 MB memory capacity was standard for mid-range products, and the 128-bit bus was a selling point. The 350.0 MPixel/s pixel rate allowed for playable frame rates at lower resolutions (though no specific resolutions are provided in the data). The 50th percentile ranking suggests it was competitive with contemporary mid-range cards from other manufacturers, though no specific rival names or scores are available. The lack of a launch MSRP in the data means no pricing analysis is possible, but the product’s positioning as a PCI-based, single-slot card with no power connectors indicates it was aimed at upgrades for office or home PCs.

Who Should Consider It

The benchmark results, while limited, point to a clear profile for potential users. The 50th percentile ranking means the GeForce2 MX DH Pro TV PCI is suitable for those who need basic 3D acceleration for legacy software, not for modern gaming. The card’s 32 MB memory and 2.656 GB/s bandwidth are sufficient for early 3D titles from the late 1990s and early 2000s, but they will struggle with anything more demanding. The pixel rate of 350.0 MPixel/s and texture rate of 700.0 MTexel/s suggest that the card can handle resolutions and detail settings that were common at the time of its release — likely 800x600 or 1024x768 with reduced texture quality, though exact numbers are not provided.

The PCI bus interface is both a limitation and an opportunity. It means the card cannot be used in systems without a PCI slot, but it also makes it an option for older motherboards that lack AGP. The 2x VGA and 1x S-Video outputs make it ideal for multi-monitor productivity setups or connecting to a television for media playback. The single-slot design and lack of power connectors mean it can be installed in almost any chassis, provided the power supply meets the 200 W suggested rating. This is a very low requirement, making it compatible with the vast majority of power supplies from its era.

For users running Windows 98 or early Windows XP systems with DirectX 7.0 games, this card is a viable choice. The OpenGL 1.2 support also covers many early OpenGL titles and professional applications from that period. However, the absence of tensor and RT cores means it is entirely unsuitable for any modern workload — no machine learning, no ray-traced rendering, no hardware-accelerated video encoding. The card’s 50th percentile ranking is a double-edged sword: it shows the card was competent for its time, but it also underscores how far GPUs have advanced. Anyone building a retro gaming rig or a period-correct Windows 98 system should consider this card, but modern users will find it obsolete.

Power and Cooling

The GeForce2 MX DH Pro TV PCI has a suggested PSU rating of 200 W, which is extremely modest by any standard. This figure indicates that the card draws very little power, though the exact TDP is not provided in the data. The lack of any power connectors on the card reinforces this: it draws all its power from the PCI slot itself, which is limited to 25 W in the PCI specification. The 200 W system power supply recommendation is a safe margin that accounts for the rest of the system components, such as the CPU and motherboard, not just the GPU.

Cooling is handled by a passive or low-profile active solution, given the single-slot design. The 180 nm process node produces less heat than smaller nodes of later generations, but the 20 million transistors still generate some thermal output. The card’s 350.0 MPixel/s and 700.0 MTexel/s rates are low enough that the GPU does not require aggressive cooling. A simple heatsink with a small fan, or even a passive heatsink, is likely sufficient — though the exact cooler size is not specified in the data. The card’s dimensions are not listed, but the single-slot width means it will fit in most standard cases without blocking adjacent slots.

The PCI bus interface provides all necessary power, eliminating the need for 6-pin or 8-pin connectors. This is a significant advantage for older power supplies that lack such connectors. The 2x VGA and 1x S-Video outputs do not add any power draw, as they are simple analog signals. The memory’s 166 MHz clock and 128-bit bus are low-power operations, contributing to the overall efficiency. The 200 W PSU recommendation is generous for a card of this class, and most systems from the year 2000 would have met or exceeded this requirement. Users with modern power supplies will have no issues, as even a 300 W unit provides ample headroom. The end-of-life production status means no new units are available, so any purchase is for a used or new-old-stock card, but the cooling and power requirements remain trivial to satisfy.