NVIDIA Quadro K200M

The NVIDIA Quadro K200M is a Kepler-generation mobile workstation GPU fabricated by TSMC on a 28 nm process. It integrates 1,270 million transistors on a 118 mm² die, achieving a transistor density of 10.8 million per mm². Released in 2014 and now classified as end-of-life, the K200M holds the 50th percentile position among all GPUs in the database, though its average benchmark score is 0 — an indication that no measured performance entries are recorded for this part.

Memory Subsystem

The K200M is equipped with 2 GB of DDR3 memory on a 64-bit bus, yielding a peak bandwidth of 14.40 GB/s. The memory clock runs at 900 MHz, which translates to 1800 Mbps effective data rate. This is a deliberately constrained memory configuration. A 64-bit bus is narrow, and DDR3 is a low-bandwidth memory type; together, they cap the card's ability to feed its shading units.

For high-resolution workloads, the consequences are significant. The 14.40 GB/s ceiling means that large frame buffers, high-resolution textures, and multi-sample anti-aliasing will all contend for a very limited pool of bandwidth. The 2 GB capacity is generous for the card's era, but capacity without bandwidth is of limited use. At high resolutions, the pixel rate of 2.980 GPixel/s and the texture rate of 11.92 GTexel/s will saturate quickly, and the memory subsystem will become the primary bottleneck. In practice, this card is best suited to low-resolution, low-detail scenarios where the bandwidth constraint is less binding.

The 64-bit bus also has implications for power and die area. A narrower bus requires fewer I/O pins and less on-chip memory controller logic, which helps keep the TDP at 35 W. However, the trade-off is a hard ceiling on memory throughput that cannot be overcome by clock speed alone. The 900 MHz memory clock is not high enough to compensate for the narrow bus. The 14.40 GB/s figure is the single most limiting specification on this card, as it directly bounds how quickly textures and geometry data can be moved into the shading units.

Ray Tracing and Feature Set

The K200M is built around the GK107 chip, a Kepler-architecture design with 192 shading units, 16 texture mapping units, and 8 raster operation units. The fact pack lists no ray tracing cores and no tensor cores — the hardware simply has no dedicated units for these workloads. Ray tracing, if attempted, would be executed as general-purpose compute on the 192 shading units, which is impractical given the FP32 throughput of 286.1 GFLOPS.

The API support is broader than the hardware would suggest. The card supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The DirectX 12 entry is explicitly qualified with the 11_0 feature level, meaning the card is DX12-compatible at the 11_0 feature tier — a typical limitation for Kepler parts. OpenGL 4.6 and Vulkan 1.2.175 provide modern API access, but without RT or tensor hardware, the feature set remains firmly rasterization-oriented.

The 8 ROPs are the smallest configuration in the Kepler lineup, and the 16 TMUs are equally modest. These counts, combined with the 745 MHz clock (which is both base and boost, with no headroom), define a part that is built for basic graphics output rather than compute-heavy tasks. The pixel rate of 2.980 GPixel/s and texture rate of 11.92 GTexel/s are direct consequences of these limited resources. There is no dynamic clock scaling on this part; the fixed 745 MHz frequency means performance is entirely predictable but also entirely capped.

Benchmark Performance

The database records an average benchmark score of 0 for the K200M, and the benchmarks array is empty. This means there are no measured performance results to analyze. The card's performance must therefore be inferred from its architectural parameters and its 50th percentile ranking.

The 50th percentile placement is a median position: half of all GPUs in the database rank higher, and half rank lower. However, given the absence of benchmark data, this percentile is likely a specification-derived estimate rather than a measured outcome. The FP32 throughput of 286.1 GFLOPS is the key compute metric. It places the K200M firmly in the entry-level mobile segment, though no direct rivals are listed in the database to anchor the comparison.

The fixed 745 MHz clock, with boost equal to base, means the card does not dynamically overclock. This is a deliberate choice for a 35 W TDP part; there is no thermal or power headroom for boost behavior. The pixel rate of 2.980 GPixel/s and texture rate of 11.92 GTexel/s are modest figures that will limit performance in fill-rate-bound scenes, such as those with heavy alpha blending or large texture fetches. The 286.1 GFLOPS FP32 figure, when divided across the 192 shading units, yields a per-unit throughput consistent with the 745 MHz clock — there are no hidden performance reserves.

In real-world terms, the K200M would be expected to handle legacy DirectX 12 (11_0) titles at low settings and modest resolutions. The 2 GB frame buffer is sufficient for older games, but the 14.40 GB/s bandwidth will cause stuttering in scenes that stream large textures. The 50th percentile rank suggests that, within the database's universe, the K200M is an average performer — but the zero benchmark score means this rank is not backed by actual test data. The absence of measured results is itself a signal: the card is old enough and niche enough that no standardized benchmark runs have been recorded.

Who Should Consider It

The K200M is an end-of-life product, so its relevance is limited to legacy systems and refurbished mobile workstations. The MXM-A (3.0) bus interface and the MXM module slot width indicate a modular laptop design. The display outputs are listed as "portable device dependent," meaning the actual video connectors are determined by the host laptop rather than the card itself.

For users maintaining a Kepler-era workstation laptop, the K200M provides a 2 GB frame buffer and 14.40 GB/s of bandwidth — sufficient for basic CAD viewing, 2D design, and office productivity. The 35 W TDP and the absence of power connectors make it a straightforward replacement in compatible MXM slots. No auxiliary power is required, which simplifies installation and keeps thermal demands low.

The card is not suitable for high-resolution gaming, ray-traced workloads, or compute-intensive tasks. The 286.1 GFLOPS FP32 throughput and the absence of RT/tensor hardware are clear indicators. Users who need modern graphics features should look to the successor, the Quadro Maxwell-M generation, or to more recent parts. The K200M occupies a narrow niche: it is a low-power, legacy mobile workstation GPU that keeps older laptops functional. Its 50th percentile ranking suggests it is not an outlier in either direction — it is a middle-of-the-road part that does exactly what it was designed to do, no more.

How It Compares

The fact pack lists no nearest rivals for the K200M, so a direct numeric comparison against competing GPUs is not possible from the available data. The card's position is defined by its 50th percentile placement among all GPUs and by its architectural lineage.

The predecessor, Quadro Fermi-M, represents the prior architecture generation. Moving from Fermi to Kepler brought the 28 nm process and a transistor count of 1,270 million on a 118 mm² die. The successor, Quadro Maxwell-M, follows with the Maxwell architecture. Without benchmark scores for any of these parts, the comparison is architectural rather than numerical. The K200M sits between these two generations, inheriting Kepler's efficiency improvements over Fermi while lacking the feature advances of Maxwell.

Against the broader GPU landscape, the 50th percentile is a median placement. The K200M is neither a high-end part nor a bottom-tier one — it sits exactly in the middle of the database's distribution. However, the zero average benchmark score tempers this ranking; the percentile is likely a specification-based estimate rather than a measured result. The 286.1 GFLOPS FP32 throughput and 14.40 GB/s bandwidth are the numbers that will actually determine its behavior in the field.

In the absence of rival entries, the K200M is best understood as a low-power Kepler workstation part that fills a narrow niche: legacy mobile workstations that need a 2 GB frame buffer and a 35 W power draw. Its 50th percentile rank indicates that, within the database's universe of GPUs, it is an average performer — but the average is heavily influenced by the absence of benchmark data. The card's longevity in the database confirms its role as a reliable, if unremarkable, entry point in the Quadro mobile lineup.