NVIDIA Quadro K2200 vs NVIDIA Tesla M2090 Comparison

The Verdict

The recorded data presents a clear split between these two end-of-life NVIDIA workstation cards. The NVIDIA Tesla M2090 wins the only head-to-head benchmark available in the database, the Geekbench OpenCL test, with a score of 13,075 against the Quadro K2200’s 11,431, a margin of 14.4%. The Tesla M2090 also holds a higher percentile rank among all GPUs at 53, while the Quadro K2200 sits at 49. For raw compute throughput, the Tesla M2090 is the pick, as its average benchmark score of 13,075 places it within 1% of the GeForce GTX 1660 SUPER and ahead of the GeForce RTX 3050 Ti Mobile by 1%.

However, the Quadro K2200 is not without its own arguments. It is the only card in this comparison with any Vulkan support in the database, listing Vulkan 1.4, whereas the Tesla M2090 has no Vulkan API entry. The K2200 also draws 68 W versus the M2090’s 250 W, and it requires no auxiliary power connectors, while the M2090 needs one 6-pin and one 8-pin connector. The K2200 has display outputs, 1x DVI and 2x DisplayPort 1.2, while the M2090 has no outputs at all. If the workload involves driving a monitor or a Vulkan-based application, the Quadro K2200 is the only viable choice from this pair. If the workload is pure compute, the Tesla M2090’s OpenCL lead and higher percentile make it the data-backed winner.

FAQ

Q: Which card is faster in the database’s OpenCL benchmark?

A: The NVIDIA Tesla M2090 scores 13,075 in Geekbench OpenCL, which is 14.4% higher than the Quadro K2200’s 11,431 in the same test.

Q: Does either card support Vulkan?

A: Only the Quadro K2200 has a Vulkan entry in the database, listing Vulkan 1.4. The Tesla M2090 has no Vulkan API listed.

Q: Which card can connect to a display?

A: The Quadro K2200 has display outputs: 1x DVI and 2x DisplayPort 1.2. The Tesla M2090 has no display outputs.

Q: How do their power requirements differ?

A: The Tesla M2090 has a TDP of 250 W and requires one 6-pin and one 8-pin power connector, with a suggested power supply of 600 W. The Quadro K2200 has a TDP of 68 W, needs no power connectors, and suggests a 250 W power supply.

Q: Which card has a higher average benchmark score across all tests?

A: The Tesla M2090 has an average benchmark score of 13,075 from its single OpenCL result. The Quadro K2200 has an average of 10,761, which includes its OpenCL score of 11,431 and a Vulkan score of 10,090.

Q: How do they compare in memory capacity?

A: The Tesla M2090 has 6 GB of GDDR5 memory on a 384-bit bus, while the Quadro K2200 has 4 GB of GDDR5 on a 128-bit bus. The M2090’s bandwidth is 177.4 GB/s versus 80.19 GB/s for the K2200.

Architecture Differences

The two cards come from different NVIDIA architectures and process nodes, which explains most of their behavioral differences. The Tesla M2090 uses the GF110 chip built on Fermi 2.0 architecture, fabricated on a 40 nm process at TSMC. It packs 3,000 million transistors on a 520 mm² die, yielding a transistor density of 5.8M per mm². The Quadro K2200 uses the GM107 chip on Maxwell architecture, also from TSMC, but on a 28 nm process. It contains 1,870 million transistors on a 148 mm² die, giving a much higher transistor density of 12.6M per mm².

The shading unit counts differ: the M2090 has 512 shading units, while the K2200 has 640. However, the M2090 counters with more texture mapping units, 64 versus 40, and far more raster output pipelines, 48 versus 16. This structural difference suggests the M2090 is built for heavier geometry and pixel throughput, while the K2200 leans on more shader-level parallelism.

The API support reveals another split. Both cards list DirectX 12 (11_0) and OpenGL 4.6. The K2200 adds Vulkan 1.4, while the M2090 has no Vulkan entry. The M2090’s display output is listed as "No outputs", making it a pure compute or rendering accelerator. The K2200 has 1x DVI and 2x DisplayPort 1.2, so it can serve as a workstation display card. The production status for both is end-of-life, but their release dates are three years apart: the M2090 came out in July 2011, the K2200 in July 2014.

Specification Differences

The most obvious specification gap is memory. The Tesla M2090 has 6 GB of GDDR5 with a 384-bit bus, delivering 177.4 GB/s of bandwidth. The Quadro K2200 has 4 GB of GDDR5 on a 128-bit bus, delivering 80.19 GB/s. That is more than double the bandwidth for the M2090. The memory clock also differs: the M2090 runs at 924 MHz, which the database lists as 3.7 Gbps effective, while the K2200 runs at 1253 MHz, given as 5 Gbps effective. The K2200’s faster memory clock does not compensate for its narrower bus.

Clock speeds are only listed for the K2200: base 1046 MHz and boost 1124 MHz. The M2090 has no base or boost clock entries, only its memory clock. In terms of compute rates, the M2090 delivers 1,332.2 GFLOPS FP32, while the K2200 delivers 1,438.7 GFLOPS FP32. So the K2200 has a slight FP32 advantage on paper, yet it loses the OpenCL benchmark.

Pixel and texture rates also differ: the M2090 has 20.83 GPixel/s and 41.66 GTexel/s, while the K2200 has 17.98 GPixel/s and 44.96 GTexel/s. The M2090 wins pixel fill, the K2200 wins texture fill. Power and physical size separate them further: the M2090 is a dual-slot card at 250 W, 248 mm long, while the K2200 is single-slot at 68 W, 202 mm long and 111 mm high. The M2090 needs 1x 6-pin and 1x 8-pin power, the K2200 needs none. Suggested PSU is 600 W for the M2090 versus 250 W for the K2200.

Head-to-Head Benchmarks

The database contains one direct head-to-head test: Geekbench OpenCL. The Tesla M2090 scores 13,075, and the Quadro K2200 scores 11,431. The delta is 14.4% in favor of the M2090. This is a substantial margin, given that the M2090’s nearest rivals cluster very close to its score. Its nearest rival, the GeForce GTX 1660 SUPER, averages 12,986, only 0.7% behind the M2090. The next rival, the GeForce GTX 950, averages 13,189, which is 0.9% ahead of the M2090. So the M2090 sits within a tight 1% band around its peer group.

The Quadro K2200’s OpenCL score of 11,431 is its stronger result. Its Vulkan score is lower at 10,090, which pulls its average down to 10,761. The K2200’s nearest rivals are similarly clustered: the AMD Radeon Pro 450 averages 10,804, which is 0.4% behind the K2200, and the GeForce GTX 560 Ti averages 10,690, which is 0.7% behind. The K2200 also sits ahead of the Radeon RX 6600S by 1.2%, but trails the GeForce MX350 by 1.1%.

The head-to-head delta of 14.4% is much larger than the gaps within each card’s rival group. That means the M2090 and K2200 are not in the same performance class according to this metric. The M2090’s average benchmark score of 13,075 is 21.5% higher than the K2200’s average of 10,761, even though the K2200 has a higher FP32 GFLOPS rating. This suggests real-world OpenCL workloads favor the M2090’s memory bandwidth and ROP count over the K2200’s shader count.

Where Each One Wins

The Tesla M2090 wins in raw compute and memory throughput. Its OpenCL score is 14.4% higher, its average benchmark score is 21.5% higher, and its memory bandwidth is 177.4 GB/s versus 80.19 GB/s. It also has 6 GB of memory, 2 GB more than the K2200, and a 384-bit bus that the K2200 cannot match. The M2090’s pixel rate of 20.83 GPixel/s exceeds the K2200’s 17.98 GPixel/s, which matters for fill-rate-bound rendering tasks. Its percentile rank of 53 versus 49 confirms that, across the entire database, the M2090 sits above the K2200 in overall standing.

The Quadro K2200 wins in efficiency and flexibility. It draws 68 W, less than a third of the M2090’s 250 W, and requires no power connectors, so it can drop into a system with a 250 W power supply. Its single-slot design at 202 mm length is easier to fit in small chassis. It has display outputs, making it a functional workstation card for desktop use. It also has Vulkan 1.4 support, which the M2090 lacks entirely, so any Vulkan-based application only runs on the K2200. Its FP32 rating of 1,438.7 GFLOPS is actually higher than the M2090’s 1,332.2 GFLOPS, and its texture rate of 44.96 GTexel/s beats the M2090’s 41.66 GTexel/s.

The use-case split is therefore straightforward. For compute-heavy workloads that rely on OpenCL and large memory buffers, the Tesla M2090 is the data-backed choice. For a workstation that needs to output video, run Vulkan, or operate within a low-power budget, the Quadro K2200 is the only option that satisfies those requirements. The M2090’s lack of display outputs disqualifies it from any interactive desktop scenario, while the K2200’s lower bandwidth and smaller memory pool limit its appeal for large data sets. Each card wins where its architecture and feature set align with the workload.