NVIDIA GeForce GTX 950 vs NVIDIA Tesla M2090 Comparison

# NVIDIA GeForce GTX 950 vs NVIDIA Tesla M2090

The NVIDIA GeForce GTX 950 and NVIDIA Tesla M2090 occupy different corners of the GPU landscape, yet their average benchmark scores place them within 0.9% of each other. The GTX 950, a Maxwell 2.0 consumer card from 2015, edges ahead with an average benchmark score of 13,189 versus the Tesla M2090's 13,075. Both sit at the 53rd percentile among all GPUs, making them statistically equivalent in overall performance despite radically different architectures, memory configurations, and intended workloads. The head-to-head data, however, reveals a clear winner in the only directly comparable test, with the GTX 950 taking the Geekbench OpenCL contest by 19.8%.

Where Each One Wins

The GTX 950 wins the only direct benchmark comparison available, but the story extends beyond raw scores. In the Geekbench OpenCL test, the GTX 950 scores 15,662 against the Tesla M2090's 13,075, a 19.8% advantage. This consumer-oriented card also carries broader API support, with DirectX 12 (12_1) and Vulkan 1.4 support, while the Tesla M2090 is limited to DirectX 12 (11_0) with no Vulkan support at all. For gaming and general-purpose compute on modern APIs, the GTX 950 is unequivocally the stronger choice.

The Tesla M2090, by contrast, wins on memory capacity and bandwidth. Its 6 GB GDDR5 frame buffer dwarfs the GTX 950's 2 GB, and its 384-bit memory bus delivers 177.4 GB/s of bandwidth compared to 105.8 GB/s on the GTX 950. This makes the Tesla better suited for memory-bound workloads that exceed the GTX 950's capacity, despite its lower raw compute throughput. The Tesla also has a higher texture fill rate per texture mapping unit, though its overall texture rate of 41.66 GTexel/s trails the GTX 950's 57.02 GTexel/s.

In practical terms, the GTX 950 wins for interactive graphics, modern API compatibility, and compute tasks that fit within 2 GB of memory. The Tesla M2090 wins for large datasets, scientific computing, and any workload where 6 GB of VRAM is a hard requirement. The data does not show the Tesla winning a single benchmark, but its memory configuration is a differentiator that raw scores cannot capture.

Architecture Differences

The GTX 950 uses the GM206 chip built on Maxwell 2.0 architecture, manufactured by TSMC on a 28 nm process. It packs 2,940 million transistors into a 228 mm² die, yielding a transistor density of 12.9 million per square millimeter. The Tesla M2090 uses the GF110 chip on the older Fermi 2.0 architecture, also from TSMC but on a 40 nm process. Its 3,000 million transistors occupy a much larger 520 mm² die, resulting in a significantly lower density of 5.8 million per square millimeter. The process node advantage is stark: 28 nm versus 40 nm means the GTX 950 achieves comparable transistor counts in less than half the die area.

The shading unit counts differ notably. The GTX 950 has 768 shading units, while the Tesla M2090 has 512. The Tesla compensates with more texture mapping units (64 versus 48) and more ROPs (48 versus 32). Clock behavior also diverges: the GTX 950 has defined base and boost clocks of 1024 MHz and 1188 MHz, while the Tesla M2090's clocks are not listed, suggesting it relies on a fixed operating point typical of compute-focused accelerators.

Memory architectures reflect their different purposes. The GTX 950 uses a 128-bit bus with 2 GB of GDDR5 at 6.6 Gbps effective, producing 105.8 GB/s. The Tesla M2090 uses a 384-bit bus with 6 GB of GDDR5 at 3.7 Gbps effective, producing 177.4 GB/s. The Tesla's wider bus and larger capacity target bandwidth-hungry scientific workloads, while the GTX 950's narrower bus prioritizes cost and power efficiency.

Power and interface differences are equally pronounced. The GTX 950 draws a 90 W TDP with a single 6-pin connector and a 250 W suggested PSU. The Tesla M2090 draws 250 W with a 6-pin plus 8-pin configuration and a 600 W suggested PSU. The GTX 950 uses PCIe 3.0 x16, while the Tesla uses the older PCIe 2.0 x16. The Tesla has no display outputs, confirming its role as a headless compute accelerator, while the GTX 950 offers 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2.

Head-to-Head Benchmarks

The sole direct comparison in the data is the Geekbench OpenCL test. The GTX 950 scores 15,662 against the Tesla M2090's 13,075, a 19.8% margin. This is a substantial gap that reflects the GTX 950's architectural advantages: more shading units (768 versus 512), higher pixel rate (38.02 GPixel/s versus 20.83 GPixel/s), higher texture rate (57.02 GTexel/s versus 41.66 GTexel/s), and higher FP32 throughput (1.825 TFLOPS versus 1,332.2 GFLOPS). The GTX 950's Maxwell design extracts more compute per clock, and its boost clock capability adds further headroom.

The deltaPct values from the nearest rivals put this in context. The GTX 950's average score of 13,189 is 0.9% above the Tesla M2090's 13,075, yet the OpenCL head-to-head shows a 19.8% advantage. This discrepancy arises because the GTX 950 has additional benchmark results — Geekbench Metal at 7,172 and Geekbench Vulkan at 16,734 — that lower its average relative to its OpenCL peak. The Tesla M2090's average is entirely derived from its single OpenCL score of 13,075. The GTX 950's Vulkan score of 16,734 is particularly telling, as it demonstrates strong performance on a modern API that the Tesla cannot even run.

The GTX 950 also shows relative strength against other rivals. It trails the GeForce GTX 480 by 0.8% (13,300 versus 13,189) and the AMD FirePro M6100 by 1.2% (13,354 versus 13,189), but leads the AMD Radeon Pro 555X by 1% (13,321 versus 13,189). The Tesla M2090, meanwhile, leads the GeForce GTX 1660 SUPER by 0.7% (13,075 versus 12,986) and the RTX 3050 Ti Mobile by 1% (13,075 versus 12,940), but trails the Radeon RX 580 by 1.1% (13,075 versus 12,928). These margins are all within noise, reinforcing that both cards sit in a tightly packed performance tier.

Specification Differences

The two cards differ across nearly every specification category, reflecting their divergent design goals.

• Process node: GTX 950 uses 28 nm; Tesla M2090 uses 40 nm.
• Transistor density: GTX 950 has 12.9M / mm²; Tesla M2090 has 5.8M / mm².
• Die size: GTX 950 is 228 mm²; Tesla M2090 is 520 mm².
• Shading units: GTX 950 has 768; Tesla M2090 has 512.
• Texture mapping units: GTX 950 has 48; Tesla M2090 has 64.
• ROP count: GTX 950 has 32; Tesla M2090 has 48.
• Memory size: GTX 950 has 2 GB; Tesla M2090 has 6 GB.
• Memory bus width: GTX 950 has 128 bit; Tesla M2090 has 384 bit.
• Memory bandwidth: GTX 950 has 105.8 GB/s; Tesla M2090 has 177.4 GB/s.
• FP32 throughput: GTX 950 has 1.825 TFLOPS; Tesla M2090 has 1,332.2 GFLOPS.
• Pixel rate: GTX 950 has 38.02 GPixel/s; Tesla M2090 has 20.83 GPixel/s.
• Texture rate: GTX 950 has 57.02 GTexel/s; Tesla M2090 has 41.66 GTexel/s.
• TDP: GTX 950 is 90 W; Tesla M2090 is 250 W.
• Power connectors: GTX 950 uses 1x 6-pin; Tesla M2090 uses 1x 6-pin + 1x 8-pin.
• Suggested PSU: GTX 950 is 250 W; Tesla M2090 is 600 W.
• Bus interface: GTX 950 uses PCIe 3.0 x16; Tesla M2090 uses PCIe 2.0 x16.
• Display outputs: GTX 950 has 1x DVI, 1x HDMI 2.0, 3x DisplayPort 1.2; Tesla M2090 has none.
• DirectX support: GTX 950 supports 12 (12_1); Tesla M2090 supports 12 (11_0).
• Vulkan support: GTX 950 supports 1.4; Tesla M2090 does not support it.
• Card length: GTX 950 is 202 mm (8 inches); Tesla M2090 is 248 mm (9.8 inches).
• Release date: GTX 950 launched 2015-08-19; Tesla M2090 launched 2011-07-24.
• Launch MSRP: GTX 950 was 159 USD; Tesla M2090 has no listed MSRP.

FAQ

Q: Which GPU has higher raw FP32 compute performance?

A: The GTX 950 delivers 1.825 TFLOPS, which is 37% higher than the Tesla M2090's 1,332.2 GFLOPS.

Q: Can the Tesla M2090 be used for display output?

A: No. The Tesla M2090 has no display outputs, whereas the GTX 950 provides 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2.

Q: How much memory bandwidth does each card provide?

A: The Tesla M2090 provides 177.4 GB/s over a 384-bit bus, while the GTX 950 provides 105.8 GB/s over a 128-bit bus. The Tesla's bandwidth advantage is 67.7%.

Q: Which card has better modern API support?

A: The GTX 950 supports DirectX 12 (12_1) and Vulkan 1.4, while the Tesla M2090 only supports DirectX 12 (11_0) and has no Vulkan support.

Q: How do their average benchmark scores compare?

A: The GTX 950 averages 13,189, which is 0.9% higher than the Tesla M2090's 13,075. Both cards rank at the 53rd percentile among all GPUs.

Q: What is the power consumption difference?

A: The GTX 950 has a 90 W TDP with a 250 W suggested PSU, while the Tesla M2090 has a 250 W TDP with a 600 W suggested PSU. The Tesla requires significantly more power.