NVIDIA GeForce GTX 680 vs NVIDIA Quadro M2000 Comparison

Head-to-Head Benchmarks

The data available for direct comparison shows a clear pattern: the NVIDIA GeForce GTX 680 wins both benchmark tests included in the head-to-head set. These are not marginal victories; the deltas are substantial and consistent across different compute APIs.

Starting with Geekbench OpenCL, the GTX 680 scores 16,906 points while the Quadro M2000 scores 14,588. That represents a 13.7% advantage for the older GeForce card. The gap is even wider in Geekbench Vulkan, where the GTX 680 posts 17,646 against the Quadro's 14,475, a delta of 18%. Both tests point in the same direction: the GTX 680 delivers higher raw compute throughput in these synthetic workloads.

The average benchmark scores tell a slightly different story. The Quadro M2000 averages 14,532 across its two recorded tests, while the GTX 680 averages 14,150 across three tests. The GTX 680's average is dragged down by its Geekbench Metal score of 7,897, a test the Quadro does not have. If we isolate only the shared benchmarks, the GTX 680 is ahead in both, but the overall average is closer than the head-to-head deltas suggest. This is because the Vulkan and OpenCL results for the GTX 680 are higher than its Metal result, which pulls its mean downward relative to the Quadro's more consistent pair of scores.

Looking at percentile rankings, the two cards are nearly level. The Quadro M2000 sits at the 56th percentile among all GPUs, while the GTX 680 sits at the 55th. These are effectively identical positions in the overall distribution, despite the GTX 680's clear wins in the direct comparisons. The nearest rival data reinforces this picture. The Quadro M2000's closest competitors include the GeForce GTX 965M (0.9% higher average score), the Radeon RX Vega 11 (1% higher), and the GeForce GTX TITAN (1.1% higher), with the Radeon RX 5500 XT trailing by 1.1%. For the GTX 680, the nearest rivals are the Tesla K10 (0.9% higher), the GeForce GTX 1070 Ti (0.9% lower), the Intel Iris Xe MAX Graphics (1.2% lower), and the Radeon Vega 11 (1.4% lower). Both cards occupy a narrow performance band where small percentage swings separate neighbors.

The key takeaway from the head-to-head data is that the GTX 680 holds a decisive edge in the two tests both cards took. The OpenCL margin of 13.7% and the Vulkan margin of 18% are not close calls. These are sizable gaps that indicate the GTX 680's architecture handles these workloads more efficiently, despite being released four years earlier and drawing significantly more power.

The Verdict

From the benchmark data alone, the choice is straightforward for anyone prioritizing raw compute performance: the NVIDIA GeForce GTX 680 wins both shared tests and should be selected when OpenCL or Vulkan throughput is the primary concern. The GTX 680 is 13.7% ahead in OpenCL and 18% ahead in Vulkan, margins that would be noticeable in any workload sensitive to these APIs.

However, the Quadro M2000 has its own case. Its average benchmark score of 14,532 is higher than the GTX 680's 14,150, driven largely by the GTX 680's weaker Metal showing. The Quadro also posts a slightly better percentile rank at 56 versus 55. For users who need consistent performance across a broader set of tests, or who specifically value the higher average, the Quadro is the defensible pick.

The data does not support a universal recommendation. The GTX 680 is the clear winner in the head-to-head compute tests, but the Quadro M2000 offers a higher average score and a marginally better percentile position. The deciding factor should be which benchmark matters more: if Vulkan or OpenCL performance is critical, the GTX 680 is the answer; if overall average score across mixed workloads is the metric, the Quadro M2000 edges ahead. The GTX 680's launch MSRP of 499 USD is the only pricing data available, and no similar figure exists for the Quadro.

Architecture Differences

The two cards come from different architectural generations, and the data shows how that plays out. The Quadro M2000 uses the GM206 chip on the Maxwell 2.0 architecture, while the GTX 680 uses the GK104 chip on the Kepler architecture. Both are built on a 28 nm process at TSMC, so the manufacturing node is identical, but the underlying designs differ substantially.

Transistor counts and die sizes reflect the generational gap. The GTX 680 packs 3,540 million transistors onto a 294 mm² die, while the Quadro M2000 has 2,940 million transistors on a smaller 228 mm² die. The transistor density tells the story: the Quadro achieves 12.9 million transistors per square millimeter, while the GTX 680 manages 12.0 million. The newer Maxwell design fits more logic into less space, a sign of architectural efficiency improvements even on the same process node.

The memory systems diverge in ways that matter for bandwidth-sensitive tasks. The GTX 680 uses a 256-bit memory bus with 2 GB of GDDR5, delivering 192.3 GB/s of bandwidth. The Quadro M2000 uses a 128-bit bus with 4 GB of GDDR5, but its bandwidth drops to 105.8 GB/s. The GTX 680 has nearly double the memory bandwidth, which helps explain its advantage in compute benchmarks.

Shader configuration also differs sharply. The GTX 680 has 1,536 shading units and 128 texture mapping units, while the Quadro M2000 has 768 shading units and 48 TMUs. Both have 32 ROPs. The GTX 680's higher shader count directly contributes to its FP32 throughput of 3.250 TFLOPS, compared to the Quadro's 1.786 TFLOPS. The GTX 680 also leads in texture rate at 135.4 GTexel/s versus 55.82 GTexel/s, though the Quadro edges ahead in pixel rate at 37.22 GPixel/s against 33.86 GPixel/s.

API support shows a split. The Quadro M2000 supports DirectX 12 with feature level 12_1, while the GTX 680 supports DirectX 12 with feature level 11_0. Vulkan support also differs: the Quadro lists Vulkan 1.4, while the GTX 680 lists Vulkan 1.2.175. OpenGL 4.6 is common to both. These differences reflect the newer Maxwell architecture's more complete support for modern graphics APIs.

Specification Differences

The spec sheet highlights several areas where the two cards diverge beyond the architectural basics. Clock speeds differ notably: the GTX 680 runs at 1006 MHz base and 1058 MHz boost, while the Quadro M2000 runs at 796 MHz base and 1163 MHz boost. The Quadro has a higher boost clock but a lower base clock, while the GTX 680 has a narrower frequency range overall.

Memory configuration is another differentiator. The Quadro M2000 has 4 GB of GDDR5 on a 128-bit bus with 6.6 Gbps effective speed, yielding 105.8 GB/s bandwidth. The GTX 680 has 2 GB of GDDR5 on a 256-bit bus with 6 Gbps effective speed, yielding 192.3 GB/s. The Quadro has double the capacity but roughly half the bandwidth.

Power and cooling requirements are starkly different. The Quadro M2000 draws 75 W with no power connectors and a suggested PSU of 250 W, fitting in a single-slot design. The GTX 680 draws 195 W, requires two 6-pin power connectors, and needs a 450 W PSU with a dual-slot cooler. Physical dimensions follow suit: the Quadro is 201 mm long and 111 mm tall, while the GTX 680 is 256 mm long, 111 mm tall, and 38 mm wide. Both use PCIe 3.0 x16 interfaces.

Display outputs differ as well. The Quadro M2000 offers four DisplayPort 1.2 outputs, while the GTX 680 provides two DVI, one HDMI 1.4a, and one DisplayPort 1.2. The Quadro's multi-display focus is evident from its output configuration, while the GTX 680 targets more conventional consumer setups.

FAQ

Q: Which card performs better in Geekbench Vulkan?

A: The NVIDIA GeForce GTX 680 scores 17,646 in Vulkan, which is 18% higher than the Quadro M2000's 14,475.

Q: Does the Quadro M2000 have a higher average benchmark score?

A: Yes, the Quadro M2000 averages 14,532 across its tests, while the GTX 680 averages 14,150. However, the GTX 680's average includes a lower Metal score of 7,897, which the Quadro does not have.

Q: What is the memory bandwidth difference between the two cards?

A: The GTX 680 has 192.3 GB/s of bandwidth from its 256-bit bus, while the Quadro M2000 has 105.8 GB/s from its 128-bit bus.

Q: Which card has more shading units?

A: The GTX 680 has 1,536 shading units, exactly double the Quadro M2000's 768 units.

Q: How do their power requirements compare?

A: The Quadro M2000 draws 75 W with no power connectors and a 250 W suggested PSU, while the GTX 680 draws 195 W with two 6-pin connectors and a 450 W suggested PSU.

Q: What DirectX feature levels do they support?

A: The Quadro M2000 supports DirectX 12 with feature level 12_1, while the GTX 680 supports DirectX 12 with feature level 11_0.

Where Each One Wins

The GTX 680 wins decisively in the two shared benchmark tests. Its 18% Vulkan advantage and 13.7% OpenCL advantage make it the clear choice for compute-heavy workloads using those APIs. The GTX 680 also leads in raw specifications that drive compute performance: higher FP32 throughput (3.250 TFLOPS versus 1.786 TFLOPS), more shading units (1,536 versus 768), more TMUs (128 versus 48), and significantly higher memory bandwidth (192.3 GB/s versus 105.8 GB/s). Users running OpenCL or Vulkan workloads that scale with shader count and bandwidth will see better results on the GTX 680.

The Quadro M2000 wins on average benchmark score, posting 14,532 against the GTX 680's 14,150. It also holds a one-point percentile advantage at 56 versus 55. The Quadro's higher memory capacity (4 GB versus 2 GB) makes it better suited for tasks that need to hold larger datasets in VRAM, even if the bandwidth is lower. Its 75 W power draw, single-slot design, and lack of power connectors make it far easier to integrate into space- or power-constrained systems. The Quadro's four DisplayPort outputs also make it the better option for multi-display professional setups. For users who value the higher average score, larger memory pool, or lower power footprint, the Quadro M2000 is the stronger pick.

The data does not crown a single winner for all scenarios. The GTX 680 dominates the compute benchmarks and the Quadro M2000 holds the edge in average score and practical deployment flexibility. Choose based on which of these factors matters most for the intended workload.