GPU Comparison

# NVIDIA Quadro 5000 vs NVIDIA GeForce GTX 670M

The data places these two Fermi-based NVIDIA parts in adjacent performance tiers, but with very different design priorities. The Quadro 5000, a professional workstation card from 2011, scores 7,289 in Geekbench OpenCL, which is 11.9% ahead of the GTX 670M's 6,513. The GTX 670M, a mobile part from 2012, sits at the 38th percentile of all GPUs, while the Quadro 5000 reaches the 40th percentile. Both are end-of-life products, but the benchmark results indicate that the Quadro 5000 holds a clear, if modest, compute advantage. The verdict from the data is straightforward: the Quadro 5000 is the stronger performer for OpenCL workloads, while the GTX 670M is the more power-efficient mobile solution.

The Verdict

For any workload measured by Geekbench OpenCL, the Quadro 5000 is the superior card. Its score of 7,289 beats the GTX 670M's 6,513 by 11.9%, a meaningful gap in compute performance. The Quadro 5000 also outranks its nearest rivals, sitting just 0.9% above the GeForce GTX 750 and 1.2% above the AMD Radeon Vega 8 Mobile, confirming it competes with a slightly higher tier of hardware. The GTX 670M, by contrast, is only 0.3% ahead of the GeForce GT 555M and 0.5% ahead of the Quadro M5000M, indicating it is clustered with lower-end parts. However, the GTX 670M's 75 W TDP versus the Quadro 5000's 152 W TDP makes it the only sensible choice for portable systems. The Quadro 5000 requires a dual-slot chassis, a 6-pin power connector, and a 450 W power supply, none of which fit in a laptop. The GTX 670M, as an MXM module with no power connectors, is designed for exactly that environment. Strictly from the data, desktop users with a need for OpenCL performance pick the Quadro 5000; laptop users pick the GTX 670M.

Architecture Differences

Both GPUs are built on TSMC's 40 nm process and share the Fermi family lineage, but they diverge significantly in their specific implementations. The Quadro 5000 uses the GF100 chip, the original Fermi architecture, while the GTX 670M uses GF114, which NVIDIA labels as "Fermi 2.0." This architectural revision is evident in the transistor counts: the Quadro 5000 packs 3,100 million transistors on a 529 mm² die, whereas the GTX 670M has 1,950 million transistors on a 332 mm² die. Both have identical transistor densities at 5.9M per mm², but the GF100 is a much larger, more complex chip. The Quadro 5000's memory configuration reflects its professional orientation: 2.5 GB of GDDR5 on a 320-bit bus, yielding 120.0 GB/s of bandwidth. The GTX 670M is more modest, with 1536 MB of GDDR5 on a 192-bit bus, delivering 72.00 GB/s. Both run memory at 750 MHz (3 Gbps effective). The shading unit counts are close, 352 on the Quadro 5000 versus 336 on the GTX 670M, but the GTX 670M has more texture mapping units (56 versus 44) and fewer ROPs (24 versus 40). The GTX 670M's higher texture rate (33.49 GTexel/s versus 22.57 GTexel/s) suggests it was optimized for fill-rate-heavy gaming workloads, while the Quadro 5000's higher pixel rate (11.29 GPixel/s versus 8.372 GPixel/s) points to its rendering-focused professional role. The Quadro 5000 also supports 1x DVI and 2x DisplayPort outputs, while the GTX 670M's display outputs are listed as "Portable Device Dependent," reflecting its mobile nature.

Head-to-Head Benchmarks

The only direct benchmark comparison in the data is Geekbench OpenCL, and the Quadro 5000 wins it decisively. The Quadro 5000 scores 7,289 against the GTX 670M's 6,513, a delta of 11.9%. This is not a marginal victory; it is a substantial performance gap that places the Quadro 5000 roughly one full tier above its mobile counterpart. To contextualize, the GTX 670M's score of 6,513 puts it within 0.3% of the GeForce GT 555M and 0.5% of the Quadro M5000M, meaning it is essentially at parity with those parts. The Quadro 5000's 7,289, meanwhile, is 0.9% above the GeForce GTX 750 and 1.6% above the GeForce GTX 560 SE. The delta between the two cards in question is larger than the delta between the Quadro 5000 and any of its four nearest rivals, which underscores that the GTX 670M is not merely a slightly slower version of the Quadro 5000, it is a fundamentally different class of hardware. The Quadro 5000's FP32 throughput of 722.3 GFLOPS is lower than the GTX 670M's 803.7 GFLOPS, yet the Quadro still wins the OpenCL test, indicating that memory bandwidth (120.0 GB/s versus 72.00 GB/s) and ROP count (40 versus 24) play a larger role in this particular workload than raw shader math. This is a notable inversion: the GTX 670M has higher theoretical compute but loses in practice.

Specification Differences

The two cards differ in nearly every measurable specification except for memory clock, process node, foundry, and transistor density. The Quadro 5000 uses the GF100 chip; the GTX 670M uses GF114. The Quadro 5000's architecture is "Fermi," while the GTX 670M's is "Fermi 2.0." The Quadro 5000 has 3,100 million transistors on a 529 mm² die; the GTX 670M has 1,950 million on 332 mm². Memory size differs: 2.5 GB versus 1536 MB. Bus width differs: 320-bit versus 192-bit. Bandwidth differs: 120.0 GB/s versus 72.00 GB/s. Shading units are 352 versus 336. TMUs are 44 versus 56. ROPs are 40 versus 24. Pixel rate is 11.29 GPixel/s versus 8.372 GPixel/s. Texture rate is 22.57 GTexel/s versus 33.49 GTexel/s. FP32 is 722.3 GFLOPS versus 803.7 GFLOPS. TDP is 152 W versus 75 W. Slot width is dual-slot versus MXM module. Power connectors are 1x 6-pin versus none. The Quadro 5000 requires a 450 W suggested PSU; the GTX 670M has no such requirement. Bus interface is PCIe 2.0 x16 versus MXM-B (3.0). Display outputs are 1x DVI and 2x DisplayPort versus "Portable Device Dependent." The Quadro 5000 has dimensions of 248 mm length and 111 mm height; the GTX 670M has no listed dimensions. Release dates differ: the Quadro 5000 launched on 2011-02-22, the GTX 670M on 2012-03-21. The Quadro 5000 has a launch MSRP of 2,499 USD; the GTX 670M has no listed launch MSRP. The Quadro 5000's predecessor is Quadro FX Tesla and successor is Quadro Kepler; the GTX 670M's predecessor is GeForce 500M and successor is GeForce 700M. Both support DirectX 12 (11_0) and OpenGL 4.6; neither lists Vulkan support.

FAQ

Q: Which GPU has higher OpenCL benchmark scores?

A: The NVIDIA Quadro 5000 scores 7,289 in Geekbench OpenCL, which is 11.9% higher than the NVIDIA GeForce GTX 670M's score of 6,513.

Q: Are both GPUs based on the same architecture?

A: Both are based on Fermi, but the Quadro 5000 uses the original Fermi architecture with the GF100 chip, while the GTX 670M uses Fermi 2.0 with the GF114 chip.

Q: Which GPU consumes less power?

A: The GTX 670M has a 75 W TDP, which is exactly half of the Quadro 5000's 152 W TDP. The GTX 670M also has no power connectors, while the Quadro 5000 requires one 6-pin connector.

Q: How do these cards compare to their nearest rivals?

A: The Quadro 5000 is 0.9% ahead of the GeForce GTX 750 and 1.2% ahead of the AMD Radeon Vega 8 Mobile. The GTX 670M is 0.3% ahead of the GeForce GT 555M and 0.5% ahead of the Quadro M5000M, but 0.6% behind the Intel UHD Graphics P750.

Q: What are the memory bandwidth differences?

A: The Quadro 5000 has 120.0 GB/s bandwidth from 2.5 GB of GDDR5 on a 320-bit bus. The GTX 670M has 72.00 GB/s bandwidth from 1536 MB of GDDR5 on a 192-bit bus.

Q: Which GPU has more texture mapping units?

A: The GTX 670M has 56 TMUs, compared to the Quadro 5000's 44 TMUs. This gives the GTX 670M a higher texture rate of 33.49 GTexel/s versus 22.57 GTexel/s.

Where Each One Wins

The Quadro 5000 wins in the only direct benchmark comparison available, taking the Geekbench OpenCL test by 11.9%. It also wins on memory capacity (2.5 GB versus 1536 MB), memory bandwidth (120.0 GB/s versus 72.00 GB/s), ROP count (40 versus 24), and pixel rate (11.29 GPixel/s versus 8.372 GPixel/s). Its 40th percentile ranking versus the GTX 670M's 38th percentile confirms that it sits in a slightly higher performance tier overall. The Quadro 5000 is the choice for any desktop workstation where OpenCL compute performance and high-bandwidth memory access are critical. Its dual-slot form factor, 6-pin power connector, and 450 W suggested PSU requirement mean it needs a full desktop chassis with adequate power delivery.

The GTX 670M wins on power efficiency, with a 75 W TDP that is less than half of the Quadro 5000's 152 W. It also wins on texture throughput, delivering 33.49 GTexel/s versus 22.57 GTexel/s, and on raw FP32 compute at 803.7 GFLOPS versus 722.3 GFLOPS. Its MXM Module form factor with no power connectors makes it the only one of the two that can be installed in a laptop. The GTX 670M is the choice for mobile gaming systems where power draw and physical size are paramount constraints, even if it sacrifices some OpenCL performance. The data shows a clear split: the Quadro 5000 is a professional desktop card that wins on memory and rendering throughput, while the GTX 670M is a mobile part that wins on efficiency and texture fill. Each dominates in its intended environment, and neither is a substitute for the other.