NVIDIA GeForce GTX 460 vs NVIDIA GeForce GTX 670M Comparison

FAQ

Q: Which GPU has the higher OpenCL benchmark score?

A: The NVIDIA GeForce GTX 460 scores 7,925 in Geekbench OpenCL, while the NVIDIA GeForce GTX 670M scores 6,513. This puts the GTX 460 ahead by 21.7% in the recorded head-to-head measurement.

Q: How does the GTX 460 compare to its closest rivals?

A: The GTX 460 sits just above the NVIDIA Quadro K4100M (7,906, a 0.2% lead) and trails the NVIDIA Quadro P5000 (8,039, a 1.4% gap) and the NVIDIA GeForce GTX 880M (8,040, also a 1.4% gap). It also edges out the NVIDIA GeForce GTX 650 Ti (8,053, a 1.6% difference in favor of the 650 Ti).

Q: Where does the GTX 670M rank among its nearest competitors?

A: The GTX 670M outperforms the NVIDIA GeForce GT 555M (6,493, a 0.3% lead) and the NVIDIA Quadro M5000M (6,481, a 0.5% lead). It also beats the AMD Radeon Vega 10 Mobile (6,476, a 0.6% lead). The Intel UHD Graphics P750 (6,554) is slightly ahead, showing a 0.6% gap in that comparison.

Q: What are the memory capacities of these two cards?

A: The GTX 460 comes with 768 MB of GDDR5 memory on a 192-bit bus, while the GTX 670M carries 1536 MB (1.5 GB) of GDDR5 on the same 192-bit bus width.

Q: Do both GPUs use the same manufacturing process?

A: Yes, both are built on TSMC's 40 nm process node, and both integrate 1,950 million transistors on a 332 mm² die, resulting in a transistor density of 5.9M per mm² for each.

Q: Which card has higher memory bandwidth?

A: The GTX 460 achieves 86.40 GB/s of bandwidth, while the GTX 670M delivers 72.00 GB/s. The GTX 460's advantage comes from its higher effective memory clock of 3.6 Gbps compared to 3 Gbps on the GTX 670M.

Where Each One Wins

The benchmark data points to a clear but narrow set of conditions for each card. The NVIDIA GeForce GTX 460 wins the single recorded OpenCL test, and it does so decisively. In the head-to-head comparison, the GTX 460 posts 7,925 versus the GTX 670M's 6,513, a 21.7% margin. This victory is consistent with its higher pixel rate (9.450 GPixel/s versus 8.372 GPixel/s), higher texture rate (37.80 GTexel/s versus 33.49 GTexel/s), and higher FP32 throughput (907.2 GFLOPS versus 803.7 GFLOPS). Every computational rate metric favors the GTX 460, so its win in the OpenCL workload is not a fluke but a reflection of raw processing capability.

The GTX 670M, despite losing the benchmark, has its own territory: mobility. It is a 75 W part with no external power connectors, formatted as an MXM Module, and its display outputs are "Portable Device Dependent." The GTX 460, by contrast, is a 160 W dual-slot desktop card requiring two 6-pin power connectors and a 450 W suggested PSU. The GTX 670M also doubles the memory capacity (1536 MB versus 768 MB), which can matter for workloads that need larger framebuffers, even if the bandwidth is lower. For a laptop or compact system where power and physical space are constraints, the GTX 670M is the only viable choice in this pairing.

Architecture Differences

Both GPUs trace their lineage to NVIDIA's Fermi architecture, but they are not identical implementations. The GTX 460 uses the GF104 chip and is explicitly labeled as "Fermi," while the GTX 670M uses the GF114 chip and is designated "Fermi 2.0." This revision marks a refinement of the original Fermi design rather than a full architectural overhaul, and it shows in the feature set and power behavior.

The two cards share identical compute resources: 336 shading units, 56 texture mapping units, and 24 ROPs. They also share the same transistor count (1,950 million), die size (332 mm²), and manufacturing process (40 nm at TSMC). The architectural difference is therefore not in the core configuration but in the clock speeds and efficiency targets. The GTX 460 runs its memory at 900 MHz (3.6 Gbps effective), while the GTX 670M runs at 750 MHz (3 Gbps effective). This clock reduction on the mobile part directly drives its lower bandwidth and pixel/texture rates.

The API support is identical: both support DirectX 12 (11_0) and OpenGL 4.6, with no Vulkan support recorded. Neither card lists ray tracing or tensor cores, which is expected for this generation. The key architectural split is thus between a desktop-optimized Fermi design tuned for peak throughput and a mobile-optimized Fermi 2.0 design tuned for power efficiency, with the latter sacrificing some rate metrics to fit a 75 W envelope.

Specification Differences

The recorded data shows several fields where the two cards diverge. The most obvious is memory size: 768 MB on the GTX 460 versus 1536 MB on the GTX 670M. Both use GDDR5 and a 192-bit bus, but the GTX 460's memory clock is 900 MHz (3.6 Gbps effective) versus 750 MHz (3 Gbps effective) on the GTX 670M. This yields bandwidth of 86.40 GB/s versus 72.00 GB/s, respectively.

The power profile differs sharply. The GTX 460 has a TDP of 160 W, uses two 6-pin power connectors, and lists a suggested PSU of 450 W. The GTX 670M has a TDP of 75 W, uses no power connectors, and has no suggested PSU listed. The GTX 460 is dual-slot and measures 210 mm (8.3 inches) in length; the GTX 670M is an MXM Module with no length recorded. The bus interface also differs: PCIe 2.0 x16 on the GTX 460 versus MXM-B (3.0) on the GTX 670M.

Display outputs are another distinguishing field. The GTX 460 offers 2x DVI and 1x mini-HDMI 1.3a, while the GTX 670M's outputs are described as "Portable Device Dependent." The release dates are also distinct: 2010-07-11 for the GTX 460 and 2012-03-21 for the GTX 670M. The GTX 460 has a launch MSRP of 199 USD; the GTX 670M has no launch MSRP recorded.

Head-to-Head Benchmarks

The only recorded head-to-head test is Geekbench OpenCL, and it is a decisive win for the GTX 460. The scores are 7,925 for the GTX 460 and 6,513 for the GTX 670M, a delta of 21.7% in favor of the desktop card. This is not a marginal victory; it is a substantial gap that places the two cards in different performance classes despite their shared core counts.

Breaking down the rate metrics explains why. The GTX 460's pixel rate of 9.450 GPixel/s is 12.9% higher than the GTX 670M's 8.372 GPixel/s. Its texture rate of 37.80 GTexel/s is 12.9% higher than 33.49 GTexel/s. Its FP32 output of 907.2 GFLOPS is 12.9% higher than 803.7 GFLOPS. The consistency of that roughly 13% advantage across all three rate metrics points to a uniform clock speed deficit on the GTX 670M, which runs its cores at lower frequencies to stay within the 75 W power budget.

The memory bandwidth gap is steeper: 86.40 GB/s versus 72.00 GB/s is a 20% difference. This compounds the core clock deficit, making the GTX 460 stronger in both compute-bound and memory-bound operations. The GTX 670M's larger 1536 MB buffer does not compensate in this test, as the OpenCL workload measured here does not appear to be capacity-limited.

The percentile data reinforces the separation. The GTX 460 sits at the 41st percentile among all GPUs, while the GTX 670M sits at the 38th percentile. That three-percentage-point gap is small in percentile terms but reflects the fact that both cards are mid-pack performers in the broader database, with the GTX 460 holding a modest edge over a large field.

The Verdict

The data supports a straightforward recommendation for each use case. If the priority is raw OpenCL compute performance, the NVIDIA GeForce GTX 460 is the clear choice. It delivers a 21.7% higher benchmark score, higher pixel and texture rates, higher FP32 throughput, and higher memory bandwidth. The GTX 460's 41st percentile ranking versus the GTX 670M's 38th percentile confirms that its advantage is real in the broader GPU landscape, not just in this direct comparison.

If the priority is mobility, power efficiency, or memory capacity, the NVIDIA GeForce GTX 670M is the only sensible pick. Its 75 W TDP, lack of power connectors, MXM Module form factor, and portable-device-dependent outputs make it a laptop-oriented part. The GTX 460, at 160 W with two 6-pin connectors and a 450 W suggested PSU, cannot be installed in such an environment. The GTX 670M also offers double the memory (1536 MB versus 768 MB), which may be advantageous for workloads that exceed the smaller framebuffer, even at lower bandwidth.

The benchmark results indicate that the GTX 460 wins the only head-to-head test, but the GTX 670M wins the practical battle for portable systems. A desktop user with power headroom should choose the GTX 460. A laptop user or a system builder constrained to MXM modules has no choice but the GTX 670M. The recorded data does not show any scenario where the GTX 670M outperforms the GTX 460 in compute, so the decision rests entirely on form factor and power requirements, not on performance.