NVIDIA GeForce GTX 670M vs NVIDIA GeForce GTX 970 Comparison

The NVIDIA GeForce GTX 970 and the NVIDIA GeForce GTX 670M represent two distinct eras of mobile and desktop graphics. The GTX 970 is a desktop-focused card built on the Maxwell 2.0 architecture, while the GTX 670M is a notebook-oriented solution based on the older Fermi 2.0 design. The benchmark data shows a decisive performance gap, but the architectural and specification differences explain why these cards are suited for entirely different purposes.

FAQ

Q: How do the two cards compare in raw compute performance?

A: The GTX 970 delivers 3.920 TFLOPS of FP32 performance, which is nearly five times the 803.7 GFLOPS of the GTX 670M. This massive difference in shading throughput is the primary driver of the performance gap observed in benchmarks.

Q: Which card has a higher memory bandwidth?

A: The GTX 970 offers 224.4 GB/s of bandwidth, compared to just 72.00 GB/s on the GTX 670M. This is a 3.1x advantage for the desktop card, stemming from its wider 256-bit bus and faster 7 Gbps effective memory clock.

Q: What is the transistor density difference between the two chips?

A: The GTX 970's GM204 chip packs 5,200 million transistors into a 398 mm² die, resulting in a density of 13.1M transistors per mm². The GTX 670M's GF114 chip has 1,950 million transistors on a 332 mm² die, yielding a density of just 5.9M per mm².

Q: Do both cards support DirectX 12?

A: Yes, both support DirectX 12, but at different feature levels. The GTX 970 supports version 12 (12_1), while the GTX 670M supports version 12 (11_0), meaning the newer card has access to more advanced rendering features.

Q: In the single shared benchmark test, what was the score difference?

A: In the Geekbench OpenCL test, the GTX 970 scored 29,982 points, while the GTX 670M scored 6,513 points. This represents a 360.3% advantage for the GTX 970, making it the clear winner in the only head-to-head comparison available.

Q: How does the GTX 970's average benchmark score compare to its rivals?

A: The GTX 970 has an average benchmark score of 7,157, placing it within 0.9% of the NVIDIA GeForce GTX 750 (7,222) and 0.6% behind the AMD Radeon Vega 8 Mobile (7,203). Its percentile rank among all GPUs is 39.

Architecture Differences

The fundamental architectural gap is vast. The GTX 970 uses the GM204 chip built on the Maxwell 2.0 architecture, fabricated on a 28 nm process at TSMC. In contrast, the GTX 670M uses the GF114 chip with the Fermi 2.0 architecture, also from TSMC but on a larger 40 nm process node. This process shrink is a key reason the GTX 970 can pack 5,200 million transistors into a 398 mm² die, while the GTX 670M manages 1,950 million transistors on a larger 332 mm² die. The resulting transistor density is 13.1M per mm² for the GTX 970 versus 5.9M per mm² for the GTX 670M.

The execution resources differ dramatically. The GTX 970 features 1,664 shading units, 104 texture mapping units, and 56 ROPs. The GTX 670M is equipped with only 336 shading units, 56 TMUs, and 24 ROPs. This translates to pixel rates of 65.97 GPixel/s for the GTX 970 versus 8.372 GPixel/s for the GTX 670M, and texture rates of 122.5 GTexel/s versus 33.49 GTexel/s.

Memory architecture also sets them apart. The GTX 970 uses 4 GB of GDDR5 on a 256-bit bus, while the GTX 670M uses 1536 MB of GDDR5 on a 192-bit bus. The memory clocks are 7 Gbps effective for the GTX 970 and 3 Gbps effective for the GTX 670M, leading to the bandwidth figures of 224.4 GB/s and 72.00 GB/s, respectively. API support is another differentiator: the GTX 970 supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, while the GTX 670M supports DirectX 12 (11_0) and OpenGL 4.6, but has no Vulkan support listed.

Where Each One Wins

The GTX 970 wins decisively in every measurable category where data exists. Its only shared benchmark with the GTX 670M is the Geekbench OpenCL test, where it scores 29,982 versus 6,513, a 360.3% difference. The GTX 970 also wins on compute capability, with its 3.920 TFLOPS FP32 output dwarfing the 803.7 GFLOPS of the GTX 670M. For desktop gaming or workstation use at higher resolutions and detail settings, the GTX 970's larger memory pool, higher bandwidth, and greater shading power make it the only viable option in this comparison.

The GTX 670M, however, holds an advantage in power efficiency and form factor. Its 75 W TDP is half the 148 W of the GTX 970, and its MXM Module slot width is designed for portable devices, whereas the GTX 970 is a dual-slot card with 2x 6-pin power connectors and a suggested 300 W PSU. The GTX 670M has no power connectors and is "Portable Device Dependent" for display outputs, making it the appropriate choice for laptops where space and battery life are critical. There is no benchmark category where the GTX 670M beats the GTX 970.

Specification Differences

The two cards differ in nearly every specification field. The GTX 970 has a base clock of 1050 MHz and a boost clock of 1178 MHz, while the GTX 670M has no base or boost clock listed. Memory size is 4 GB versus 1536 MB, and bus width is 256 bit versus 192 bit. The GTX 970's memory clock is 1753 MHz (7 Gbps effective) versus 750 MHz (3 Gbps effective) for the GTX 670M.

Shading units are 1,664 versus 336, TMUs are 104 versus 56, and ROPs are 56 versus 24. Pixel rate is 65.97 GPixel/s versus 8.372 GPixel/s, and texture rate is 122.5 GTexel/s versus 33.49 GTexel/s. FP32 performance is 3.920 TFLOPS versus 803.7 GFLOPS. TDP is 148 W versus 75 W. Slot width is dual-slot versus MXM Module, and power connectors are 2x 6-pin versus none. The suggested PSU is 300 W for the GTX 970, with no value for the GTX 670M. Bus interfaces are PCIe 3.0 x16 versus MXM-B (3.0). Display outputs are 1x DVI, 1x HDMI 2.0, 3x DisplayPort 1.2 for the GTX 970, versus "Portable Device Dependent" for the GTX 670M. The GTX 970 has physical dimensions of 267 mm length, 111 mm height, and 40 mm width, while the GTX 670M has no dimensions listed. Release dates are 2014-09-18 for the GTX 970 and 2012-03-21 for the GTX 670M. The GTX 970 has a launch MSRP of 329 USD, while the GTX 670M has none.

Head-to-Head Benchmarks

The only direct benchmark comparison available is the Geekbench OpenCL test. In this test, the GTX 970 scores 29,982 points, while the GTX 670M scores 6,513 points. The delta percentage is 360.3%, meaning the GTX 970 is more than four and a half times faster in this compute workload. This single data point encapsulates the generational leap between the two architectures.

Beyond the direct comparison, the GTX 970 has a broader benchmark profile. Its average benchmark score is 7,157, placing it at the 39th percentile of all GPUs. Its nearest rivals include the Intel Iris Pro Graphics P580 (7,170, -0.2% delta), the NVIDIA GeForce GTX 560 SE (7,171, -0.2% delta), the AMD Radeon Vega 8 Mobile (7,203, -0.6% delta), and the NVIDIA GeForce GTX 750 (7,222, -0.9% delta). This suggests the GTX 970 sits in a competitive mid-range desktop segment.

The GTX 670M, on the other hand, has an average benchmark score of 6,513, placing it at the 38th percentile. Its nearest rivals include the NVIDIA GeForce GT 555M (6,493, +0.3% delta), the NVIDIA Quadro M5000M (6,481, +0.5% delta), the AMD Radeon Vega 10 Mobile (6,476, +0.6% delta), and the Intel UHD Graphics P750 (6,554, -0.6% delta). The GTX 670M is slightly above or below these rivals within a narrow 0.6% band, indicating it is a middle-of-the-pack mobile solution.

The Verdict

The data presents a clear hierarchy. The GTX 970 is the superior performer in every shared metric, with a 360.3% lead in the only head-to-head benchmark. Its architecture, memory subsystem, and compute resources are all generations ahead. For anyone building a desktop system that requires high-resolution gaming or compute workloads, the GTX 970 is the only logical choice between these two cards.

The GTX 670M, however, is not without purpose. Its 75 W TDP and MXM form factor make it a suitable component for laptops, where the GTX 970's dual-slot size and 148 W power draw are impractical. The GTX 670M's nearest rivals are all mobile or integrated solutions, confirming its role in the portable segment. Its performance, while far below the GTX 970, is competitive with other mobile GPUs of its era, sitting within 0.6% of the AMD Radeon Vega 10 Mobile and 0.5% of the NVIDIA Quadro M5000M.

The choice depends entirely on the use case. Desktop users with access to a 300 W PSU and space for a 267 mm card should select the GTX 970. Laptop users constrained by power and thermal limits will find the GTX 670M adequate, though benchmark results indicate it is firmly in the mid-range of mobile performance. The GTX 970's launch MSRP of 329 USD reflects its positioning as a mainstream desktop card, while the GTX 670M, with no listed MSRP, was typically bundled into complete mobile systems. There is no scenario in this data where the GTX 670M outperforms the GTX 970 on raw performance, but there are scenarios where only the GTX 670M fits the physical and power constraints of the device.