NVIDIA GeForce GTX 670M vs NVIDIA T600 Comparison

The NVIDIA T600 and NVIDIA GeForce GTX 670M represent two very different eras of mobile and workstation graphics, separated by nearly a decade of architectural evolution. The data shows a clear generational gap, with the T600’s Turing architecture delivering a massive performance advantage over the older Fermi 2.0-based GTX 670M, despite the latter having a wider memory bus and more texture units. This analysis explores the benchmark results, architectural shifts, and specification differences that define this matchup.

FAQ

Q: How much faster is the NVIDIA T600 than the GTX 670M in the only shared benchmark?

A: In the Geekbench OpenCL test, the T600 scores 27,875 compared to the GTX 670M’s 6,513. This translates to a delta of 328%, meaning the T600 is over four times faster in this compute workload.

Q: What is the performance percentile ranking for each GPU?

A: The T600 sits at the 39th percentile of all GPUs, while the GTX 670M is at the 38th percentile. Despite the massive OpenCL score gap, their overall percentile rankings are nearly identical, suggesting the GTX 670M’s average score is pulled by other factors or that the T600’s dataset includes more diverse tests.

Q: Which GPU has a higher average benchmark score?

A: The T600 has an average benchmark score of 7,035, while the GTX 670M averages 6,513. This difference is about 8% in favor of the T600, though the single head-to-head test shows a far larger disparity.

Q: How do the nearest rivals compare to the T600?

A: The T600’s closest rival is the NVIDIA GeForce GTX 680M, which scores 7,023, just 0.2% lower. The AMD Radeon R5 M240 is 0.9% behind, and the NVIDIA GeForce GTX 675M trails by 1.3%. The GTX 970 is 1.7% ahead of the T600.

Q: What are the closest competitors to the GTX 670M?

A: The GTX 670M’s nearest rival is the NVIDIA GeForce GT 555M, which scores 6,493, only 0.3% lower. The NVIDIA Quadro M5000M is 0.5% behind, and the AMD Radeon Vega 10 Mobile is 0.6% behind. The Intel UHD Graphics P750 is 0.6% ahead.

Q: What is the manufacturing process difference between the two?

A: The T600 uses a 12 nm process at TSMC, while the GTX 670M uses a 40 nm process. This is a substantial difference in transistor density, with the T600 packing 23.5 million transistors per square millimeter versus 5.9 million for the GTX 670M.

Architecture Differences

The architectural gap between these two GPUs is profound. The T600 is built on the Turing architecture, specifically the TU117 chip, which represents a modern design optimized for efficiency and compute performance. In contrast, the GTX 670M uses the GF114 chip based on the Fermi 2.0 architecture, a design that dates back to 2012. This generational leap is evident in nearly every aspect of their design.

The manufacturing process tells a stark story. The T600 is fabricated on a 12 nm process at TSMC, allowing for 4,700 million transistors on a 200 mm² die. This yields a transistor density of 23.5 million per square millimeter. The GTX 670M, on a 40 nm process, packs 1,950 million transistors onto a much larger 332 mm² die, resulting in a density of just 5.9 million per square millimeter. The T600 achieves more than double the transistor count on a smaller physical footprint, a signal of process node advancement.

Clock speeds also differ significantly. The T600 has a base clock of 735 MHz and a boost clock of 1,335 MHz, while the GTX 670M has no listed base or boost clocks, only a memory clock of 750 MHz. The T600’s memory runs at 1,250 MHz with 10 Gbps effective speed, whereas the GTX 670M’s memory operates at 750 MHz with 3 Gbps effective. This directly impacts bandwidth, though the GTX 670M’s wider 192-bit bus partially compensates.

Feature support diverges sharply. The T600 supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The GTX 670M only supports DirectX 12 (11_0) and OpenGL 4.6, with no Vulkan support listed. The T600 also offers 4x mini-DisplayPort 1.4a outputs, while the GTX 670M’s display outputs are described as "Portable Device Dependent," reflecting its mobile MXM form factor.

Head-to-Head Benchmarks

The only direct benchmark comparison available is Geekbench OpenCL, and the results are lopsided. The T600 scores 27,875, while the GTX 670M scores 6,513. This is a 328% delta in favor of the T600, meaning the T600 delivers more than four times the OpenCL compute performance. This is not a marginal improvement; it is a generational obliteration.

This OpenCL result aligns with the theoretical compute specifications. The T600’s FP32 performance is 1.709 TFLOPS, while the GTX 670M is rated at 803.7 GFLOPS. The T600 is more than double the raw FP32 throughput. The T600 also has higher pixel and texture rates: 42.72 GPixel/s versus 8.372 GPixel/s, and 53.40 GTexel/s versus 33.49 GTexel/s. The pixel rate advantage is particularly stark, reflecting the T600’s modern ROP design.

The T600 wins the sole head-to-head benchmark, giving it a 1-0 record in wins. However, the broader benchmark suite shows a more nuanced picture. The T600 has a passmark G3D score of 6,479, which is its strongest legacy DirectX result, while its passmark DirectX 9 score is 114. The GTX 670M lacks these data points, making a full comparison impossible. Still, the OpenCL result alone is decisive for compute-oriented workloads.

Specification Differences

The specification sheets reveal fundamental differences in memory configuration, power, and interface.

Memory: The T600 has 4 GB of GDDR6 memory on a 128-bit bus, yielding 160.0 GB/s bandwidth. The GTX 670M has 1,536 MB (1.5 GB) of GDDR5 on a 192-bit bus, with 72.00 GB/s bandwidth. The T600 has more capacity and more than double the bandwidth, despite the narrower bus, thanks to faster GDDR6 memory.

Compute Units: The T600 has 640 shading units, 40 TMUs, and 32 ROPs. The GTX 670M has 336 shading units, 56 TMUs, and 24 ROPs. The T600 has nearly double the shading units, but the GTX 670M has more TMUs, which could benefit texture-heavy workloads if they were not bottlenecked elsewhere.

Power and Physical: The T600 has a TDP of 40 W, while the GTX 670M draws 75 W. The T600 is single-slot with no power connectors and a suggested PSU of 200 W. The GTX 670M is an MXM module with no power connectors and no suggested PSU listed. The T600 is far more power-efficient, delivering higher performance at nearly half the power draw.

Interface: The T600 uses PCIe 3.0 x16, while the GTX 670M uses MXM-B (3.0). This reflects their different target platforms: the T600 is a workstation card for standard PCIe slots, while the GTX 670M is a mobile module.

Release and Status: The T600 was released in April 2021 and is end-of-life, with a predecessor of Quadro Volta and successor of Workstation Ampere. The GTX 670M was released in March 2012, also end-of-life, with a predecessor of GeForce 500M and successor of GeForce 700M.

Where Each One Wins

The T600 wins decisively in compute workloads, as evidenced by the Geekbench OpenCL result. Its 328% lead over the GTX 670M suggests it is far better suited for tasks that leverage GPGPU compute, such as rendering, scientific simulation, or machine learning inference. The T600’s higher FP32 throughput and modern architecture make it the clear choice for any compute-intensive application.

The T600 also wins in memory capacity and bandwidth. With 4 GB versus 1.5 GB, it can handle larger datasets and textures. The 160.0 GB/s bandwidth is more than double the GTX 670M’s 72.00 GB/s, which is critical for memory-bound workloads. The T600’s support for Vulkan 1.4 also gives it an advantage in modern games and applications that use this API, whereas the GTX 670M has no Vulkan support.

The GTX 670M’s potential wins are harder to identify from the data. It has more TMUs (56 versus 40), which theoretically could improve texture filtering rates, but its texture rate is actually lower (33.49 GTexel/s versus 53.40 GTexel/s) due to lower clocks. Its wider 192-bit memory bus is a theoretical advantage, but the slower GDDR5 memory negates it. The GTX 670M’s average benchmark score is lower, and it has no other benchmark wins. The data does not support any meaningful advantage for the GTX 670M.

The Verdict

The data points to a single conclusion: the NVIDIA T600 is the superior GPU in every measurable way. Its 328% lead in Geekbench OpenCL is the headline result, but the broader picture confirms this. The T600 has higher FP32 performance (1.709 TFLOPS versus 803.7 GFLOPS), higher pixel rate (42.72 GPixel/s versus 8.372 GPixel/s), higher texture rate (53.40 GTexel/s versus 33.49 GTexel/s), more memory (4 GB versus 1.5 GB), and higher bandwidth (160.0 GB/s versus 72.00 GB/s). It also does all this at a lower TDP (40 W versus 75 W).

The GTX 670M is a relic of the Fermi era, and its only theoretical advantages are more TMUs and a wider memory bus, both of which fail to translate into real-world performance. Its average benchmark score of 6,513 is below the T600’s 7,035, and its percentile ranking is nearly identical despite the massive OpenCL gap.

For users choosing between these two, the T600 is the obvious pick for workstation tasks, compute workloads, or any application requiring modern API support. The GTX 670M would only be relevant in legacy systems where its MXM form factor is required, and even then, its performance is severely limited. The verdict is clear: the T600 wins outright, and the only question is why the GTX 670M is still being compared to it. The data suggests the T600 is not just better; it is in a different class entirely.