NVIDIA GeForce GTX 670 vs NVIDIA Quadro K4200 Comparison

The NVIDIA GeForce GTX 670 and NVIDIA Quadro K4200 share the same GK104 Kepler silicon, but they are tuned for different missions. The GTX 670 is a consumer gaming card from 2012, while the K4200 is a professional workstation card from 2014. Despite the two-year gap, their underlying hardware is nearly identical, making their differences in clocks, memory, and power the deciding factors. Benchmark data shows the GTX 670 leads in both available compute tests, but the K4200 counters with double the VRAM and a far lower power draw. This comparison is about whether raw performance or professional feature-set matters more to you.

Head-to-Head Benchmarks

The two cards were tested head-to-head on two Geekbench compute workloads: OpenCL and Vulkan. In both cases, the NVIDIA GeForce GTX 670 wins decisively, with identical margins. In Geekbench OpenCL, the GTX 670 scores 15,341 against the K4200’s 12,313. That is a 24.6% advantage for the GTX 670, a substantial gap that reflects its higher clock speeds. The GTX 670’s base clock is 915 MHz with a boost of 980 MHz, while the K4200 runs at just 771 MHz base and 784 MHz boost. This clock difference directly translates into compute throughput, as the GTX 670 delivers 2.634 TFLOPS of FP32 performance versus the K4200’s 2.107 TFLOPS.

The Vulkan test tells the same story. The GTX 670 scores 15,553, while the K4200 trails at 12,482, again a 24.6% delta. This consistency across APIs suggests the performance gap is purely a function of clock speed and memory bandwidth, not architectural efficiency. The GTX 670’s memory runs at 6 Gbps effective, producing 192.3 GB/s of bandwidth, whereas the K4200’s 5.4 Gbps effective memory yields 172.8 GB/s. That 11% bandwidth advantage, combined with the higher clocks, explains why the GTX 670 pulls ahead so consistently.

It is worth remembering the K4200 has no Geekbench Metal score listed, while the GTX 670 posts 7,424 in that test. The two cards do share the same 1,344 shading units, 112 texture mapping units, and 32 ROPs, so the raw compute hardware is identical. Every performance difference in these benchmarks comes down to operating frequencies. The GTX 670’s pixel rate of 27.44 GPixel/s and texture rate of 109.8 GTexel/s are both higher than the K4200’s 21.95 GPixel/s and 87.81 GTexel/s, reinforcing the clock-driven nature of the gap.

FAQ

Q: Which card has more VRAM?

A: The NVIDIA Quadro K4200 has 4 GB of GDDR5 memory, while the NVIDIA GeForce GTX 670 has 2 GB. Both use a 256-bit bus, but the GTX 670 compensates with higher effective memory speed (6 Gbps vs 5.4 Gbps), giving it more bandwidth despite half the capacity.

Q: Are these cards based on the same chip?

A: Yes. Both use the GK104 chip from NVIDIA’s Kepler architecture, fabricated by TSMC on a 28 nm process. They share identical transistor counts of 3,540 million, a die size of 294 mm², and a transistor density of 12.0M per mm².

Q: What is the average benchmark score for each card?

A: The GTX 670 has an average benchmark score of 12,773, while the K4200 scores 12,398. This puts the GTX 670 about 2.9% ahead of the K4200 in the nearest rival comparison, with the K4200’s own rival list showing a -2.9% delta against the GTX 670.

Q: Does the Quadro K4200 support the same APIs as the GTX 670?

A: Yes, both cards support DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The API feature sets are identical, meaning software compatibility is not a differentiator between them.

Q: Which card requires less power?

A: The Quadro K4200 has a 108 W TDP, compared to the GTX 670’s 170 W. The K4200 needs only a single 6-pin power connector and a 300 W suggested PSU, whereas the GTX 670 requires two 6-pin connectors and a 450 W suggested PSU.

Q: How do these cards rank against other GPUs?

A: Both sit at the 52nd percentile among all GPUs. The GTX 670’s nearest rivals include the GeForce GTX 590 (12830, -0.4%), Radeon Pro 455 (12831, -0.4%), and RX 7600M XT (12710, +0.5%). The K4200’s rivals include the Tesla K20Xm (12625, -1.8%), RX 7600M XT (12710, -2.5%), and GTX 670 (12773, -2.9%).

The Verdict

The data points to a clear performance winner: the NVIDIA GeForce GTX 670. It beats the K4200 by 24.6% in both OpenCL and Vulkan benchmarks, and its average score of 12,773 exceeds the K4200’s 12,398. If your priority is raw compute throughput, the GTX 670 is the better choice. However, the K4200 is not without merit. Its 4 GB VRAM doubles the GTX 670’s 2 GB, which matters for large datasets or high-resolution textures that exceed 2 GB. The K4200 also draws 62 W less power (108 W vs 170 W), allowing for a single-slot cooler and a 300 W PSU instead of 450 W. For a workstation that runs 24/7, that power saving is significant.

The GTX 670 wins on every benchmark metric available, but the K4200 wins on capacity and efficiency. If you need maximum compute performance per clock, the GTX 670 is the pick. If you need more memory and lower power in a compact single-slot form factor, the K4200 is the practical choice. Neither card is current — both are end-of-life — but the GTX 670’s higher clocks make it the faster of the two in every tested workload.

Specification Differences

The two cards diverge on several key specifications. Clock speeds are the biggest gap: the GTX 670 runs at 915 MHz base and 980 MHz boost, while the K4200 runs at 771 MHz base and 784 MHz boost. Memory speed also differs, with the GTX 670 at 1502 MHz (6 Gbps effective) versus the K4200’s 1350 MHz (5.4 Gbps effective). Memory capacity is 2 GB on the GTX 670 and 4 GB on the K4200, though both use GDDR5 on a 256-bit bus. Bandwidth favors the GTX 670 at 192.3 GB/s versus 172.8 GB/s.

The pixel rate is 27.44 GPixel/s on the GTX 670 versus 21.95 GPixel/s on the K4200. Texture rate is 109.8 GTexel/s versus 87.81 GTexel/s. FP32 compute is 2.634 TFLOPS versus 2.107 TFLOPS. Power consumption is starkly different: 170 W TDP for the GTX 670 with dual-slot cooling and 2x 6-pin connectors, versus 108 W TDP for the K4200 with single-slot cooling and 1x 6-pin. The suggested PSU is 450 W for the GTX 670 and 300 W for the K4200. Bus interface differs too: PCIe 3.0 x16 on the GTX 670, PCIe 2.0 x16 on the K4200. Display outputs are 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2 on the GTX 670, while the K4200 has 1x DVI and 2x DisplayPort 1.2.

Architecture Differences

Both cards use the exact same GK104 chip from NVIDIA’s Kepler architecture, built by TSMC on a 28 nm process. They share identical transistor counts (3,540 million), die size (294 mm²), and transistor density (12.0M / mm²). The shading units, TMUs, and ROPs are identical at 1,344, 112, and 32 respectively. Neither card has ray tracing or tensor cores. The architecture itself is the same — any differences in performance come from clock speeds and memory configuration, not from core count or design.

The GTX 670 is part of the GeForce 600 generation, while the K4200 belongs to the Quadro Kepler (Kx200) generation. The GTX 670’s predecessor is GeForce 500 and its successor is GeForce 700. The K4200’s predecessor is Quadro Fermi and its successor is Quadro Maxwell. The GTX 670 was released on May 9, 2012, and the K4200 on July 21, 2014. Both are end-of-life. The GTX 670 had a launch MSRP of 399 USD; the K4200 has no listed launch MSRP.

Where Each One Wins

The GTX 670 wins in every benchmark category where both are tested. It scores 15,341 in OpenCL and 15,553 in Vulkan, beating the K4200 by 24.6% in each. Its higher memory bandwidth (192.3 GB/s) and faster clocks make it the choice for compute-heavy workloads that are bound by throughput and latency. If you are running OpenCL or Vulkan-based applications that are not limited by VRAM capacity, the GTX 670 will deliver noticeably better performance.

The K4200 wins on the specifications that matter outside raw compute. Its 4 GB VRAM is double the GTX 670’s 2 GB, making it the better option for workloads that require large framebuffers or datasets exceeding 2 GB. Its 108 W TDP is 62 W lower, which means it runs cooler and quieter in a single-slot form factor. It also requires only a 300 W PSU versus 450 W, making it easier to integrate into compact or legacy systems. The K4200’s PCIe 2.0 x16 interface is older than the GTX 670’s PCIe 3.0 x16, but that is unlikely to matter in most real-world scenarios where both cards are bottlenecked by their own compute limits.

For a gaming or general compute scenario, the GTX 670’s higher clocks win outright. For a professional workstation where memory capacity, power efficiency, and single-slot cooling are priorities, the K4200 is the more sensible choice, even at the cost of 24.6% compute performance. The data is unambiguous: the GTX 670 is faster, the K4200 is more efficient and has more memory. Your workload determines which metric you value more.