NVIDIA GeForce GTX 675M vs NVIDIA GRID K2 Comparison

Head-to-Head Benchmarks

The data records a single head-to-head benchmark between the NVIDIA GRID K2 and the NVIDIA GeForce GTX 675M: Geekbench OpenCL. The GRID K2 dominates this test decisively, scoring 10,602 points against the GTX 675M's 6,946 points. This translates to a 52.6% advantage for the GRID K2, a margin that places the two cards in entirely different performance tiers despite both being NVIDIA products from the same era.

The GRID K2's OpenCL score is not just a win, it is a near-doubling of the GTX 675M's output. In practical terms, compute-heavy workloads such as OpenCL-accelerated rendering, physics simulation, or data-parallel processing would complete in roughly two-thirds of the time on the GRID K2. The GTX 675M cannot match this throughput, and the benchmark data shows no other test where it closes the gap.

Context from the database's nearest rival data reinforces the scale of this victory. The GRID K2's average benchmark score is 8,080, placing it at the 42nd percentile of all GPUs. Its closest rivals are the GeForce GTX 650 Ti Boost (8,067, within 0.2%), the GeForce 945M (8,099, within 0.2%), and the GeForce GTX 650 Ti (8,053, within 0.3%). The GRID K2 sits in a tight cluster of mid-range desktop parts. In contrast, the GTX 675M's average score is 6,946, at the 39th percentile, with rivals like the AMD Radeon R5 M240 (6,975, within 0.4%) and the NVIDIA T600 (7,035, within 1.3%). The GTX 675M is grouped with lower-tier mobile and entry-level parts, roughly 14% behind the GRID K2's average score.

The only benchmark where both cards share a test is OpenCL, and the GRID K2 wins it outright. The Geekbench Metal score of 5,557 for the GRID K2 is not comparable to the GTX 675M because the latter has no Metal result recorded. Thus, the head-to-head record stands at 1 win for the GRID K2 and 0 for the GTX 675M.

Where Each One Wins

The GRID K2 wins the only directly comparable workload category: OpenCL compute. This is a significant victory because OpenCL is a cross-platform, heterogeneous computing standard used for general-purpose GPU tasks. The GRID K2's 2.289 TFLOPS of FP32 throughput and 160.0 GB/s of memory bandwidth make it a substantially more capable compute engine than the GTX 675M's 952.3 GFLOPS and 96.00 GB/s. For workloads that scale with raw floating-point operations and memory bandwidth, the GRID K2 is the clear choice.

The GTX 675M, however, has its own domain where it holds an advantage, though not in raw performance. It is a mobile GPU with a 100 W TDP, while the GRID K2 draws 225 W. The GTX 675M is designed as an MXM module, meaning it is intended for notebooks and portable workstations. It has no external power connectors, relying on the host laptop's power delivery, and its display outputs are "Portable Device Dependent," indicating it drives laptop panels. The GRID K2, by contrast, has no display outputs at all; it is a server-side compute card. So the GTX 675M wins in the category of physical integration into mobile systems, a use case the GRID K2 cannot serve.

For desktop or server environments where compute density matters, the GRID K2 wins decisively. For laptop gaming or mobile workstation use, the GTX 675M is the only viable option. The data does not show any workload where the GTX 675M outperforms the GRID K2 in raw speed, but it fills a form-factor niche that the GRID K2 cannot.

Architecture Differences

The two GPUs come from different NVIDIA architectures and process nodes. The GRID K2 uses the GK104 chip built on the Kepler architecture, fabricated on a 28 nm process at TSMC. The GTX 675M uses the GF114 chip on the older Fermi 2.0 architecture, also from TSMC but on a larger 40 nm node. This generational leap is visible in the transistor density: the GRID K2 packs 3,540 million transistors into a 294 mm² die, yielding 12.0 million transistors per square millimeter. The GTX 675M has 1,950 million transistors spread across a larger 332 mm² die, for a density of just 5.9 million per square millimeter. The Kepler design achieves more than double the transistor density, which explains its superior compute throughput despite a similar physical footprint.

The GRID K2 features 1,536 shading units, 128 texture mapping units (TMUs), and 32 render output units (ROPs). The GTX 675M offers only 384 shading units, 64 TMUs, and 32 ROPs. The GRID K2 has four times the shader count and double the TMUs, which directly drives its fill rate advantages: 23.84 GPixel/s pixel rate and 95.36 GTexel/s texture rate versus the GTX 675M's 9.92 GPixel/s and 39.68 GTexel/s. Both have 32 ROPs, but the GRID K2's higher clocked memory and wider internal execution make its pixel throughput more than double.

Memory configurations also differ significantly. The GRID K2 has 4 GB of GDDR5 on a 256-bit bus, running at 1250 MHz (5 Gbps effective), yielding 160.0 GB/s of bandwidth. The GTX 675M has 2 GB of GDDR5 on the same 256-bit bus, but at 750 MHz (3 Gbps effective), resulting in only 96.00 GB/s. The GRID K2 offers double the capacity and 66.7% more bandwidth.

Both cards support DirectX 12 (11_0) and OpenGL 4.6. The GRID K2 adds Vulkan 1.2.175 support, while the GTX 675M has no Vulkan support recorded. This is a notable API difference: the GRID K2 can run modern Vulkan-based titles or compute applications, whereas the GTX 675M cannot. The GRID K2 also has a higher TDP at 225 W versus 100 W, reflecting its server-oriented design with a dual-slot cooler and 1x 6-pin plus 1x 8-pin power connectors, whereas the GTX 675M is an MXM module with no external power connectors.

FAQ

Q: Which GPU has the higher OpenCL benchmark score?

A: The NVIDIA GRID K2 scores 10,602 in Geekbench OpenCL, while the NVIDIA GeForce GTX 675M scores 6,946. The GRID K2 leads by 52.6%.

Q: Do both GPUs support the same APIs?

A: Both support DirectX 12 (11_0) and OpenGL 4.6. The GRID K2 additionally supports Vulkan 1.2.175, while the GTX 675M has no Vulkan support recorded.

Q: What are the memory sizes and bandwidths?

A: The GRID K2 has 4 GB of GDDR5 with 160.0 GB/s bandwidth. The GTX 675M has 2 GB of GDDR5 with 96.00 GB/s bandwidth. Both use a 256-bit bus.

Q: Which card is more power-efficient?

A: The GTX 675M has a 100 W TDP, while the GRID K2 has a 225 W TDP. The GTX 675M draws less power, making it suitable for mobile systems, but it also delivers far lower compute performance.

Q: Are there any display outputs on these cards?

A: The GRID K2 has no display outputs; it is designed for compute workloads in servers. The GTX 675M has display outputs that are "Portable Device Dependent," meaning it drives laptop displays.

Q: What is the transistor count difference?

A: The GRID K2 has 3,540 million transistors, while the GTX 675M has 1,950 million. The GRID K2 uses a 28 nm process, while the GTX 675M uses a 40 nm process.

The Verdict

The data is unambiguous: the NVIDIA GRID K2 is the superior GPU in raw performance. It wins the only head-to-head benchmark by 52.6%, has double the memory capacity and significantly more bandwidth, and offers a 2.4x advantage in FP32 compute (2.289 TFLOPS versus 952.3 GFLOPS). Its architecture is newer, denser, and more capable, with Vulkan support that the GTX 675M lacks. Any workload that can use a discrete GPU, whether rendering, simulation, or compute, will execute faster on the GRID K2.

The GTX 675M's only advantages are its lower power draw (100 W versus 225 W) and its mobile form factor. It is an MXM module designed for laptops, with no external power connectors and portable-device-dependent outputs. The GRID K2 cannot be installed in a laptop, and it has no display outputs whatsoever. So for a mobile workstation or gaming notebook, the GTX 675M is the only choice between these two.

However, the benchmark data does not show a single test where the GTX 675M beats the GRID K2. The GTX 675M's average benchmark score of 6,946 is roughly 14% lower than the GRID K2's 8,080. Its nearest rivals are entry-level mobile parts and low-end desktop cards, while the GRID K2 sits with mid-range desktop GPUs. The GTX 675M is an older Fermi-generation part, and its 40 nm process and 384 shading units cannot compete with the Kepler-based GRID K2's 1,536 shading units.

Users should pick the GRID K2 if they need maximum compute throughput in a server or desktop chassis and can supply 225 W of power. Users should pick the GTX 675M only if they require a GPU that fits into a laptop's MXM slot and operates within a 100 W power envelope. For all performance-sensitive tasks, the GRID K2 is the clear winner.

Specification Differences

| Specification | NVIDIA GRID K2 | NVIDIA GeForce GTX 675M |

|---------------|---------------|------------------------|

| Architecture | Kepler | Fermi 2.0 |

| Process Node | 28 nm | 40 nm |

| Transistors | 3,540 million | 1,950 million |

| Die Size | 294 mm² | 332 mm² |

| Transistor Density | 12.0M / mm² | 5.9M / mm² |

| Memory Size | 4 GB | 2 GB |

| Memory Clock | 1250 MHz (5 Gbps effective) | 750 MHz (3 Gbps effective) |

| Memory Bandwidth | 160.0 GB/s | 96.00 GB/s |

| Shading Units | 1536 | 384 |

| TMUs | 128 | 64 |

| ROPs | 32 | 32 |

| Pixel Rate | 23.84 GPixel/s | 9.920 GPixel/s |

| Texture Rate | 95.36 GTexel/s | 39.68 GTexel/s |

| FP32 Performance | 2.289 TFLOPS | 952.3 GFLOPS |

| TDP | 225 W | 100 W |

| Slot Width | Dual-slot | MXM Module |

| Power Connectors | 1x 6-pin + 1x 8-pin | None |

| Suggested PSU | 550 W | None |

| Bus Interface | PCIe 3.0 x16 | MXM-B (3.0) |

| Display Outputs | No outputs | Portable Device Dependent |

| Vulkan Support | 1.2.175 | None |

| Release Date | 2013-05-10 | 2012-03-21 |

| Launch MSRP | 5,199 USD | None |