GPU Comparison

The NVIDIA GeForce GT 640 and NVIDIA Quadro P600 are both end-of-life, single-slot offerings from NVIDIA, but they were designed for different eras and purposes. The GT 640 is a 2012 Kepler-era card aimed at the entry-level desktop market, while the P600 is a 2017 Pascal-era professional workstation card. The benchmark data reveals a stark generational gap, with the P600 dominating in compute and API-specific workloads, despite both cards sharing a 384-shading-unit count and 2 GB of memory.

Head-to-Head Benchmarks

The data shows a decisive victory for the Quadro P600 in the two head-to-head tests available. In Geekbench OpenCL, the P600 scores 11,181, while the GT 640 manages only 3,736. This represents a 66.6% delta, meaning the P600 is roughly three times faster in this compute-oriented workload. The Geekbench Vulkan test tells a similar story: the P600 scores 9,755 against the GT 640's 3,809, a 61% delta. These are not marginal wins; they are overwhelming generational leaps.

Looking at the broader benchmark suite, the P600 also has a much wider range of data. Its Passmark G3D score is 3,317, and its Passmark GPU Compute score is 1,415. While the GT 640 does not have these specific Passmark results in the data, its average benchmark score of 3,210 is actually higher than the P600's average of 2,923. This is a crucial nuance: the P600 wins the raw compute tests, but the GT 640 holds a higher average score across all recorded benchmarks. The discrepancy stems from the P600's extremely low Passmark DirectX scores (14 in DX10, 24 in DX11, 14 in DX12, and 56 in DX9), which drag down its average considerably. The GT 640, with only Geekbench scores (Metal, OpenCL, Vulkan) in its record, avoids this penalty.

The percentile rankings reflect this mixed picture. The GT 640 sits at the 20th percentile of all GPUs, while the P600 sits at the 19th percentile. Despite the P600's massive wins in OpenCL and Vulkan, its overall standing relative to the entire GPU universe is nearly identical to the older card. This suggests that while the P600 is a superior compute engine, its DirectX performance in the Passmark suite is anomalously poor, pulling its percentile down to match the GT 640's level.

Where Each One Wins

The Quadro P600 is the clear winner in raw compute throughput and modern API support. Its Geekbench OpenCL score of 11,181 is not just ahead of the GT 640; it is in a different performance class. Similarly, its Vulkan score of 9,755 shows it can handle modern, low-level graphics APIs with far greater efficiency than the GT 640's 3,809. For any task that leverages OpenCL for general-purpose GPU computing, or Vulkan for modern game engines or professional visualization, the P600 is the only viable choice between the two.

The GT 640's only "win" is its higher average benchmark score of 3,210 versus the P600's 2,923. This is a statistical artifact rather than a real performance advantage. The GT 640 also has a Geekbench Metal score of 2,086, a test that the P600 does not appear in the data for, which contributes to its higher average. In practical terms, the GT 640 does not win any compute or graphics workload in the head-to-head tests. However, its higher percentile rank (20 vs 19) and higher average score suggest that its performance is more consistent across the limited set of tests it was subjected to, whereas the P600's results are highly variable depending on the API.

For a user on a legacy system that only needs basic display output and runs older software that relies on DirectX 9 or 10, the GT 640's Passmark scores (if they existed) would likely be more competitive, but the data does not provide them. The P600's Passmark DirectX scores are catastrophically low (single digits to mid-50s), which indicates it is a poor choice for legacy DirectX gaming or applications. The GT 640, with its Kepler architecture and DirectX 12 (11_0) support, is not tested in those specific Passmark benchmarks, leaving its legacy performance an open question.

Architecture Differences

The two cards represent completely different architectural generations. The GT 640 uses the GK107 chip built on a 28 nm process at TSMC, featuring 1,270 million transistors on a 118 mm² die. The P600 uses the GP107 chip built on a 14 nm process at Samsung, packing 3,300 million transistors into a 132 mm² die. This is a massive difference in transistor density: the GT 640 has 10.8M transistors per mm², while the P600 has 25.0M per mm². The newer, denser process allows the P600 to deliver far more compute performance per watt and per square millimeter.

The memory subsystems differ significantly. The GT 640 uses 2 GB of DDR3 on a 128-bit bus, yielding 28.51 GB/s of bandwidth. The P600 uses 2 GB of GDDR5 on the same 128-bit bus, but with 64.13 GB/s of bandwidth, more than double. The memory clock speeds reflect this: the GT 640 runs at 891 MHz (1782 Mbps effective), while the P600 runs at 1002 MHz (4 Gbps effective). This bandwidth advantage is critical for compute workloads and explains a large part of the P600's OpenCL and Vulkan dominance.

Core configuration shows a split. Both cards have 384 shading units and 16 ROPs. However, the GT 640 has 32 TMUs, while the P600 has 24 TMUs. This is a rare case where the older card has more texture units. Despite having more TMUs, the GT 640's texture rate is only 28.86 GTexel/s, compared to the P600's 37.37 GTexel/s. The P600's higher clock speeds (1329 MHz base, 1557 MHz boost) compensate for its fewer TMUs. The pixel rate tells a similar story: the P600 hits 24.91 GPixel/s versus the GT 640's 7.216 GPixel/s, a massive gap driven by the clock speed difference. FP32 performance is 1,195.8 GFLOPS for the P600 versus 692.7 GFLOPS for the GT 640, again a near-doubling of compute throughput.

FAQ

Q: Is the Quadro P600 faster than the GT 640 in all benchmarks?

A: No. In the head-to-head Geekbench tests, the P600 wins decisively (11,181 vs 3,736 in OpenCL; 9,755 vs 3,809 in Vulkan). However, the GT 640 has a higher average benchmark score overall (3,210 vs 2,923) and a slightly higher percentile rank (20th vs 19th), driven by the P600's very low Passmark DirectX scores.

Q: What is the biggest performance gap between the two cards?

A: The largest delta is in Geekbench OpenCL, where the P600 is 66.6% ahead of the GT 640. The Vulkan gap is also large at 61%. These are the only two head-to-head tests recorded.

Q: Do both cards support the same modern APIs?

A: No. The P600 supports DirectX 12 (12_1) and Vulkan 1.4, while the GT 640 only supports DirectX 12 (11_0) and Vulkan 1.2.175. Both support OpenGL 4.6.

Q: Why does the P600 have a lower average benchmark score if it wins the head-to-head tests?

A: The P600 has additional Passmark scores (DirectX 9/10/11/12, G2D, G3D, GPU Compute) that are included in its average. Its DirectX scores are extremely low (e.g., 14 in DX10 and DX12, 24 in DX11), which drags its average down to 2,923. The GT 640 only has Geekbench scores in its record, keeping its average at 3,210.

Q: Which card has more texture units?

A: The GT 640 has 32 TMUs, while the P600 has 24 TMUs. However, the P600's higher clock speeds (1557 MHz boost vs none listed for the GT 640) give it a higher texture rate of 37.37 GTexel/s versus 28.86 GTexel/s.

Q: What are the power requirements for each card?

A: The P600 has a 40 W TDP and a suggested PSU of 200 W. The GT 640 has a 65 W TDP and a suggested PSU of 250 W. Neither card requires external power connectors.

The Verdict

The data is unambiguous for compute-heavy workflows: the Quadro P600 is the superior card. Its OpenCL and Vulkan scores are roughly three times and two and a half times higher, respectively, than the GT 640's. Its FP32 throughput (1,195.8 GFLOPS) and memory bandwidth (64.13 GB/s) are nearly double the GT 640's figures. If you are running GPU-accelerated applications that utilize OpenCL, or modern software built on Vulkan, the P600 is the only sensible choice. The P600 also has a lower TDP (40 W vs 65 W) and a lower suggested PSU (200 W vs 250 W), making it more efficient despite its higher performance.

The GT 640's only statistical advantage is its higher average benchmark score (3,210 vs 2,923) and its slightly better percentile rank. This is a misleading metric, as it is skewed by the P600's anomalous Passmark DirectX results. If you are dealing with legacy DirectX 9/10/11 applications, the P600's Passmark scores (56, 14, 24) are so low that the GT 640, despite lacking those specific scores, would likely be a better choice for older software. However, the GT 640 is a 2012 card with a 28 nm process, and it lacks the P600's modern architectural features. For any modern professional workload, the P600 wins outright. For anyone building a retro system or needing basic 2D output, the GT 640's higher average score suggests it may be more consistent, but the P600's raw power in modern APIs makes it the better investment for future-proofing.

Specification Differences

| Specification | NVIDIA GeForce GT 640 | NVIDIA Quadro P600 |

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

| Chip | GK107 | GP107 |

| Architecture | Kepler | Pascal |

| Generation | GeForce 600 | Quadro Pascal (Px000) |

| Process Node | 28 nm | 14 nm |

| Foundry | TSMC | Samsung |

| Transistors | 1,270 million | 3,300 million |

| Die Size | 118 mm² | 132 mm² |

| Transistor Density | 10.8M / mm² | 25.0M / mm² |

| Base Clock | None listed | 1329 MHz |

| Boost Clock | None listed | 1557 MHz |

| Memory Clock | 891 MHz / 1782 Mbps effective | 1002 MHz / 4 Gbps effective |

| Memory Type | DDR3 | GDDR5 |

| Memory Bandwidth | 28.51 GB/s | 64.13 GB/s |

| TMUs | 32 | 24 |

| Pixel Rate | 7.216 GPixel/s | 24.91 GPixel/s |

| Texture Rate | 28.86 GTexel/s | 37.37 GTexel/s |

| FP32 | 692.7 GFLOPS | 1,195.8 GFLOPS |

| FP16 | None listed | 18.68 GFLOPS (1:64) |

| TDP | 65 W | 40 W |

| Suggested PSU | 250 W | 200 W |

| Display Outputs | 1x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 | 4x mini-DisplayPort 1.4a |

| DirectX Support | 12 (11_0) | 12 (12_1) |

| Vulkan Support | 1.2.175 | 1.4 |

| Dimensions | 145 mm (5.7 inches) | 150 mm (5.9 inches), 69 mm (2.7 inches) height |

| Launch MSRP | 99 USD | None listed |