NVIDIA GeForce GT 640 vs NVIDIA Quadro 2000D Comparison

# NVIDIA Quadro 2000D vs NVIDIA GeForce GT 640: Database Analysis

The NVIDIA Quadro 2000D and the NVIDIA GeForce GT 640 occupy similar positions in their respective lineups: both are entry-level parts built on the same generation of NVIDIA architecture, but aimed at different markets. The Quadro 2000D is the professional workstation card, while the GT 640 is the consumer desktop part. The database shows that in OpenCL compute workloads, the Quadro 2000D edges out the GT 640 by 5.2%, a margin that is modest but consistent with its higher positioning in the product stack. The following analysis breaks down the architectural differences, the benchmark results, and the practical implications for each card.

FAQ

Q: What architectures do the NVIDIA Quadro 2000D and GeForce GT 640 use?

A: The Quadro 2000D is built on the Fermi architecture with the GF106 chip and is manufactured on a 40 nm process at TSMC. The GeForce GT 640 uses the Kepler architecture with the GK107 chip, also made by TSMC, but on a 28 nm node.

Q: How much memory do these cards have?

A: The Quadro 2000D ships with 1024 MB of GDDR5 memory across a 128 bit bus, providing 41.60 GB/s of bandwidth. The GT 640 has 2 GB of DDR3 memory on a 128 bit bus, with 28.51 GB/s of bandwidth.

Q: What are the core counts and clock frequencies?

A: The Quadro 2000D includes 192 shading units, 32 texture mapping units, and 16 ROPs. Its memory clock is listed at 650 MHz with 2.6 Gbps effective. The GT 640 has 384 shading units, 32 TMUs, and 16 ROPs, with a memory clock of 891 MHz and 1782 Mbps effective.

Q: What is the power draw and physical size?

A: The Quadro 2000D is rated for a 62 W TDP, uses a single-slot cooler, needs no auxiliary power connectors, and the system power supply recommendation is 250 W. The GT 640 has a 65 W TDP, also a single-slot design, no power connectors, and a 250 W recommended PSU.

Q: What display outputs do they offer?

A: The Quadro 2000D features dual DVI outputs. The GT 640 provides one DVI, one HDMI 1.4a, and one DisplayPort 1.2 output.

Architecture Differences

The two cards split along generation lines. The Quadro 2000D is a Fermi architecture product, a 40 nm design with 1,170 million transistors on a 238 mm² die, resulting in a transistor density of 4.9 million per mm². The GT 640 moves to the newer Kepler architecture, built at 28 nm with 1,270 million transistors on a smaller 118 mm² die, which yields a much higher density of 10.8 million transistors per mm². This generational jump explains much of the efficiency difference between them.

The memory subsystems also differ. The Quadro uses GDDR5, a high-bandwidth memory type, on a 128-bit bus. The GT 640 sticks with DDR3 on a 128-bit bus. GDDR5 runs at a higher effective data rate (2.6 Gbps effective for the Quadro vs 1782 Mbps for the GT), which makes the Quadro's bandwidth 41.60 GB/s, a decisive advantage over the GT's 28.51 GB/s.

Compute resources are split differently: the Quadro packs 192 shading units, while the GT 640 doubles that to 384. Both share the same TMU and ROP counts (32 and 16, respectively). The Quadro's pixel throughput is 5.000 GPixel/s and texture rate is 20.00 GTexel/s, while the GT can output 7.216 GPixel/s and 28.86 GTexel/s. FP32 compute is 480.0 GFLOPS for the Quadro vs 692.7 GFLOPS for the GT.

There is no difference in the API support: both cards run DirectX 12 (11_0) and OpenGL 4.6. The GT 640 also supports Vulkan 1.2.175, a feature the Quadro lacks. The Quadro 2000D was released later (October 2011) and is now at end-of-life; the GT 640 came out in June 2012 and is also at end-of-life.

The Verdict

From the raw data, the Quadro 2000D wins in exactly one of the head-to-head benchmark tests recorded: the Geekbench OpenCL test, where it scores 3930 against the GT 640's 3736, a 5.2% lead. Across all recorded tests, the Quadro takes 1 win and the GT 640 takes 0 wins. The database's percentile ranks place the Quadro at the 23rd percentile of all GPUs, while the GT 640 sits at the 20th percentile. This is a narrow margin, but it is in the Quadro's favor.

Despite the similar compute positioning, the two cards are not equal. The Quadro 2000D's advantage comes almost entirely from its memory system: GDDR5 with a much higher effective data rate and bandwidth. That pays off in memory-intensive compute workloads. The GT 640, on the other hand, has more shading units and a much higher FP32 peak (692.7 GFLOPS vs 480.0 GFLOPS), and it comes with 2 GB of VRAM, double the Quadro's 1024 MB. For raw compute throughput and future-proofing memory capacity, the GT 640 is clearly better.

The nearest rivals in the database are informative. The Quadro K2000D averages 3919 (0.3% higher), the GeForce GT 745M scores 3953 (0.6% higher), and the Radeon R5 M420 scores 3956 (0.7% lower). The GT 640's closest rival, the Intel Arc Pro B60, averages 3182, 0.9% lower. The Quadro P1000 is 1.5% higher, and the GeForce 920M is 2.3% above, while the GT 730M is 3.2% lower. These are all within a tight band, meaning the two cards are placed in a very dense performance cluster.

For the workstation user that needs certified drivers and ISV support, the Quadro 2000D is the pick. For a general-purpose consumer card with more memory and higher compute density, the GT 640 is the pick. The data favors the Quadro by a single benchmark, but the GT offers a much bigger memory footprint and higher raw FP32 throughput.

Specification Differences

| Field | NVIDIA Quadro 2000D | NVIDIA GeForce GT 640 |

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

| Chip | GF106 | GK107 |

| Architecture | Fermi | Kepler |

| Process | 40 nm | 28 nm |

| Transistors | 1,170 million | 1,270 million |

| Die size | 238 mm² | 118 mm² |

| Transistor density | 4.9M / mm² | 10.8M / mm² |

| Memory size | 1024 MB | 2 GB |

| Memory type | GDDR5 | DDR3 |

| Memory bus width | 128 bit | 128 bit |

| Memory bandwidth | 41.60 GB/s | 28.51 GB/s |

| Memory clock | 850 MHz | 891 MHz |

| Effective memory clock | 2.6 Gbps | 1782 Mbps |

| Shading units | 192 | 384 |

| TMUs | 32 | 32 |

| ROPs | 16 | 16 |

| Pixel rate | 5.000 GPixel/s | 7.216 GPixel/s |

| Texture rate | 20.00 GTexel/s | 28.86 GTexel/s |

| FP32 | 480.0 GFLOPS | 692.7 GFLOPS |

| TDP | 62 W | 65 W |

| Slot width | Single-slot | Single-slot |

| Power connectors | None | None |

| Suggested PSU | 250 W | 250 W |

| Bus interface | PCIe 2.0 x16 | PCIe 3.0 x16 |

| Display outputs | 2x DVI | 1x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 |

| DirectX | 12 (11_0) | 12 (11_0) |

| OpenGL | 4.6 | 4.6 |

| Vulkan | N/A | 1.2.175 |

| Length | 178 mm (7.0 in) | 145 mm (5.7 in) |

| Release | 2011-10-04 | 2012-06-04 |

| Launch MSRP | 599 USD | 99 USD |

| Predecessor | Quadro FX Tesla | GeForce 500 |

| Successor | Quadro Kepler | GeForce 700 |

Head-to-Head Benchmarks

The single recorded head-to-head benchmark result in the database is the Geekbench OpenCL test. The Quadro 2000D returned a score of 3930, while the GT 640 scored 3736. That is a 5.2% difference in favor of the Quadro. This is the only direct comparison on record, and it shows the Quadro's memory advantage translating into a measurable lead on a compute task, since OpenCL workloads are often bandwidth-sensitive.

The other benchmark results in the database, while not head-to-head runs, place the two cards in the same ranking band. The Quadro's average score across all recorded tests is 3930, and its percentile is the 23rd. The GT 640's average is 3210, at the 20th percentile. The nearest rivals to the Quadro are the K2000D (3919, +0.3%), the GeForce GTX 745M (3953, -0.6%), and the R5 M420 (3956, -0.7%). Around the GT 640, the Intel Arc Pro B60 is 0.9% lower, the Quadro 1000 is 1.5% higher, and the GeForce GTX 920M is 2.3% lower. The GT 730M is 3.2% lower. These clusters of scores tell the compatibility story: neither card is a runaway performer, with the Quadro having a small consistent edge on one test and the GeForce having a strong raw-compute advantage on paper.

The head-to-head table tells the story in one line: the Quadro wins 1, the GT wins 0. The database's overall percentile ranking (23rd vs 20th) aligns with the 5.2% delta in the head-to-head test. But the GT's 384 shading units (double the Quadro's 192) and 692.7 GFLOPS of FP32 (vs 480.0 GFLOPS) show a card that is compute-heavy but memory-light. If a workload is mostly shader math, the GT will pull ahead. If it is memory movement, the Quadro's 41.60 GB/s of bandwidth will dominate. This is a classic bandwidth-vs-shading balance, and the correct choice depends on where the workload sits.