GPU Comparison

The NVIDIA GeForce GT 640 and NVIDIA Quadro P1000 occupy the same 20th percentile of all GPUs, with average benchmark scores of 3210 and 3163, respectively. Despite this statistical near-tie, the two cards are separated by five years of architecture evolution, a 2:1 memory capacity gap, and wildly different compute profiles. The data shows two products that land at the same percentile but achieve it through entirely different means.

Head-to-Head Benchmarks

The head-to-head results reveal a decisive advantage for the Quadro P1000 in every shared test. In Geekbench OpenCL, the Quadro P1000 scores 13,584 against the GT 640's 3,736, representing a 72.5% margin of victory. This is not a marginal win; it is a three-and-a-half-fold increase in raw compute throughput. The Geekbench Vulkan test tells a similar story, with the Quadro P1000 posting 7,739 versus 3,809 for the GT 640, a 50.8% advantage.

The deltaPct values in the rival data frame these results in context. The Quadro P1000 sits only 0.6% behind the Intel Arc Pro B60 in average score, while the GT 640 sits 0.9% ahead of that same Arc Pro B60. In other words, the two NVIDIA cards are essentially swapped positions around the same baseline. The GT 640's closest rival, the GeForce 920M, scores 3,287, which is 2.3% higher, while the Quadro P1000 trails that same 920M by 3.8%. When the two cards face each other directly, the Quadro P1000 wins both available benchmarks outright, with zero wins for the GT 640.

The OpenCL gap is particularly telling for compute workloads. A 72.5% deficit means the GT 640 would take roughly three and a half times as long to complete the same OpenCL task, assuming linear scaling. The Vulkan gap of 50.8% is smaller but still substantial, indicating the Pascal architecture's superior graphics pipeline efficiency. Notably, the GT 640 lacks benchmark entries for DirectX 10, 11, 12, or 9, while the Quadro P1000 has scores for all of those tests, suggesting the older card was not even run through those suites.

FAQ

Q: Which card has the higher average benchmark score?

A: The NVIDIA GeForce GT 640 has a marginally higher average benchmark score of 3,210, compared to the Quadro P1000's 3,163. The difference is 1.5% in favor of the GT 640 when measured from the Quadro's perspective, but the head-to-head tests show the Quadro winning by 72.5% and 50.8% in OpenCL and Vulkan, respectively.

Q: How do the two cards compare in terms of memory bandwidth?

A: The Quadro P1000 offers 80.19 GB/s of bandwidth from 4 GB of GDDR5 memory on a 128-bit bus, while the GT 640 provides 28.51 GB/s from 2 GB of DDR3 on the same 128-bit bus. The Quadro's bandwidth is nearly three times higher, which directly contributes to its superior OpenCL and Vulkan performance.

Q: What is the thermal design power difference between the two cards?

A: The GT 640 has a 65 W TDP with a suggested 250 W power supply, while the Quadro P1000 has a 47 W TDP with a suggested 200 W power supply. The newer Pascal card draws 18 watts less and requires a smaller power supply, despite delivering substantially higher compute performance.

Q: Do both cards support the same DirectX version?

A: No. The GT 640 supports DirectX 12 (11_0), while the Quadro P1000 supports DirectX 12 (12_1). Both cards support OpenGL 4.6, but the Vulkan support differs: the GT 640 supports Vulkan 1.2.175, while the Quadro P1000 supports Vulkan 1.4.

Q: Which card has more shading units and texture mapping units?

A: The Quadro P1000 has 640 shading units and 40 TMUs, compared to the GT 640's 384 shading units and 32 TMUs. The Quadro also has 32 ROPs versus the GT 640's 16 ROPs, giving it double the pixel throughput capability.

Q: What are the display output configurations for each card?

A: The GT 640 has one DVI, one HDMI 1.4a, and one DisplayPort 1.2 output. The Quadro P1000 has four mini-DisplayPort 1.4a outputs, offering more display connectivity and a newer DisplayPort standard.

Architecture Differences

The GT 640 is built on the GK107 chip using the Kepler architecture, fabricated by TSMC on a 28 nm process. It integrates 1,270 million transistors on a 118 mm² die, yielding a transistor density of 10.8 million per square millimeter. The Quadro P1000 uses the GP107 chip on the Pascal architecture, manufactured by Samsung on a 14 nm process. That chip packs 3,300 million transistors into a 132 mm² die, achieving a density of 25.0 million per square millimeter. The Pascal chip more than doubles the transistor count while only increasing die size by 14 mm², a direct consequence of the smaller process node.

The memory subsystems diverge sharply. The GT 640 uses 2 GB of DDR3 with a 128-bit bus, producing 28.51 GB/s of bandwidth. The Quadro P1000 uses 4 GB of GDDR5 with the same 128-bit bus, but delivers 80.19 GB/s. This is a 51.68 GB/s absolute difference, and the GDDR5 memory operates at 1253 MHz with 5 Gbps effective speed, versus 891 MHz with 1782 Mbps effective speed for the DDR3 on the GT 640.

Compute resources differ across the board. The GT 640 has 384 shading units, 32 TMUs, and 16 ROPs. The Quadro P1000 has 640 shading units, 40 TMUs, and 32 ROPs. Pixel rate jumps from 7.216 GPixel/s on the GT 640 to 47.36 GPixel/s on the Quadro P1000, while texture rate rises from 28.86 GTexel/s to 59.20 GTexel/s. FP32 performance scales from 692.7 GFLOPS to 1.894 TFLOPS, a 2.73x increase. The Quadro P1000 also lists FP16 performance at 29.60 GFLOPS (1:64), while the GT 640 has no FP16 data. Neither card has ray tracing or tensor cores.

Specification Differences

The two cards differ in nearly every measurable specification. The GT 640 uses a 28 nm TSMC process; the Quadro P1000 uses a 14 nm Samsung process. Transistor count is 1,270 million versus 3,300 million, and die size is 118 mm² versus 132 mm². Transistor density is 10.8M/mm² versus 25.0M/mm². The Quadro P1000 has base and boost clocks of 1266 MHz and 1480 MHz, while the GT 640 has no listed base or boost clocks, only a memory clock of 891 MHz. The Quadro's memory clock is 1253 MHz.

Memory capacity doubles from 2 GB to 4 GB, and type changes from DDR3 to GDDR5. Bandwidth increases from 28.51 GB/s to 80.19 GB/s. Shading units rise from 384 to 640, TMUs from 32 to 40, and ROPs from 16 to 32. Pixel rate goes from 7.216 GPixel/s to 47.36 GPixel/s, texture rate from 28.86 GTexel/s to 59.20 GTexel/s, and FP32 from 692.7 GFLOPS to 1.894 TFLOPS. The Quadro P1000 lists FP16 at 29.60 GFLOPS; the GT 640 has none. TDP drops from 65 W to 47 W, and suggested PSU from 250 W to 200 W. Both are single-slot with no power connectors.

Display outputs change from 1x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 to 4x mini-DisplayPort 1.4a. DirectX support advances from 12 (11_0) to 12 (12_1), and Vulkan from 1.2.175 to 1.4. Both share OpenGL 4.6 and PCIe 3.0 x16 interfaces. Dimensions are similar: 145 mm (5.7 inches) length for the GT 640 versus 150 mm (5.9 inches) for the Quadro P1000, with the Quadro also listing a height of 69 mm (2.7 inches). The GT 640 has a launch MSRP of 99 USD; the Quadro P1000 has no launch MSRP listed. Release dates are 2012-06-04 for the GT 640 and 2017-02-06 for the Quadro P1000. Both are end-of-life.

The Verdict

The data supports a clear division of purpose. The GT 640, with its 99 USD launch MSRP and 2012 release date, was positioned as an entry-level desktop card. Its 20th percentile ranking and 3,210 average score place it alongside the GeForce 920M, which scores 3,287. For legacy applications that rely on its specific Kepler-era feature set, the GT 640 remains functional, but its 28.51 GB/s bandwidth and 692.7 GFLOPS FP32 are binding constraints.

The Quadro P1000, despite scoring 1.5% lower on average, is the superior performer in every shared benchmark. Its 72.5% OpenCL advantage and 50.8% Vulkan advantage are decisive. The 4 GB GDDR5 memory with 80.19 GB/s bandwidth, combined with 1.894 TFLOPS FP32 and 47.36 GPixel/s pixel rate, makes it a far more capable compute and graphics device. The 47 W TDP and 200 W suggested PSU also mean it delivers this performance at lower power draw.

Users should choose the Quadro P1000 for any workload involving OpenCL compute, modern Vulkan applications, or multi-display setups requiring four mini-DisplayPort outputs. Its DirectX 12 (12_1) support and Vulkan 1.4 are also ahead of the GT 640's DirectX 12 (11_0) and Vulkan 1.2.175. The GT 640 should only be selected for systems that specifically require its DVI output, its 2012-era driver stack, or its particular Kepler-based feature set. For raw performance, the Quadro P1000 wins both head-to-head tests, and the architecture data confirms why: double the shading units, double the ROPs, 2.73x the FP32 throughput, and nearly three times the memory bandwidth. The percentile ranking is coincidental; the benchmark deltas are not.