NVIDIA GeForce 920M vs NVIDIA Quadro 2000D Comparison

The Verdict

The data presents a clear, if narrow, split between these two end-of-life NVIDIA parts. The NVIDIA Quadro 2000D, a Fermi-era workstation card, takes the single recorded head-to-head victory in Geekbench OpenCL test with a score of 3930, which is 5.5% ahead of the GeForce 920M's score of 3725. The Quadro 2000D wins the 1 recorded benchmark comparison, while the GeForce 920M wins 0 recorded comparisons. The Quadro 2000D sits at the 23rd percentile of all GPUs, while the 920M sits at the 20th percentile, indicating the Quadro is rated higher overall in the database. For users whose workload is represented by OpenCL compute, the Quadro 2000D is the better pick based on its benchmark victory and higher average score. The GeForce 920M, however, is the only one of the two with a recorded Vulkan benchmark score, 2849, making it the only option for that API in this pairing. The 920M also offers double the memory at 2 GB compared to the Quadro's 1024 MB. The verdict is that the Quadro 2000D is the stronger compute performer in OpenCL, but the GeForce 920M provides a more modern feature set and larger memory buffer, with a much lower 33 W TDP compared to the Quadro's 62 W.

Where Each One Wins

The NVIDIA Quadro 2000D wins in raw compute performance as measured by Geekbench OpenCL. Its score of 3930 places it 5.5% ahead of the GeForce 920M's 3725 in the head-to-head benchmark. This is its sole recorded benchmark victory. It also has a higher average benchmark score of 3930 compared to the 920M's 3287, a substantial 16.4% difference in average performance. The Quadro 2000D also offers a significantly higher memory bandwidth of 41.60 GB/s on a 128-bit bus, compared to the 920M's 14.40 GB/s on a 64-bit bus. This bandwidth advantage, coupled with the GDDR5 memory type, gives the workstation card a distinct edge in memory-intensive tasks. It also has double the ROPs, 16 versus the 920M's 8, and a higher pixel rate of 5.000 GPixel/s, though the 920M actually has a higher pixel rate at 7.632 GPixel/s, so the ROP count is not the sole determinant. The Quadro also has a higher texture rate of 20.00 GTexel/s, though this is lower than the 920M's 30.53 GTexel/s.

The NVIDIA GeForce 920M wins in efficiency and modern features. Its TDP is 33 W, nearly half the Quadro 2000D's 62 W, making it far more suitable for power-constrained environments. It is the only one of the two with a recorded Vulkan benchmark score of 2849, which is a significant advantage for any application using the Vulkan API. It has a larger memory capacity of 2 GB versus 1 GB, allowing for larger datasets. The 920M is built on a smaller 28 nm process, which contributes to its lower power consumption. It also has double the shading units, 384 versus 192, and a higher FP32 compute of 732.7 GFLOPS, indicating a higher theoretical compute ceiling in some workloads, despite losing the recorded OpenCL test. It also supports PCIe 3.0 x8, while the older Quadro is limited to PCIe 2.0 x16.

Architecture Differences

The two GPUs come from different NVIDIA architectures and generations. The Quadro 2000D uses the GF106 chip based on the Fermi architecture, while the GeForce 920M uses the GK208B chip based on the Kepler 2.0 architecture. The Quadro 2000D is from the Quadro Fermi (x000) generation, released on 2011-10-04, whereas the GeForce 920M is from the GeForce 900M generation, released later on 2015-03-12.

The process node differs significantly. The Quadro 2000D is fabricated on a 40 nm process at TSMC, while the GeForce 920M uses a more advanced 28 nm process at the same foundry. This results in different transistor densities: the Quadro has 1,170 million transistors on a 238 mm² die, for a density of 4.9 million transistors per mm². The 920M has fewer total transistors, 1,020 million, but on a much smaller 87 mm² die, yielding a density of 11.7 million transistors per mm². This indicates the Keplper 2.0 part is a much denser, more modern design.

Memory architecture is a major differentiator. The Quadro 2000D has 1024 MB of GDDR5 memory on a 128-bit bus, providing 41.60 GB/s of bandwidth. The GeForce 920M has 2 GB of DDR3 memory on a 64-bit bus, providing only 14.40 GB/s of bandwidth. The Quadro's memory clock is listed as 650 MHz (2.6 Gbps effective), while the 920M's is 900 MHz (1800 Mbps effective). The compute resources also differ: the Quadro has 192 shading units, 32 TMUs, and 16 ROPs, while the 920M has 384 shading units, 32 TMUs, and only 8 ROPs. The 920M has a higher FP32 rating at 732.7 GFLOPS versus the Quadro's 480.0 GFLOPS, and a higher texture rate at 30.53 GTexel/s versus 20.00 GTexel/s. The 920M also has a higher pixel rate at 7.632 GPixel/s versus the Quadro's 5.000 GPixel/s.

In terms of API support, both support DirectX 12 (11_0 and OpenGL 4.6. The 920M adds Vulkan support at version 1.2.175, while the Quadro 2000D has no recorded Vulkan support. The Quadro 2000D is a single-slot card with 2x DVI display outputs, a 178 mm length, and a 111 mm height, and it requires a 250 W suggested PSU. The GeForce 920M is an integrated GPU (IGP) with portable device dependent display outputs, and it has no suggested PSU. Neither card requires external power connectors. The Quadro 2000D has a launch MSRP of 599 USD, while the 920M has no recorded launch MSRP.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The NVIDIA Quadro 2000D has an average benchmark score of 3930, which is higher than the NVIDIA GeForce 920M's average of 3287.

Q: Does the GeForce 920M support Vulkan?

A: Yes, the GeForce 920M has a recorded Geekbench Vulkan score of 2849 and supports Vulkan version 1.2.175. The Quadro 2000D has no recorded Vulkan support or score.

Q: How much faster is the Quadro 2000D in the head-to-head benchmark?

A: The Quadro 2000D is 5.5% faster than the GeForce 920M, scoring 3930 versus 3725 in the Geekbench OpenCL test.

Q: What are the TDP differences between the two cards?

A: The Quadro 2000D has a TDP of 62 W, while the GeForce 920M has a much lower TDP of 33 W, making the 920M more power-efficient.

Q: Which card has more memory bandwidth and what type of memory does each use, and what is the size difference?

A: The Quadro 2000D has 1024 MB of GDDR5, whereas the GeForce 920M has 2 GB of DDR3 memory. The Quadro's bandwidth is 41.60 GB/s, and the 920M has 14.40 GB/s.

Q: What is the transistor count on each chip?

A: The Quadro 2000D has 1,170 million transistors on a 238 mm² die, while the GeForce 920M has 1,020 million transistors on a much smaller 87 mm² die.

Head-to-Head Benchmarks

The database only records a single head-to-head benchmark between the two GPUs, which is Geekbench OpenCL. In this test, the NVIDIA Quadro 2000D wins decisively, posting a score of 3930 against the GeForce 920M's 3725. This translates to a 5.5% delta percentage difference between the two. This is the only recorded win for either GPU, with the Quadro taking 1 win and the 920M taking 0 wins.

The results are consistent with the broader benchmark averages. The Quadro 2000D's average score of 3930 is consistent with its single recorded OpenCL run, while the 920M's average score of 3287 is pulled down from its OpenCL score of 3725 by its lower Vulkan score of 2849. The 920M's nearest rival, the NVIDIA GeForce GT 730M, has an average score of 3316, which is 0.9% lower than the 920M, while the Intel HD Graphics 530 is 1.4% lower at 3332, and the Intel HD Graphics P4600 is 3% lower at 3389. The only rival with a higher average score is the NVIDIA GeForce GT 640 at 3210, which is 2.4% higher than the 920M. The Quadro 2000D's nearest rival, the NVIDIA Quadro K2000D, has an average score of 3919, only 0.3% lower than the Quadro 2000D, making it an extremely close match. The GeForce GT 745M is 0.6% higher at 3953, the AMD Radeon R5 M420 is 0.7% higher at 3956, and the GeForce 830M is also 0.7% higher at 3957.

The 5.5% lead in OpenCL computes is the single most important data point for this pair, but the broader picture shows the Quadro 2000D at the 23rd percentile versus the 920M at the 20th. This indicates that while the 920M has a higher theoretical FP32 throughput (732.7 GFLOPS) and more shading units (384), the real-world OpenCL workload represented by the Geekbench test favors the Quadro 2000D's architecture and memory subsystem. The Quadro's 41.60 GB/s of bandwidth, enabled by GDDR5 on a 128-bit bus, likely contributes to its OpenCL win over the 920M's 14.40 GB/s of DDR3 bandwidth on a 64-bit bus. The 920M does have a higher pixel rate at 7.632 GPixel/s and texture rate at 30.53 GTexel/s, which are not captured in the head-to-head but are noted in the full database.