GPU Comparison

The NVIDIA Quadro K2000D and NVIDIA GeForce GT 635M are both end-of-life mobile and workstation graphics solutions from NVIDIA, but they target very different use cases. In the single available benchmark, the Quadro K2000D is the clear winner, posting a Geekbench OpenCL score of 3919 against the GT 635M's 3740, a 4.8% advantage. However, the data shows that this is not a simple story of one card being universally better; the architectural and specification differences create distinct strengths that matter depending on the workload.

Where Each One Wins

The Quadro K2000D wins the only head-to-head benchmark available, the Geekbench OpenCL test, with a score of 3919 compared to the GT 635M's 3740. This 4.8% lead is the sole direct comparison point, but it reflects a substantial underlying hardware advantage. The K2000D's victory is driven by its compute resources: it has 384 shading units and 32 texture mapping units, versus the GT 635M's 96 shading units and 16 TMUs. This 4x difference in shading units directly translates to the higher FP32 performance of 732.7 GFLOPS versus 182.4 GFLOPS. The K2000D also has a much higher memory bandwidth at 64.00 GB/s compared to the GT 635M's 28.80 GB/s, which helps in memory-intensive OpenCL workloads.

The GT 635M, while losing the benchmark war, wins in the efficiency and deployment domain. Its TDP is 35 W, which is lower than the K2000D's 51 W. More importantly, its slot width is listed as "IGP" (Integrated Graphics Processor), meaning it is designed to be soldered onto a motherboard in a laptop, while the K2000D is a single-slot discrete card measuring 202 mm in length. The GT 635M's integrated nature means it has no power connectors and no suggested PSU requirement, whereas the K2000D suggests a 250 W power supply. This makes the GT 635M the only option for ultra-thin, power-constrained portable devices, while the K2000D is a card designed for a desktop workstation slot.

The Verdict

The data is decisive: pick the NVIDIA Quadro K2000D if your priority is raw compute performance in a workstation environment. Its 4.8% benchmark lead is backed by a 4x advantage in shading units, a 4x advantage in TMUs, and more than double the memory bandwidth. With a Geekbench score of 3919, it sits at the 23rd percentile of all GPUs, outperforming its direct rivals like the NVIDIA Quadro 2000D (3930, -0.3%) and the NVIDIA GeForce GT 745M (3953, -0.9%). The K2000D is the choice for anyone needing a capable OpenCL compute device with professional display outputs (2x DVI and 1x mini-DisplayPort 1.2).

Pick the NVIDIA GeForce GT 635M only if the form factor dictates the decision. The data shows it is an integrated part for portable devices, requiring no power connectors and drawing 35 W, which is 16 W less than the K2000D. Its Geekbench score of 3740 places it at the 22nd percentile, and it is competitive with its own rivals, such as the NVIDIA Quadro 3000M (3718, 0.6%) and the NVIDIA GeForce GT 740M (3717, 0.6%). However, its compute capabilities are dramatically lower, with only 96 shading units and a pixel rate of 1.900 GPixel/s compared to the K2000D's 7.632 GPixel/s. The GT 635M is a legacy mobile part for basic graphics acceleration, not a compute or professional visualization tool.

Head-to-Head Benchmarks

The only direct benchmark result is the Geekbench OpenCL test. The Quadro K2000D scored 3919, while the GeForce GT 635M scored 3740, giving the K2000D a 4.8% victory. This is a small percentage delta, but the underlying capabilities suggest the gap would widen in more demanding tasks. The K2000D offers 732.7 GFLOPS of FP32 compute, which is 4 times the 182.4 GFLOPS of the GT 635M. This massive raw compute advantage is somewhat mitigated in the OpenCL test, which likely has memory bandwidth and latency components, but the K2000D still wins on that front with 64.00 GB/s versus 28.80 GB/s.

The pixel and texture rates further illustrate the performance gap. The K2000D can fill 7.632 GPixel/s and 30.53 GTexel/s, while the GT 635M manages only 1.900 GPixel/s and 7.600 GTexel/s. These are 4x and 4x differences, respectively, matching the shading unit and TMU counts. The K2000D's 16 ROPs are also 4x the 4 ROPs on the GT 635M, which directly impacts fill-rate bound scenarios. In summary, the 4.8% OpenCL delta is a compressed representation of a much larger architectural performance gap; the K2000D is fundamentally a more powerful processor.

FAQ

Q: Which GPU has a higher Geekbench OpenCL score?

A: The NVIDIA Quadro K2000D scores 3919, which is 4.8% higher than the GeForce GT 635M's score of 3740.

Q: How do their shading unit counts compare?

A: The Quadro K2000D has 384 shading units, while the GeForce GT 635M has 96 shading units, a 4x difference in favor of the K2000D.

Q: Is the GeForce GT 635M suitable for a desktop workstation?

A: No. The GT 635M has a slot width of "IGP" and its display outputs are "Portable Device Dependent," indicating it is an integrated graphics processor for laptops. The K2000D is a single-slot, 202 mm discrete card with fixed display outputs.

Q: What is the memory bandwidth of each card?

A: The Quadro K2000D has a memory bandwidth of 64.00 GB/s using GDDR5 memory, while the GeForce GT 635M has a bandwidth of 28.80 GB/s using DDR3 memory.

Q: Which GPU has a higher transistor density?

A: The Quadro K2000D, built on a 28 nm process, has a transistor density of 10.8M / mm², compared to the GT 635M's 5.0M / mm² on a 40 nm process.

Q: Do both GPUs support DirectX 12?

A: Yes, both the Quadro K2000D and the GeForce GT 635M are listed as supporting DirectX 12 (11_0) and OpenGL 4.6.

Architecture Differences

The two GPUs are built on fundamentally different architectures from NVIDIA. The Quadro K2000D uses the Kepler architecture with the GK107 chip, while the GeForce GT 635M uses the older Fermi architecture with the GF108 chip. This architectural generation gap is significant. The K2000D is fabricated on a 28 nm process at TSMC, allowing for 1,270 million transistors on a 118 mm² die. The GT 635M uses a 40 nm process, also at TSMC, with 585 million transistors on a 116 mm² die. This means the K2000D packs more than double the transistors into a similar die size, resulting in a transistor density of 10.8M / mm² versus the GT 635M's 5.0M / mm².

The compute core configurations differ drastically. The K2000D features 384 shading units, 32 TMUs, and 16 ROPs, while the GT 635M has 96 shading units, 16 TMUs, and only 4 ROPs. This 4x disparity in core counts is the primary driver of the performance difference. Neither GPU has dedicated RT cores or tensor cores. In terms of software support, the K2000D supports Vulkan 1.2.175, while the GT 635M has no Vulkan support listed. Both support DirectX 12 (11_0) and OpenGL 4.6, but the newer Kepler architecture generally has better feature support and efficiency.

Specification Differences

The specification sheets reveal a clear divergence in design goals. The Quadro K2000D is a professional workstation card with a launch MSRP of 599 USD. It has a TDP of 51 W and requires a suggested 250 W power supply, using no external power connectors. It is a single-slot card with a length of 202 mm and a height of 111 mm, featuring 2x DVI and 1x mini-DisplayPort 1.2 outputs. Its memory runs at 1000 MHz (4 Gbps effective) with a 128-bit bus, delivering 64.00 GB/s of bandwidth from 2 GB of GDDR5.

The GeForce GT 635M is a mobile integrated part with no launch MSRP listed. It has a lower TDP of 35 W and no suggested PSU, as it is an IGP. It is portable-device dependent in its display outputs and has no physical dimensions listed. Its memory is 2 GB of DDR3 running at 900 MHz (1800 Mbps effective) on a 128-bit bus, yielding only 28.80 GB/s of bandwidth. The K2000D also has a higher pixel rate (7.632 GPixel/s vs 1.900 GPixel/s) and texture rate (30.53 GTexel/s vs 7.600 GTexel/s). The K2000D's release date is later (2013-02-28) than the GT 635M (2012-03-21), and its predecessor is Quadro Fermi, while the GT 635M's predecessor is GeForce 500M.