NVIDIA GeForce GTX 690 vs NVIDIA T400 Comparison

The NVIDIA GeForce GTX 690 and NVIDIA T400 represent two vastly different eras and purposes in NVIDIA’s lineup, separated by nearly a decade of architectural evolution. The GTX 690 is a dual-GPU flagship from the Kepler era, built to dominate at the high end of the consumer market, while the T400 is a low-profile Turing-based workstation card designed for efficiency and basic professional display tasks. Benchmark data shows the older card retains a slight edge in raw compute, but the two are far closer in average scores than their generational gap would suggest, with the GTX 690 averaging 17037 and the T400 averaging 16508.

FAQ

Q: Which card has the higher average benchmark score?

A: The NVIDIA GeForce GTX 690 has a higher average benchmark score of 17037, compared to the NVIDIA T400’s 16508. This places both cards in the 60th percentile of all GPUs, though the GTX 690 edges out its rival by roughly 3.2% in overall average performance.

Q: How do the two cards compare in OpenCL performance specifically?

A: In the Geekbench OpenCL test, the GTX 690 scores 17399, narrowly beating the T400’s 17039. This represents a 2.1% lead for the GTX 690, making it the winner in this particular workload.

Q: What is the difference in Vulkan performance between the two?

A: The GTX 690 also wins the Vulkan test, scoring 16675 against the T400’s 15976. The delta here is larger at 4.4%, showing the older Kepler card has a more pronounced advantage in this API.

Q: Which card consumes less power?

A: The T400 is dramatically more power-efficient, with a TDP of 30 W compared to the GTX 690’s 300 W. The T400 requires no power connectors and a suggested PSU of only 200 W, while the GTX 690 needs two 8-pin connectors and a 700 W power supply.

Q: What are the key architectural differences between the two?

A: The GTX 690 uses the Kepler architecture on a 28 nm TSMC process with 3,540 million transistors on a 294 mm² die. The T400 uses Turing on a 12 nm node with 4,700 million transistors packed into a 200 mm² die, giving it a much higher transistor density of 23.5M per mm² versus 12.0M per mm².

Q: Which card has better API support for modern software?

A: The T400 supports DirectX 12 (12_1) and Vulkan 1.4, while the GTX 690 is limited to DirectX 12 (11_0) and Vulkan 1.2.175. Both share OpenGL 4.6 support, but the T400 is clearly better prepared for current-generation graphics APIs.

Architecture Differences

The architectural gap between these two cards is generational and profound. The GTX 690 is built on Kepler, fabricated on a 28 nm process at TSMC, housing 3,540 million transistors across a 294 mm² die. In contrast, the T400 uses the Turing architecture on a much more advanced 12 nm node, with 4,700 million transistors squeezed into a smaller 200 mm² die. This gives the T400 a transistor density of 23.5M per mm², nearly double the GTX 690’s 12.0M per mm², reflecting the massive manufacturing improvements between 2012 and 2021.

The compute resources tell a story of brute force versus efficiency. The GTX 690 packs 1536 shading units, 128 TMUs, and 32 ROPs, while the T400 offers just 384 shading units, 24 TMUs, and 16 ROPs. Despite having a quarter of the shader count, the T400 achieves comparable benchmark scores thanks to its much higher boost clock of 1425 MHz versus the GTX 690’s 1019 MHz boost. The base clocks are even more divergent, with the T400 idling at 420 MHz and the GTX 690 at 915 MHz.

Memory architecture also differs fundamentally. The GTX 690 uses 2 GB of GDDR5 on a 256-bit bus, delivering 192.3 GB/s of bandwidth. The T400 also has 2 GB, but it is GDDR6 on a 64-bit bus, yielding only 80.00 GB/s. The T400 compensates with a higher effective memory speed of 10 Gbps versus the GTX 690’s 6 Gbps, but the narrower bus limits overall throughput. Neither card features ray tracing or tensor cores, as both predate or omit those specialized units.

Head-to-Head Benchmarks

The head-to-head data shows a clear, if modest, victory for the GTX 690 in both tested workloads. In Geekbench OpenCL, the GTX 690 scores 17399 against the T400’s 17039, a 2.1% advantage. This is a surprisingly narrow margin given the GTX 690’s massive lead in shading units and memory bandwidth, but the T400’s architectural efficiency and higher clocks narrow the gap considerably. The OpenCL result places both cards within a hair of each other, with the GTX 690’s nearest rivals including the AMD Radeon HD 7970M at 17019 and the NVIDIA Tesla M4 at 16932, while the T400 sits near the NVIDIA GeForce RTX 5090 D V2 at 16504 and the AMD Radeon PRO W7500 at 16415.

The Vulkan benchmark tells a similar story with a slightly larger spread. The GTX 690 scores 16675, while the T400 manages 15976, giving the older card a 4.4% win. This delta is more substantial and suggests that the GTX 690’s raw compute throughput translates better to Vulkan’s lower-level API access. However, it is worth remembering the T400’s Vulkan driver support is far newer, with Vulkan 1.4 support versus the GTX 690’s 1.2.175, so software maturity may be a factor in this result. Overall, the GTX 690 wins both head-to-head tests, taking 2 wins to 0 for the T400, but the performance gap is far smaller than the nine-year release date difference would imply.

Specification Differences

The specification sheet reveals stark contrasts across nearly every category. The GTX 690 uses the GK104 chip under the Kepler architecture, while the T400 uses TU117 under Turing. Process nodes differ significantly: 28 nm for the GTX 690 against 12 nm for the T400. Transistor counts also diverge, with the GTX 690 at 3,540 million and the T400 at 4,700 million, despite the latter having a smaller die size of 200 mm² versus 294 mm².

Clock speeds show an interesting inversion. The GTX 690 has a higher base clock at 915 MHz versus the T400’s 420 MHz, but the T400’s boost clock of 1425 MHz exceeds the GTX 690’s 1019 MHz. Memory configurations differ in type and bus width: GDDR5 on a 256-bit bus for the GTX 690, GDDR6 on a 64-bit bus for the T400. Effective memory speed is 6 Gbps on the GTX 690 and 10 Gbps on the T400, but bandwidth heavily favors the GTX 690 at 192.3 GB/s versus 80.00 GB/s.

Power requirements could not be more different. The GTX 690 draws 300 W and requires two 8-pin power connectors plus a 700 W suggested PSU, while the T400 sips just 30 W, needs no external power connectors, and works with a 200 W PSU. Physical dimensions also differ, with the GTX 690 being a dual-slot card measuring 279 mm in length, while the T400 is a single-slot design with unspecified dimensions. Display outputs favor the T400 with three mini-DisplayPort 1.4a connectors, whereas the GTX 690 offers three DVI and one mini-DisplayPort 1.2. The GTX 690 had a launch MSRP of 999 USD, while the T400 has no recorded launch price.

Where Each One Wins

The GTX 690 wins in raw performance across both benchmark tests, making it the stronger choice for compute-heavy workloads that can leverage its 3.130 TFLOPS of FP32 power and 130.4 GTexel/s texture rate. Its 192.3 GB/s memory bandwidth is a massive advantage for tasks that are bandwidth-sensitive, such as high-resolution texture streaming or large data sets in OpenCL applications. The dual-GPU design, while not explicitly listed in the specs, is reflected in the high shading unit count and makes it suitable for scenarios where parallel throughput is paramount. Its 60th percentile ranking and proximity to the NVIDIA GeForce RTX 3070 in average score (17208, a -1% delta) show that even a decade later, it holds its own against modern mid-range hardware in synthetic benchmarks.

The T400 wins decisively in efficiency and modern feature support. Its 30 W TDP makes it suitable for systems with minimal power delivery, and its single-slot, connector-free design allows for installation in compact workstations. The Turing architecture brings DirectX 12 (12_1) and Vulkan 1.4 support, which are essential for modern professional applications and newer game engines. Its FP16 performance of 2.189 TFLOPS (2:1) is a capability the GTX 690 lacks entirely, offering advantages in AI inference and certain compute tasks that utilize half-precision arithmetic. The T400’s nearest rivals in average score include the NVIDIA RTX PRO 6000 Blackwell at 16408 and the AMD Radeon RX 5700 XT at 16361, showing it competes with much more powerful cards in overall compute despite its low-power design. For professional environments prioritizing quiet operation, minimal heat output, and broad software compatibility, the T400 is the pragmatic choice, even if it loses the raw performance race.