AMD Radeon HD 7790 vs NVIDIA GeForce GTX 690 Comparison

The AMD Radeon HD 7790 and the NVIDIA GeForce GTX 690 represent two very different approaches to GPU design from the same 28 nm TSMC manufacturing process. The HD 7790 is a compact, single-chip solution built on GCN 2.0 architecture, while the GTX 690 is a dual-GPU flagship built on Kepler. The benchmark data shows the GTX 690 holds a narrow overall edge in average score, but the HD 7790 posts a higher single-test result, creating a split that depends entirely on which workload and API is being examined.

Where Each One Wins

The AMD Radeon HD 7790 wins on the sole benchmark result recorded for it, the Geekbench Metal test, where it scores 17666. That score places it in the 61st percentile of all GPUs, which is a strong showing for a card with 896 shading units and a 128-bit memory bus. Its nearest rival in that result is the NVIDIA GeForce RTX 4060, which scores 17639, a delta of just 0.2 percent in favor of the HD 7790. The HD 7790 also edges out the AMD Radeon 780M (17588, 0.4 percent slower) and the AMD Radeon Pro 560 (17551, 0.7 percent slower). The only card in its immediate rival group that beats it is the NVIDIA Quadro RTX 4000, which scores 17789, a 0.7 percent advantage. This makes the HD 7790 the winner in Metal-specific workloads, which are generally associated with Apple platforms and compute tasks.

The NVIDIA GeForce GTX 690 wins on its two recorded benchmarks, Geekbench OpenCL and Geekbench Vulkan. It scores 17399 in OpenCL and 16675 in Vulkan, giving it an average benchmark score of 17037. That average places it in the 60th percentile of all GPUs, just one percentile point below the HD 7790. The GTX 690’s nearest rival is the AMD Radeon HD 7970M, which scores 17019, a 0.1 percent difference in favor of the GTX 690. It also leads the NVIDIA Tesla M4 (16932, 0.6 percent slower) but trails the AMD Radeon RX 7600 XT (17083, 0.3 percent faster) and the NVIDIA GeForce RTX 3070 (17208, 1 percent faster). The GTX 690’s wins come in cross-platform compute APIs, where its dual-GPU configuration and higher raw throughput — 3.130 TFLOPS FP32 versus 1.792 TFLOPS — show up more clearly.

The split is clear: the HD 7790 wins in Metal, the GTX 690 wins in OpenCL and Vulkan. For users whose workloads rely on Metal, the HD 7790 is the better performer. For those using OpenCL or Vulkan, the GTX 690 takes the lead. The GTX 690 also has a higher average score overall, 17037 versus 17666 for the HD 7790, but that average is pulled down by its Vulkan result, which is 724 points lower than its OpenCL score.

The Verdict

The data presents a trade-off rather than a clear overall winner. The AMD Radeon HD 7790 should be chosen by users who prioritize Metal performance, as it scores 17666 in that test, which is higher than any single score recorded for the GTX 690. Its 61st percentile ranking also suggests it sits slightly higher in the overall distribution than the GTX 690’s 60th percentile. The HD 7790 is also far more efficient on paper, with a TDP of 85 watts compared to 300 watts for the GTX 690, and it requires only a single 6-pin power connector versus two 8-pin connectors on the GTX 690. Its 250-watt suggested PSU is dramatically lower than the GTX 690’s 700-watt recommendation, making it the sensible choice for systems with modest power delivery.

The NVIDIA GeForce GTX 690 should be chosen by users who need OpenCL or Vulkan performance. Its OpenCL score of 17399 is 267 points lower than the HD 7790’s Metal score, but its Vulkan score of 16675 is still respectable. More importantly, the GTX 690 has double the memory of the HD 7790 — 2 GB versus 1 GB — and a 256-bit bus that delivers 192.3 GB/s of bandwidth compared to 96.00 GB/s. It also has more shading units (1536 versus 896), more texture mapping units (128 versus 56), and more render output units (32 versus 16). For compute-heavy tasks in OpenCL, the GTX 690’s 3.130 TFLOPS FP32 throughput is 74.7 percent higher than the HD 7790’s 1.792 TFLOPS, which explains its advantage in that API.

The verdict depends on workload. For Metal users, the HD 7790 is the superior card. For OpenCL and Vulkan users, the GTX 690 is the superior card. The HD 7790 is also the better fit for power-constrained builds, while the GTX 690 is the better fit for high-throughput compute tasks. Neither card dominates the other across all metrics, and the 1-percentile gap in overall standing is negligible in practice.

Head-to-Head Benchmarks

There are no direct head-to-head benchmark results shared between the two cards in the data. Instead, each card has its own set of tests: the HD 7790 has one Geekbench Metal result (17666), and the GTX 690 has Geekbench OpenCL (17399) and Geekbench Vulkan (16675) results. Comparing these across APIs shows the largest single-score gap is in favor of the HD 7790: its Metal score of 17666 is 991 points higher than the GTX 690’s Vulkan score of 16675. That is a 5.9 percent difference based on the raw numbers, and it demonstrates that the HD 7790 can outperform the GTX 690 in the right API context.

The biggest win for the GTX 690 comes when comparing its OpenCL score to its own Vulkan score, but that is an internal comparison. Against the HD 7790, the GTX 690’s OpenCL score of 17399 is 267 points lower than the HD 7790’s Metal score, so the GTX 690 does not beat the HD 7790 on any direct cross-API comparison. However, the GTX 690’s average score of 17037 is derived from two tests, while the HD 7790’s average equals its single Metal score of 17666. The GTX 690’s higher memory bandwidth (192.3 GB/s versus 96.00 GB/s) and higher texture rate (130.4 GTexel/s versus 56.00 GTexel/s) suggest it should excel in bandwidth-bound workloads, but the benchmark data does not include a test that captures that advantage.

The closest comparison in the data is between the HD 7790’s Metal score (17666) and the GTX 690’s OpenCL score (17399), a difference of 267 points or 1.5 percent in favor of the HD 7790. The GTX 690’s Vulkan score (16675) is 991 points lower than the HD 7790’s Metal score, a 5.9 percent gap. This indicates that the HD 7790 has a higher peak performance in its best API, while the GTX 690 is more consistent across its two tested APIs, differing by 724 points between OpenCL and Vulkan.

FAQ

Q: Which card has the higher average benchmark score?

A: The AMD Radeon HD 7790 has an average benchmark score of 17666, while the NVIDIA GeForce GTX 690 has an average of 17037. The HD 7790’s average is based on a single Metal test, while the GTX 690’s is based on OpenCL and Vulkan tests.

Q: How do the two cards compare in percentile ranking?

A: The HD 7790 sits in the 61st percentile of all GPUs, and the GTX 690 sits in the 60th percentile. The 1-percentile gap is small, but it favors the HD 7790.

Q: What is the memory configuration difference between the two cards?

A: The HD 7790 has 1024 MB of GDDR5 memory on a 128-bit bus, delivering 96.00 GB/s of bandwidth. The GTX 690 has 2 GB of GDDR5 memory on a 256-bit bus, delivering 192.3 GB/s of bandwidth.

Q: Which card has higher FP32 compute throughput?

A: The GTX 690 has 3.130 TFLOPS FP32, which is 74.7 percent higher than the HD 7790’s 1.792 TFLOPS. This is reflected in the GTX 690’s higher shading unit count (1536 versus 896).

Q: What are the power requirements for each card?

A: The HD 7790 has a TDP of 85 watts and requires a 250-watt suggested PSU with one 6-pin connector. The GTX 690 has a TDP of 300 watts and requires a 700-watt suggested PSU with two 8-pin connectors.

Q: Which card supports newer API versions?

A: The HD 7790 supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.2.170. The GTX 690 supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The GTX 690 has a slightly newer Vulkan version, but the HD 7790 has a higher DirectX feature level.

Architecture Differences

The AMD Radeon HD 7790 uses the Bonaire chip, built on GCN 2.0 architecture, and belongs to the Southern Islands generation (HD 7700 family). The NVIDIA GeForce GTX 690 uses the GK104 chip, built on Kepler architecture, and belongs to the GeForce 600 generation. Both are fabricated on a 28 nm process at TSMC, but the GTX 690’s chip is substantially larger: it contains 3,540 million transistors on a 294 mm² die, while the HD 7790 contains 2,080 million transistors on a 160 mm² die. Transistor density is slightly higher on the HD 7790 at 13.0M per mm², versus 12.0M per mm² on the GTX 690.

The GTX 690 is a dual-GPU card, which explains its much higher transistor count and die size. It has 1536 shading units, 128 texture mapping units, and 32 render output units. The HD 7790 has 896 shading units, 56 texture mapping units, and 16 render output units. The GTX 690 also has higher clock speeds, with a base of 915 MHz and a boost of 1019 MHz, while the HD 7790 has no listed base or boost clocks. The GTX 690’s pixel rate is 32.61 GPixel/s and its texture rate is 130.4 GTexel/s, both roughly double the HD 7790’s 16.00 GPixel/s and 56.00 GTexel/s.

Neither card has ray tracing cores or tensor cores. The HD 7790 supports DirectX 12 (12_0), while the GTX 690 supports DirectX 12 (11_0), indicating the HD 7790 has a higher feature level. Both support OpenGL 4.6, but the GTX 690 has a slightly newer Vulkan version (1.2.175 versus 1.2.170). The HD 7790 was released on 2013-03-21, following the Northern Islands generation and succeeded by Sea Islands. The GTX 690 was released on 2012-05-02, following GeForce 500 and succeeded by GeForce 700. Both are end-of-life products.

Specification Differences

The two cards differ in nearly every measurable specification. The HD 7790 has 1024 MB of memory, while the GTX 690 has 2 GB. The HD 7790’s memory runs at 1500 MHz (6 Gbps effective) on a 128-bit bus, yielding 96.00 GB/s bandwidth. The GTX 690’s memory runs at 1502 MHz (6 Gbps effective) on a 256-bit bus, yielding 192.3 GB/s bandwidth. The GTX 690 has exactly double the bus width and double the bandwidth.

Shading units differ significantly: 896 on the HD 7790 versus 1536 on the GTX 690. Texture mapping units are 56 versus 128, and render output units are 16 versus 32. The HD 7790 produces 16.00 GPixel/s and 56.00 GTexel/s, while the GTX 690 produces 32.61 GPixel/s and 130.4 GTexel/s. FP32 compute is 1.792 TFLOPS on the HD 7790 and 3.130 TFLOPS on the GTX 690, a 74.7 percent advantage for the GTX 690.

Power consumption is a major differentiator. The HD 7790 has a TDP of 85 watts, a single 6-pin power connector, and a suggested PSU of 250 watts. The GTX 690 has a TDP of 300 watts, two 8-pin power connectors, and a suggested PSU of 700 watts. Both are dual-slot cards, but the GTX 690 is physically larger: 279 mm long, 111 mm tall, and 38 mm wide, versus 183 mm long for the HD 7790 (height and width not listed). The HD 7790 has display outputs of 1x DVI, 1x HDMI 1.4a, and 2x mini-DisplayPort 1.2. The GTX 690 has 3x DVI and 1x mini-DisplayPort 1.2. Both use PCIe 3.0 x16 interfaces. The HD 7790 has a launch MSRP of 149 USD, while the GTX 690 has a launch MSRP of 999 USD.