AMD Radeon HD 7790 vs NVIDIA GeForce GTX 580 Comparison

Head-to-Head Benchmarks

The benchmark data available for the AMD Radeon HD 7790 and the NVIDIA GeForce GTX 580 comes from different testing frameworks, which makes a direct apples-to-apples comparison difficult. The HD 7790 was measured using the Geekbench Metal test, while the GTX 580 was measured using the Geekbench OpenCL test. Within these separate tests, each card posts a respectable score that places it within the same general performance tier, but the lack of a shared workload means the database does not record a head-to-head matchup where both cards ran the same benchmark.

Looking at the recorded numbers, the AMD Radeon HD 7790 achieved a Geekbench Metal score of 17,666. This places it at the 61st percentile among all GPUs in the database. Its nearest rivals in this test are the NVIDIA GeForce RTX 4060 with an average score of 17,639, putting the HD 7790 just 0.2% ahead, and the AMD Radeon 780M with 17,588, where the HD 7790 leads by 0.4%. The AMD Radeon Pro 560 trails at 17,551, a 0.7% gap in favor of the HD 7790. On the other side, the NVIDIA Quadro RTX 4000 posts 17,789, which is 0.7% higher than the HD 7790. These are extremely tight margins, indicating that the HD 7790 sits in a narrow band of performance where a few points can shift the ranking.

The NVIDIA GeForce GTX 580 recorded a Geekbench OpenCL score of 15,283, placing it at the 58th percentile among all GPUs. Its nearest rivals cluster even closer together. The NVIDIA GeForce RTX 2060 scores 15,290, a 0% delta relative to the GTX 580, meaning the two are effectively tied in this test. The AMD Radeon 680M posts 15,270, just 0.1% behind, while the NVIDIA GeForce RTX 3050 OEM scores 15,199, a 0.6% deficit. The AMD Radeon RX 7600 trails at 15,171, which is 0.7% behind the GTX 580. Again, the GTX 580 finds itself in a tightly contested cluster where the difference between first and last among its nearest rivals is under one percent.

When comparing the two cards directly using their respective benchmark scores, the HD 7790's 17,666 is higher than the GTX 580's 15,283 by roughly 15.6%. However, this gap must be interpreted with caution because the tests are not the same. The Metal framework is optimized for AMD hardware in many workloads, and OpenCL is a separate compute API. The database does not claim that one card is definitively faster than the other; it simply records two different test results. The percentile rankings do show a small edge for the HD 7790, 61st versus 58th, but that is a narrow difference in the broader GPU landscape.

Architecture Differences

The two cards represent fundamentally different generations of GPU design. The AMD Radeon HD 7790 is built on the Bonaire chip using the GCN 2.0 architecture, part of the Southern Islands generation under the HD 7700 family. It is manufactured on a 28 nm process at TSMC. The chip contains 2,080 million transistors on a die size of 160 mm², giving it a transistor density of 13.0 million per square millimeter. The NVIDIA GeForce GTX 580, by contrast, uses the GF110 chip with the Fermi 2.0 architecture, belonging to the GeForce 500 generation. It is built on a 40 nm process, also at TSMC. The GF110 packs 3,000 million transistors onto a much larger die of 520 mm², resulting in a transistor density of only 5.8 million per square millimeter.

The process node difference is significant. The 28 nm node used for the HD 7790 is denser and more power-efficient than the 40 nm node used for the GTX 580. This explains why the HD 7790 can deliver comparable compute performance with far fewer transistors and a much smaller die. The GTX 580's larger, older chip consumes considerably more power and produces more heat, but it also allows for a wider memory bus and more ROPs.

In terms of compute resources, the HD 7790 features 896 shading units, 56 texture mapping units, and 16 ROPs. The GTX 580 has 512 shading units, 64 TMUs, and 48 ROPs. The GTX 580 has fewer shaders but more texture units and triple the ROP count. This configuration suggests the GTX 580 was designed for higher fill-rate workloads, while the HD 7790 leans on its larger shader array for compute throughput.

Memory architecture also differs substantially. The HD 7790 uses 1024 MB of GDDR5 on a 128-bit bus, delivering 96.00 GB/s of bandwidth. The GTX 580 uses 1536 MB of GDDR5 on a 384-bit bus, delivering 192.4 GB/s. The GTX 580 has 50% more memory capacity and double the memory bandwidth. The memory clock differs as well: the HD 7790 runs at 1500 MHz (6 Gbps effective), while the GTX 580 runs at 1002 MHz (4 Gbps effective). The HD 7790's memory runs at a higher clock speed, but the GTX 580's wider bus more than compensates.

The API support reflects the different design eras. The HD 7790 supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.2.170. The GTX 580 supports DirectX 12 (11_0) and OpenGL 4.6, but has no Vulkan support listed. This is a meaningful gap for modern titles that rely on Vulkan for cross-platform rendering. The HD 7790 also has a more modern display output set: one DVI, one HDMI 1.4a, and two mini-DisplayPort 1.2 connections. The GTX 580 offers two DVI ports and one mini-HDMI 1.3a. The HD 7790 supports PCIe 3.0 x16, while the GTX 580 uses the older PCIe 2.0 x16 interface.

Where Each One Wins

The AMD Radeon HD 7790 wins in compute-oriented workloads that favor its GCN 2.0 architecture and higher shader count. Its 1.792 TFLOPS of FP32 performance exceeds the GTX 580's 1.581 TFLOPS by approximately 13.3%. This gives the HD 7790 a measurable advantage in general-purpose compute tasks, particularly those that scale with shader throughput. The HD 7790 also supports Vulkan 1.2.170, which opens the door to modern cross-platform graphics APIs that the GTX 580 cannot access. For users running Vulkan-based games or compute applications, the HD 7790 is the only one of the two that can participate.

The HD 7790 also wins decisively on power efficiency. Its TDP is 85 W versus 244 W for the GTX 580, meaning the GTX 580 draws nearly three times as much power. The HD 7790 requires only a single 6-pin power connector and a 250 W suggested PSU, while the GTX 580 needs one 6-pin and one 8-pin connector with a 550 W suggested PSU. The HD 7790 is also shorter at 183 mm (7.2 inches) compared to 267 mm (10.5 inches) for the GTX 580, making it easier to fit in compact cases.

The NVIDIA GeForce GTX 580 wins in memory-intensive workloads. Its 192.4 GB/s of bandwidth is exactly double the HD 7790's 96.00 GB/s. This matters for high-resolution textures, heavy anti-aliasing, and any workload that streams large amounts of data through memory. The GTX 580's 1536 MB of VRAM also provides 50% more capacity than the HD 7790's 1024 MB, which helps in scenarios where texture sets exceed 1 GB. The wider 384-bit bus gives the GTX 580 a structural advantage in fill-rate-bound scenes, and its 24.70 GPixel/s pixel rate is 54.4% higher than the HD 7790's 16.00 GPixel/s.

The GTX 580 also has more TMUs, 64 versus 56, but its texture rate of 49.41 GTexel/s is actually lower than the HD 7790's 56.00 GTexel/s. This is because the HD 7790's higher clock speed compensates for having fewer texture units. So in texture-heavy workloads, the HD 7790 holds a 13.3% advantage despite the GTX 580's higher TMU count.

In terms of raw benchmark scores, the HD 7790 recorded a higher number (17,666 in Metal) than the GTX 580 (15,283 in OpenCL), but these are different tests. The HD 7790's 61st percentile versus the GTX 580's 58th percentile is a modest edge, not a dominant one. The database records zero head-to-head wins for either card, reflecting the absence of shared benchmark data.

Specification Differences

The following specifications differ between the two cards:

• Architecture: GCN 2.0 (HD 7790) versus Fermi 2.0 (GTX 580)
• Process node: 28 nm (HD 7790) versus 40 nm (GTX 580)
• Transistors: 2,080 million (HD 7790) versus 3,000 million (GTX 580)
• Die size: 160 mm² (HD 7790) versus 520 mm² (GTX 580)
• Transistor density: 13.0M / mm² (HD 7790) versus 5.8M / mm² (GTX 580)
• Memory clock: 1500 MHz / 6 Gbps effective (HD 7790) versus 1002 MHz / 4 Gbps effective (GTX 580)
• Memory size: 1024 MB (HD 7790) versus 1536 MB (GTX 580)
• Memory bus width: 128 bit (HD 7790) versus 384 bit (GTX 580)
• Memory bandwidth: 96.00 GB/s (HD 7790) versus 192.4 GB/s (GTX 580)
• Shading units: 896 (HD 7790) versus 512 (GTX 580)
• TMUs: 56 (HD 7790) versus 64 (GTX 580)
• ROPs: 16 (HD 7790) versus 48 (GTX 580)
• Pixel rate: 16.00 GPixel/s (HD 7790) versus 24.70 GPixel/s (GTX 580)
• Texture rate: 56.00 GTexel/s (HD 7790) versus 49.41 GTexel/s (GTX 580)
• FP32 performance: 1.792 TFLOPS (HD 7790) versus 1.581 TFLOPS (GTX 580)
• TDP: 85 W (HD 7790) versus 244 W (GTX 580)
• Power connectors: 1x 6-pin (HD 7790) versus 1x 6-pin + 1x 8-pin (GTX 580)
• Suggested PSU: 250 W (HD 7790) versus 550 W (GTX 580)
• Bus interface: PCIe 3.0 x16 (HD 7790) versus PCIe 2.0 x16 (GTX 580)
• Display outputs: 1x DVI, 1x HDMI 1.4a, 2x mini-DisplayPort 1.2 (HD 7790) versus 2x DVI, 1x mini-HDMI 1.3a (GTX 580)
• DirectX support: 12 (12_0) (HD 7790) versus 12 (11_0) (GTX 580)
• Vulkan support: 1.2.170 (HD 7790) versus none (GTX 580)
• Dimensions: 183 mm length (HD 7790) versus 267 mm length, 111 mm height (GTX 580)
• Release date: 2013-03-21 (HD 7790) versus 2010-11-08 (GTX 580)
• Launch MSRP: 149 USD (HD 7790) versus 499 USD (GTX 580)

Both cards are end-of-life products, use GDDR5 memory, are dual-slot designs, and have no ray tracing or tensor cores. Both support OpenGL 4.6. The HD 7790's predecessor is Northern Islands and its successor is Sea Islands. The GTX 580's predecessor is GeForce 400 and its successor is GeForce 600.

FAQ

Q: Which card has a higher raw benchmark score?

A: The AMD Radeon HD 7790 recorded a Geekbench Metal score of 17,666, while the NVIDIA GeForce GTX 580 recorded a Geekbench OpenCL score of 15,283. However, these are different benchmark tests, so the scores are not directly comparable.

Q: How do their power requirements compare?

A: The HD 7790 has a TDP of 85 W and requires a 250 W suggested PSU with a single 6-pin connector. The GTX 580 has a TDP of 244 W and requires a 550 W suggested PSU with one 6-pin and one 8-pin connector.

Q: Does the GTX 580 have more memory bandwidth?

A: Yes, significantly. The GTX 580 delivers 192.4 GB/s over a 384-bit bus, while the HD 7790 delivers 96.00 GB/s over a 128-bit bus. The GTX 580 also has 1536 MB of VRAM compared to 1024 MB on the HD 7790.

Q: Which card supports Vulkan?

A: Only the AMD Radeon HD 7790 supports Vulkan, with version 1.2.170. The NVIDIA GeForce GTX 580 has no Vulkan support listed in the database.

Q: How do their compute capabilities compare?

A: The HD 7790 has higher FP32 performance at 1.792 TFLOPS, compared to 1.581 TFLOPS for the GTX 580. The HD 7790 also has more shading units (896 versus 512). The GTX 580 counters with a higher pixel rate of 24.70 GPixel/s versus 16.00 GPixel/s.

Q: What are their release dates and MSRPs?

A: The HD 7790 launched on 2013-03-21 with a launch MSRP of 149 USD. The GTX 580 launched on 2010-11-08 with a launch MSRP of 499 USD.

The Verdict

The data paints a clear picture of two cards from different design philosophies. The AMD Radeon HD 7790 represents the modern approach: a smaller, denser chip on a 28 nm process that achieves higher compute throughput with dramatically lower power consumption. Its 1.792 TFLOPS of FP32 performance, 896 shading units, 13.3% higher texture rate, and support for Vulkan 1.2.170 make it the better choice for compute-oriented tasks and modern API compatibility. Its 85 W TDP and single 6-pin connector mean it can run in systems with a 250 W PSU, and its 183 mm length fits easily into compact builds. The recorded Metal benchmark score of 17,666 and 61st percentile ranking also suggest strong performance relative to its era.

The NVIDIA GeForce GTX 580 takes the opposite approach. Its 40 nm Fermi 2.0 chip is physically massive at 520 mm², consumes 244 W, and requires a 550 W PSU, but it delivers double the memory bandwidth at 192.4 GB/s and triple the ROP count of the HD 7790. For workloads that are bottlenecked by memory throughput or fill rate, the GTX 580's 24.70 GPixel/s pixel rate and 1536 MB VRAM give it a structural advantage. Its 58th percentile ranking and OpenCL score of 15,283 show it still holds its own in the database, but its lack of Vulkan support and older DirectX 11_0 feature level limit its reach in modern applications.

For users prioritizing power efficiency, modern API support, and compute throughput, the HD 7790 is the data-backed choice. For users who need maximum memory bandwidth and fill rate in legacy applications, the GTX 580 retains a purpose. The HD 7790's higher percentile ranking and newer architecture give it the edge in overall capability, but the GTX 580's memory subsystem remains its strongest asset. Both are end-of-life products, so the decision ultimately comes down to which workload profile matters more.