AMD Radeon 740M vs NVIDIA GeForce GTX 590 Comparison

# FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon 740M posts an average benchmark score of 12870, which is 0.3% ahead of the NVIDIA GeForce GTX 590's 12830. This places the 740M at the 53rd percentile of all GPUs, versus the GTX 590's 52nd percentile.

Q: How do the two compare in the only head-to-head benchmark available?

A: In the Geekbench OpenCL test, the NVIDIA GeForce GTX 590 scores 12830 against the AMD Radeon 740M's 10172, giving the GTX 590 a 20.7% win. This is the sole head-to-head benchmark recorded, and it favors NVIDIA by a substantial margin.

Q: What are the nearest rivals to each GPU based on average scores?

A: For the AMD Radeon 740M, the closest rival is the AMD FirePro W5100 with an average score of 12847 (a 0.2% delta). The NVIDIA GeForce GTX 590's nearest rival is the AMD Radeon Pro 455 at 12831 (a 0.0% delta). Both GPUs sit in a tight cluster around the 12830-12870 score range.

Q: Which GPU has a higher pixel fill rate?

A: The AMD Radeon 740M achieves a pixel rate of 22.40 GPixel/s, while the NVIDIA GeForce GTX 590 reaches 19.46 GPixel/s. This gives the 740M a 15% advantage in pixel throughput.

Q: What are the thermal design power differences?

A: The AMD Radeon 740M has a TDP of 45 W, while the NVIDIA GeForce GTX 590 draws 365 W. The GTX 590 also requires two 8-pin power connectors and a 750 W suggested power supply, whereas the 740M is an integrated GPU with no power connectors.

Q: Which GPU supports newer API features?

A: The AMD Radeon 740M supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, whereas the NVIDIA GeForce GTX 590 supports DirectX 12 (11_0) and has no Vulkan support listed. The 740M also includes 4 ray tracing cores, which the GTX 590 lacks entirely.

Architecture Differences

The architectural gap between these two GPUs is generational and profound. The AMD Radeon 740M is built on TSMC's 4 nm process node, packing 20,900 million transistors into a 137 mm² die, yielding a transistor density of 152.6M per mm². In contrast, the NVIDIA GeForce GTX 590 uses TSMC's 40 nm process, containing 3,000 million transistors across a 520 mm² die, with a density of just 5.8M per mm². This means the 740M packs over 26 times more transistors per square millimeter than the GTX 590.

The AMD chip is based on the RDNA 3.0 architecture from the Phoenix2 chip, part of the Navi III IGP generation. It implements 256 shading units, 16 texture mapping units, and 8 raster operation pipelines, alongside 4 dedicated ray tracing cores. The NVIDIA GPU uses the Fermi 2.0 architecture with the GF110 chip from the GeForce 500 generation, offering 512 shading units, 64 TMUs, and 48 ROPs, but no ray tracing hardware.

Clock behavior differs fundamentally. The Radeon 740M has a base clock of 800 MHz and a boost clock of 2800 MHz, while the GTX 590 lists no base or boost clocks in its specifications. Memory architecture is also divergent: the 740M uses system-shared memory with system-dependent bandwidth, whereas the GTX 590 has dedicated 1536 MB of GDDR5 memory on a 384-bit bus delivering 164.0 GB/s.

Compute output shows a reversal of the hardware resource counts. The Radeon 740M delivers 2.867 TFLOPS of FP32 performance, more than double the GTX 590's 1,244.2 GFLOPS. The 740M also achieves a texture rate of 44.80 GTexel/s versus the GTX 590's 38.91 GTexel/s. However, the GTX 590 has no listed FP16 support, while the 740M provides FP16 at a 1:1 ratio. The GTX 590 uses a PCIe 2.0 x16 interface, while the 740M runs on PCIe 4.0 x8.

The Verdict

The data paints a clear but nuanced picture. For raw computational throughput and modern feature support, the AMD Radeon 740M is the superior product. It doubles the FP32 performance of the GTX 590, offers ray tracing cores, supports DirectX 12 Ultimate, and consumes just 45 W versus 365 W. Its average benchmark score of 12870 also edges out the GTX 590's 12830, and it ranks slightly higher at the 53rd percentile versus 52nd.

However, the NVIDIA GeForce GTX 590 wins the only direct head-to-head benchmark recorded. In Geekbench OpenCL, it scores 12830 against the 740M's 10172, a 20.7% advantage. This suggests that in certain OpenCL workloads, the GTX 590's larger memory bus (384-bit versus system shared) and dedicated GDDR5 memory provide a tangible benefit, even with its older architecture.

For modern users, the Radeon 740M is the rational choice. It is an active, current product with a 2024 release date, while the GTX 590 is end-of-life from 2011. The 740M's far lower power draw, smaller footprint (IGP versus dual-slot), and superior API support make it viable for contemporary software. The GTX 590, despite its head-to-head win, is a legacy part that lacks Vulkan support and ray tracing, and its 365 W TDP with dual 8-pin connectors makes it impractical for most modern systems. Those prioritizing peak OpenCL performance in legacy workloads might favor the GTX 590, but the 740M is the better all-around GPU by every modern metric.

Specification Differences

The two GPUs differ across nearly every specification field. The AMD Radeon 740M uses a 4 nm TSMC process, while the NVIDIA GeForce GTX 590 uses a 40 nm TSMC process. Transistor counts are 20,900 million versus 3,000 million, and die sizes are 137 mm² versus 520 mm², giving transistor densities of 152.6M/mm² and 5.8M/mm² respectively.

The 740M has a base clock of 800 MHz and a boost clock of 2800 MHz; the GTX 590 lists no base or boost clocks but has a memory clock of 854 MHz (3.4 Gbps effective). Memory configurations are entirely different: the 740M uses system-shared memory, while the GTX 590 has 1536 MB of GDDR5 with a 384-bit bus and 164.0 GB/s bandwidth.

Shading units number 256 on the 740M versus 512 on the GTX 590; TMUs are 16 versus 64; ROPs are 8 versus 48. The 740M includes 4 ray tracing cores, while the GTX 590 has none. Pixel rates are 22.40 GPixel/s versus 19.46 GPixel/s, and texture rates are 44.80 GTexel/s versus 38.91 GTexel/s. FP32 output is 2.867 TFLOPS versus 1,244.2 GFLOPS, and FP16 is 2.867 TFLOPS on the 740M with no FP16 listed for the GTX 590.

Power requirements differ drastically: 45 W TDP with no power connectors for the 740M, versus 365 W TDP with two 8-pin connectors and a 750 W suggested PSU for the GTX 590. The 740M is an IGP with motherboard-dependent outputs and PCIe 4.0 x8; the GTX 590 is a dual-slot card measuring 279 mm by 111 mm by 40 mm, with 3x DVI and 1x mini-DisplayPort outputs on PCIe 2.0 x16. API support also diverges: DirectX 12 Ultimate versus DirectX 12 (11_0), both with OpenGL 4.6, but Vulkan 1.4 only on the 740M.

Head-to-Head Benchmarks

The only direct comparison available is the Geekbench OpenCL test, and it is a decisive win for the NVIDIA GeForce GTX 590. The GTX 590 scores 12830, while the AMD Radeon 740M scores 10172, resulting in a 20.7% delta in NVIDIA's favor. This is a substantial margin in a compute-focused workload, and it suggests that the GTX 590's dedicated 1536 MB of GDDR5 memory on a 384-bit bus provides a significant advantage over the 740M's system-shared memory, which has system-dependent bandwidth.

Despite this head-to-head loss, the 740M's average benchmark score of 12870 is actually higher than the GTX 590's 12830. This apparent contradiction is explained by the 740M's second benchmark, Geekbench Vulkan, where it scores 15568. That Vulkan result, combined with the OpenCL score of 10172, yields an average that surpasses the GTX 590's single OpenCL result. The GTX 590 has no Vulkan benchmark recorded and no Vulkan support listed.

The nearest-rival data reinforces how closely matched these two are in overall performance. The 740M's rivals include the AMD FirePro W5100 (12847, 0.2% delta), AMD Radeon Pro 455 (12831, 0.3% delta), NVIDIA GeForce GTX 590 (12830, 0.3% delta), and AMD Radeon RX 580 (12928, -0.4% delta). The GTX 590's rivals are nearly identical: AMD Radeon Pro 455 (12831, 0.0% delta), AMD FirePro W5100 (12847, -0.1% delta), AMD Radeon 740M (12870, -0.3% delta), and NVIDIA GeForce GTX 670 (12773, 0.4% delta). Both GPUs sit within a 1% band of each other and their closest competitors.

Where Each One Wins

The NVIDIA GeForce GTX 590 wins in the OpenCL compute arena. Its 20.7% head-to-head victory in Geekbench OpenCL demonstrates that its dedicated GDDR5 memory and 384-bit memory bus deliver real benefits in memory-bandwidth-sensitive workloads. For users running legacy OpenCL applications that cannot leverage newer APIs, the GTX 590's 12830 score is the stronger result. Its 512 shading units and 64 TMUs also provide raw parallel throughput that can be advantageous in certain compute tasks.

The AMD Radeon 740M wins in nearly every other category. Its FP32 performance of 2.867 TFLOPS more than doubles the GTX 590's 1,244.2 GFLOPS, making it better suited for modern gaming and compute workloads that rely on shader throughput. Its Vulkan score of 15568 shows exceptional performance in modern graphics APIs, which is absent on the GTX 590 entirely. The 740M also wins on pixel rate (22.40 GPixel/s versus 19.46 GPixel/s) and texture rate (44.80 GTexel/s versus 38.91 GTexel/s).

The 740M's 4 ray tracing cores give it hardware-accelerated ray tracing, a feature the GTX 590 lacks. Its DirectX 12 Ultimate support and Vulkan 1.4 compatibility make it future-proof for current software. The 45 W TDP versus 365 W TDP means the 740M is viable in thin-and-light systems, while the GTX 590 requires a desktop chassis with substantial power delivery. The 740M is an active product released in 2024, while the GTX 590 is end-of-life from 2011. For anyone building or upgrading a system today, the 740M's combination of modern features, higher average score, and dramatically lower power draw makes it the clear winner, despite the GTX 590's single head-to-head benchmark victory.