AMD Radeon 660M vs NVIDIA GeForce GTX 770 Comparison

The NVIDIA GeForce GTX 770 and AMD Radeon 660M represent two very different approaches to graphics processing: a dedicated, power-hungry desktop card from 2013 versus a modern integrated processor (IGP) from 2022. The benchmark data shows a clear and consistent performance hierarchy, but the context of each product’s design goals is crucial to understanding the numbers. The GTX 770 wins both head-to-head tests, yet the 660M operates at a fraction of the power draw, making the comparison less about outright speed and more about the trade-off between raw performance and efficiency.

Head-to-Head Benchmarks

The Geekbench results provide a direct quantitative comparison across two compute APIs. In the OpenCL test, the NVIDIA GeForce GTX 770 scores 15,707 against the AMD Radeon 660M’s 12,876. This yields a delta of 22% in favor of the GTX 770, marking a substantial lead in general-purpose compute workloads. The Vulkan test shows an even larger gap: the GTX 770 reaches 19,166 while the 660M trails at 14,748, a 30% advantage for NVIDIA. These are not marginal differences; the dedicated Kepler card outpaces the RDNA 2 IGP by a significant margin in both synthetic tests.

Looking at the broader context, the GTX 770’s average benchmark score of 14,747 places it in the 57th percentile of all GPUs. Its nearest rival, the AMD Radeon RX 5500 XT, scores 14,692, meaning the GTX 770 is just 0.4% ahead. This shows that despite its age, the GTX 770 performs on par with a much newer entry-level discrete card. The 660M, with an average score of 13,812, sits in the 55th percentile, and its closest comparison is the NVIDIA RTX A2000 Mobile at 13,821, where the 660M is essentially tied (a delta of -0.1%). This places the 660M in the same performance tier as a mobile workstation GPU, which is impressive for an integrated solution.

The head-to-head data leaves no ambiguity about the winner in raw compute. The GTX 770’s victory in both OpenCL and Vulkan is consistent with its higher transistor count (3,540 million vs. 13,100 million) dedicated to graphics, though the 660M’s chip also includes CPU cores. The 30% delta in Vulkan is particularly notable, suggesting the GTX 770’s architecture handles modern API overhead better in these tests, or that its higher memory bandwidth (224.4 GB/s vs. system dependent) provides a tangible benefit in fill-rate-heavy tasks.

Where Each One Wins

The NVIDIA GeForce GTX 770 wins in every benchmark category where data exists. Its two victories in OpenCL and Vulkan are the only head-to-head results available, and it wins both by double-digit percentages. For users prioritizing maximum compute throughput in these specific workloads, the GTX 770 is the clear choice. Its 3.333 TFLOPS of FP32 performance, combined with 128 texture mapping units and 32 ROPs, gives it a substantial hardware advantage in traditional rasterization and compute tasks. The 2 GB of GDDR5 memory on a 256-bit bus provides 224.4 GB/s of bandwidth, which is critical for high-resolution textures and complex scenes.

The AMD Radeon 660M, while losing both tests, wins in every aspect of efficiency and portability. Its 40 W TDP is a fraction of the GTX 770’s 230 W, and it requires no power connectors or a dedicated power supply recommendation. The 660M is an IGP, meaning it is designed to be integrated into a mobile processor, eliminating the need for a separate graphics card. In this context, its performance is remarkable: it achieves 87% of the GTX 770’s OpenCL score (12,876 vs. 15,707) while using 83% less power. For applications that are power-constrained or require a compact form factor, the 660M is the only viable option, regardless of its lower absolute scores.

Neither GPU supports the full feature set of modern graphics. The GTX 770 is limited to DirectX 12 (11_0) and Vulkan 1.2.175, while the 660M supports DirectX 12 Ultimate (12_2) and Vulkan 1.4. The 660M also includes 6 ray tracing cores, which the GTX 770 lacks entirely. This means the 660M can handle ray-traced workloads and newer graphical effects, while the GTX 770 cannot, even though the 660M’s raw compute is lower. In this sense, the 660M wins in architectural features, even if it loses in raw benchmark scores.

The Verdict

The data clearly indicates that the NVIDIA GeForce GTX 770 is the faster GPU in raw compute benchmarks. It wins both head-to-head tests, and its average score of 14,747 is 6.8% higher than the 660M’s 13,812. The GTX 770’s percentile rank of 57 versus the 660M’s 55 further confirms its overall performance advantage. For any user who needs maximum FP32 throughput, texture fill rate (138.9 GTexel/s vs. 45.60 GTexel/s), or pixel rate (34.72 GPixel/s vs. 30.40 GPixel/s), the GTX 770 is the superior choice. Its closest rivals are all discrete desktop or workstation cards, underscoring its position as a dedicated performance part.

However, the verdict is not a simple recommendation of one over the other. The AMD Radeon 660M is an integrated processor designed for laptops and compact devices. Its 40 W TDP, lack of power connectors, and IGP form factor make it the only option for systems where a discrete card is physically impossible or impractical. The GTX 770 is a dual-slot, 267 mm long card that requires a 550 W power supply and 1x 6-pin plus 1x 8-pin connectors. No user would choose the 660M for a desktop build with space for a dedicated card, and no user would choose the GTX 770 for a thin-and-light laptop.

The selection depends entirely on the use case. If the priority is absolute performance in OpenCL and Vulkan workloads, and the system can accommodate a large, power-hungry card, the GTX 770 is the winner. If the priority is efficiency, portability, and modern API support, the 660M is the only viable choice. The 660M’s inclusion of ray tracing cores and DirectX 12 Ultimate support makes it more future-proof for modern games, but its lower raw compute means it will struggle with demanding titles at high settings. The GTX 770 remains a capable performer for its age, but its end-of-life status and lack of modern features limit its longevity.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The NVIDIA GeForce GTX 770 has an average benchmark score of 14,747, which is 6.8% higher than the AMD Radeon 660M’s 13,812.

Q: How much faster is the GTX 770 in Vulkan?

A: In the Geekbench Vulkan test, the GTX 770 scores 19,166, which is 30% higher than the 660M’s 14,748.

Q: Does the AMD Radeon 660M support ray tracing?

A: Yes, the 660M includes 6 ray tracing cores, while the GTX 770 has no ray tracing cores at all.

Q: What are the TDP requirements for each GPU?

A: The GTX 770 has a TDP of 230 W and requires a 550 W suggested power supply, while the 660M has a TDP of 40 W and requires no external power connectors.

Q: Which GPU supports a newer DirectX version?

A: The AMD Radeon 660M supports DirectX 12 Ultimate (12_2), while the NVIDIA GeForce GTX 770 is limited to DirectX 12 (11_0).

Q: What is the release date difference between the two?

A: The GTX 770 was released on May 29, 2013, while the 660M was released on January 3, 2022, a gap of nearly nine years.

Architecture Differences

The two GPUs are built on fundamentally different architectures and process technologies. The NVIDIA GeForce GTX 770 uses the GK104 chip based on the Kepler architecture, manufactured on a 28 nm process at TSMC. This chip contains 3,540 million transistors on a 294 mm² die, resulting in a transistor density of 12.0 million per mm². In contrast, the AMD Radeon 660M uses the Rembrandt chip based on RDNA 2.0 architecture, manufactured on a 6 nm process, also at TSMC. This newer chip packs 13,100 million transistors onto a smaller 208 mm² die, achieving a much higher density of 63.0 million per mm².

The 660M’s 6 nm process allows for twice the transistor density, enabling more features in a smaller footprint. The 660M includes 6 ray tracing cores, which are absent from the GTX 770. The 660M also supports a wider range of modern APIs, including DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the GTX 770 is limited to DirectX 12 (11_0) and Vulkan 1.2.175. Both support OpenGL 4.6, but the 660M’s newer architecture allows for hardware-accelerated ray tracing and variable rate shading, which the Kepler architecture cannot handle.

The 660M’s shading units are fewer (384 vs. 1,536), and its texture units (24) and ROPs (16) are also lower than the GTX 770’s 128 TMUs and 32 ROPs. However, the 660M’s higher clock speeds (1,900 MHz boost vs. 1,085 MHz boost) partially compensate. The 660M also supports FP16 at 2.918 TFLOPS (2:1 ratio), while the GTX 770 has no listed FP16 performance. The 660M’s memory is system shared, meaning it depends on the host system’s RAM, whereas the GTX 770 has dedicated 2 GB GDDR5. This fundamental difference in memory architecture impacts bandwidth, with the GTX 770 providing a fixed 224.4 GB/s versus the 660M’s system-dependent bandwidth.

Specification Differences

The most obvious specification difference is memory. The GTX 770 comes with 2 GB of GDDR5 memory on a 256-bit bus, providing 224.4 GB/s of bandwidth. The 660M uses system shared memory, meaning its size, type, bus width, and bandwidth are all dependent on the host system’s configuration. This makes the 660M’s memory performance variable, while the GTX 770 offers consistent, dedicated bandwidth.

Clock speeds differ significantly. The GTX 770 has a base clock of 1046 MHz and a boost of 1085 MHz, with memory clocked at 1753 MHz (7 Gbps effective). The 660M has a higher base clock of 1500 MHz and a boost of 1900 MHz, but its memory clock is listed as system shared. The 660M’s higher core clocks help it compete despite having far fewer execution units.

Power and physical requirements are where the two diverge most starkly. The GTX 770 has a TDP of 230 W, is dual-slot, requires 1x 6-pin + 1x 8-pin power connectors, and needs a 550 W power supply. It measures 267 mm in length, 111 mm in height, and 38 mm in width. The 660M is an IGP with a 40 W TDP, no power connectors, and no dimensions listed, as it is integrated into a processor. The GTX 770 uses PCIe 3.0 x16, while the 660M uses PCIe 4.0 x8. Display outputs also differ: the GTX 770 offers 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2, while the 660M’s outputs are portable device dependent. The GTX 770 had a launch MSRP of 399 USD, while the 660M has no listed launch MSRP. The GTX 770’s predecessor is GeForce 600 and its successor is GeForce 900, while the 660M’s predecessor is Vega II IGP and its successor is Navi III IGP.