AMD Radeon 660M vs NVIDIA GeForce RTX 2060 Comparison

NVIDIA GeForce RTX 2060 vs AMD Radeon 660M

Head-to-Head Benchmarks

The recorded data shows a decisive victory for the NVIDIA GeForce RTX 2060 in the only two benchmarks where both products share a common test. In Geekbench OpenCL, the RTX 2060 scores 65,014 points against the Radeon 660M’s 12,876 points, a delta of 404.9%. This means the RTX 2060 delivers roughly five times the compute throughput in this API. The margin is similarly lopsided in Geekbench Vulkan, where the RTX 2060 posts 65,846 points versus 14,748 points for the Radeon 660M, a 346.5% advantage.

The win count in the head-to-head table is 2 for NVIDIA and 0 for AMD. There are no benchmark results where the Radeon 660M comes out ahead. The RTX 2060’s average benchmark score of 15,290 further reinforces its position, placing it in the 58th percentile of all GPUs. The Radeon 660M’s average score is 13,812, which puts it in the 55th percentile. While both sit in the mid-range of the overall distribution, the absolute gap of 1,478 points between their averages is substantial.

When looking at the nearest rivals for each card, the RTX 2060 is bracketed by the GeForce GTX 580 (15,283, 0% delta), the Radeon 680M (15,270, 0.1% ahead), the RTX 3050 OEM (15,199, 0.6% behind), and the RX 7600 (15,171, 0.8% behind). This clustering suggests the RTX 2060’s average score is representative of a tight performance tier. The Radeon 660M, by contrast, sits near the RTX A2000 Mobile (13,821, 0.1% behind), the RX 570X (13,871, 0.4% behind), the RX 7900 XT (13,745, 0.5% ahead), and the Tesla K10 (14,029, 1.5% behind). The Radeon 660M’s nearest rivals are generally mobile or older desktop parts, indicating its performance class is lower than the RTX 2060’s.

Architecture Differences

The two GPUs come from fundamentally different design philosophies. The RTX 2060 uses the TU106 chip built on Turing architecture, manufactured by TSMC on a 12 nm process. It integrates 10,800 million transistors on a 445 mm² die, yielding a transistor density of 24.3 million per mm². The Radeon 660M uses the Rembrandt chip based on RDNA 2.0, also from TSMC, but on a much smaller 6 nm node. It packs 13,100 million transistors into a 208 mm² die, achieving a density of 63.0 million per mm². The Radeon 660M’s higher density reflects its newer process, but its smaller die area and lower absolute transistor count relative to the RTX 2060’s die size indicate different design targets.

Core counts diverge sharply. The RTX 2060 has 1,920 shading units, 120 texture mapping units, and 48 raster output units. It also includes 30 ray tracing cores and 240 tensor cores, features absent from the Radeon 660M’s spec sheet (tensor cores are listed as null). The Radeon 660M has 384 shading units, 24 TMUs, and 16 ROPs, with 6 ray tracing cores. The RTX 2060’s raw shader count is five times higher, which directly explains its massive lead in the compute benchmarks.

Memory architecture is another major split. The RTX 2060 uses 6 GB of GDDR6 memory on a 192-bit bus, delivering 336.0 GB/s of bandwidth. The Radeon 660M has no dedicated memory; it uses System Shared memory, with bus width and bandwidth listed as System Dependent. This means the Radeon 660M’s performance is heavily influenced by the host system’s RAM speed and configuration, whereas the RTX 2060 has fixed, dedicated bandwidth. The RTX 2060’s memory clock is 1750 MHz (14 Gbps effective), while the Radeon 660M’s memory clock is simply "System Shared".

Clock speeds also differ. The RTX 2060 runs at a base of 1365 MHz and boosts to 1680 MHz. The Radeon 660M has a higher base of 1500 MHz and a boost of 1900 MHz. Despite the Radeon 660M’s higher clocks, the RTX 2060’s far larger execution resources dominate. Pixel rate for the RTX 2060 is 80.64 GPixel/s versus 30.40 GPixel/s for the Radeon 660M. Texture rate is 201.6 GTexel/s versus 45.60 GTexel/s. FP32 throughput is 6.451 TFLOPS for the RTX 2060 and 1,459.2 GFLOPS (1.4592 TFLOPS) for the Radeon 660M. FP16 rates are 12.90 TFLOPS and 2.918 TFLOPS respectively, both at a 2:1 ratio.

Where Each One Wins

The RTX 2060 wins in every scenario that relies on raw compute throughput, dedicated memory bandwidth, or sustained desktop-class performance. Its 6 GB of GDDR6 memory and 336.0 GB/s bandwidth make it suitable for high-resolution textures and bandwidth-intensive workloads. The 240 tensor cores provide hardware acceleration for AI inference and deep learning tasks, which the Radeon 660M cannot match. The RTX 2060’s 30 ray tracing cores, combined with its higher pixel and texture rates, give it a clear edge in ray-traced gaming and content creation applications.

The Radeon 660M wins in scenarios where power efficiency and integration are priorities. Its 40 W TDP is a fraction of the RTX 2060’s 160 W, and it requires no power connectors, drawing power from the host system. It is an IGP (integrated graphics processor) with no slot width, meaning it occupies no expansion slot and is portable-device dependent for display outputs. This makes it suitable for thin-and-light laptops where space and thermals are constrained. Its PCIe 4.0 x8 interface is newer than the RTX 2060’s PCIe 3.0 x16, which could reduce latency in system-shared memory configurations. The Radeon 660M’s higher boost clock of 1900 MHz also suggests it can scale with adequate cooling, but the benchmark data does not support any performance win over the RTX 2060.

The data shows the RTX 2060 wins 2 out of 2 head-to-head tests. The Radeon 660M has no recorded wins. For users needing a discrete GPU with dedicated VRAM and high compute throughput, the RTX 2060 is the only viable choice. For users needing an iGPU with low power draw and no additional hardware, the Radeon 660M is the only option, but it sacrifices performance.

FAQ

Q: How much faster is the NVIDIA GeForce RTX 2060 in Geekbench OpenCL?

A: The RTX 2060 scores 65,014 points compared to the Radeon 660M’s 12,876 points, a 404.9% advantage.

Q: Does the AMD Radeon 660M have any benchmark win over the RTX 2060?

A: No. In the head-to-head tests, the RTX 2060 wins 2 tests and the Radeon 660M wins 0.

Q: What is the memory configuration of each GPU?

A: The RTX 2060 has 6 GB of GDDR6 memory on a 192-bit bus with 336.0 GB/s bandwidth. The Radeon 660M uses System Shared memory with no dedicated VRAM, and its bandwidth is System Dependent.

Q: Which GPU has a higher transistor density?

A: The AMD Radeon 660M has a transistor density of 63.0 million per mm², while the NVIDIA GeForce RTX 2060 has 24.3 million per mm².

Q: What are the power requirements for each card?

A: The RTX 2060 has a TDP of 160 W and requires a 1x 8-pin power connector with a suggested PSU of 450 W. The Radeon 660M has a TDP of 40 W and requires no power connectors.

Q: How do their average benchmark scores compare?

A: The RTX 2060 has an average benchmark score of 15,290, which is 1,478 points higher than the Radeon 660M’s average of 13,812.

Specification Differences

• Architecture: NVIDIA Turing (TU106) vs AMD RDNA 2.0 (Rembrandt)
• Process Node: 12 nm vs 6 nm (both TSMC)
• Transistors: 10,800 million vs 13,100 million
• Die Size: 445 mm² vs 208 mm²
• Transistor Density: 24.3M / mm² vs 63.0M / mm²
• Base Clock: 1365 MHz vs 1500 MHz
• Boost Clock: 1680 MHz vs 1900 MHz
• Memory: 6 GB GDDR6 vs System Shared
• Memory Bus Width: 192 bit vs System Shared
• Memory Bandwidth: 336.0 GB/s vs System Dependent
• Shading Units: 1920 vs 384
• TMUs: 120 vs 24
• ROPs: 48 vs 16
• RT Cores: 30 vs 6
• Tensor Cores: 240 vs null
• Pixel Rate: 80.64 GPixel/s vs 30.40 GPixel/s
• Texture Rate: 201.6 GTexel/s vs 45.60 GTexel/s
• FP32: 6.451 TFLOPS vs 1,459.2 GFLOPS
• FP16: 12.90 TFLOPS vs 2.918 TFLOPS
• TDP: 160 W vs 40 W
• Slot Width: Dual-slot vs IGP
• Power Connectors: 1x 8-pin vs None
• Suggested PSU: 450 W vs null
• Bus Interface: PCIe 3.0 x16 vs PCIe 4.0 x8
• Display Outputs: 1x DVI, 1x HDMI 2.0, 2x DisplayPort 1.4a, 1x USB Type-C vs Portable Device Dependent
• Release Date: 2019-01-06 vs 2022-01-03
• Predecessor: GeForce 10 vs Vega II IGP
• Successor: GeForce 30 vs Navi III IGP
• Launch MSRP: 349 USD vs null

The Verdict

The data is unambiguous: the NVIDIA GeForce RTX 2060 is the superior performer. It wins both shared benchmarks by margins exceeding 340%, holds a higher average benchmark score (15,290 vs 13,812), and ranks higher in the overall percentile distribution (58th vs 55th). The RTX 2060’s dedicated 6 GB GDDR6 memory, 336.0 GB/s bandwidth, and 1920 shading units provide a level of compute throughput that the Radeon 660M cannot approach with its 384 shading units and system-shared memory.

The AMD Radeon 660M is the appropriate choice only for systems where power consumption is critical. Its 40 W TDP, no power connectors, and IGP form factor make it suitable for portable devices where a discrete GPU is impossible. But the benchmark results show a 404.9% gap in OpenCL and a 346.5% gap in Vulkan, meaning the Radeon 660M is not a substitute for the RTX 2060 in any performance-sensitive workload. The RTX 2060’s end-of-life status and 349 USD launch MSRP are the only caveats, but the recorded data shows it remains a far more capable part. Buyers needing GPU compute should pick the RTX 2060; buyers needing an integrated solution with minimal power draw should pick the Radeon 660M, but they should expect a substantial performance penalty.