AMD Radeon 660M vs NVIDIA GeForce RTX 3060 Ti Comparison

FAQ

Q: How does the NVIDIA GeForce RTX 3060 Ti compare to the AMD Radeon 660M in Geekbench OpenCL performance?

A: The RTX 3060 Ti scores 78,927 in Geekbench OpenCL, while the Radeon 660M scores 12,876. This gives the RTX 3060 Ti a 513% advantage, the largest single-test gap between the two.

Q: Which GPU has the higher percentile ranking among all GPUs?

A: The RTX 3060 Ti sits in the 59th percentile of all GPUs, while the Radeon 660M sits in the 55th percentile. The NVIDIA part holds a modest overall ranking advantage despite its much larger benchmark lead.

Q: What are the average benchmark scores for each GPU?

A: The RTX 3060 Ti has an average benchmark score of 16,129, while the Radeon 660M has an average of 13,812. The RTX 3060 Ti is ahead by roughly 17% in average score.

Q: Which GPU has the higher boost clock?

A: The AMD Radeon 660M has a boost clock of 1900 MHz, while the NVIDIA GeForce RTX 3060 Ti has a boost clock of 1665 MHz. The AMD part runs at a higher boost frequency.

Q: Do both GPUs support DirectX 12 Ultimate?

A: Yes, both the RTX 3060 Ti and the Radeon 660M support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Q: Which GPU has more shading units?

A: The RTX 3060 Ti has 4,864 shading units, while the Radeon 660M has 384 shading units. This is a 12.7x difference in raw shading unit count.

Architecture Differences

The NVIDIA GeForce RTX 3060 Ti is built on the GA104 chip using the Ampere architecture, manufactured on an 8 nm process at Samsung. The AMD Radeon 660M uses the Rembrandt chip with RDNA 2.0 architecture, manufactured on a 6 nm process at TSMC. Both are end-of-life products, but they target entirely different market segments, one as a discrete add-in card and the other as an integrated graphics processor.

The RTX 3060 Ti packs 17,400 million transistors into a 392 mm² die, giving it a transistor density of 44.4M per mm². The Radeon 660M contains 13,100 million transistors on a smaller 208 mm² die, achieving a higher density of 63.0M per mm². The AMD part's smaller die and tighter packing reflect its integrated nature and lower power envelope.

The RTX 3060 Ti includes 152 tensor cores alongside 38 RT cores, while the Radeon 660M has no tensor core equivalent but includes 6 RT cores. The NVIDIA GPU features 152 texture mapping units and 80 render output units, compared to 24 TMUs and 16 ROPs on the AMD part. These differences translate directly into compute throughput, with the RTX 3060 Ti delivering 16.20 TFLOPS FP32 versus 1,459.2 GFLOPS on the Radeon 660M.

Memory architecture diverges sharply. The RTX 3060 Ti has 8 GB of GDDR6 on a 256-bit bus, delivering 448.0 GB/s of bandwidth. The Radeon 660M relies on system shared memory, with bandwidth marked as system dependent. This makes the NVIDIA card suitable for data-heavy workloads, while the AMD IGP depends entirely on the host system's memory configuration.

The power envelope is equally distinct. The RTX 3060 Ti has a 200 W TDP with a dual-slot cooler and a single 12-pin power connector, plus a suggested 550 W PSU. The Radeon 660M draws only 40 W, uses no power connectors, and is classified as an IGP. The NVIDIA part requires dedicated power delivery, while the AMD part is designed for portable devices with no discrete power connection.

Head-to-Head Benchmarks

Only two benchmarks in the database directly compare these GPUs, and both show decisive NVIDIA wins. In Geekbench OpenCL, the RTX 3060 Ti scores 78,927 against 12,876 for the Radeon 660M. This 513% delta is the largest performance gap recorded between the two parts. The OpenCL test stresses general compute workloads, where the RTX 3060 Ti's 4,864 shading units and dedicated memory bandwidth give it an overwhelming advantage.

Geekbench Vulkan shows a similar pattern but a narrower margin. The RTX 3060 Ti scores 47,784, while the Radeon 660M scores 14,748, a 224% difference. Vulkan workloads benefit from lower-level API access, but the raw compute resources of the NVIDIA card still dominate. The RTX 3060 Ti wins both head-to-head tests, giving it 2 wins and 0 losses in the database.

Looking at the nearest rivals for each GPU places these scores in context. The RTX 3060 Ti's average benchmark score of 16,129 is 0.7% ahead of the AMD Radeon RX 9060 at 16,014, and 1.4% behind the AMD Radeon RX 5700 XT at 16,361. The Radeon 660M's average of 13,812 is nearly identical to the NVIDIA RTX A2000 Mobile at 13,821, just 0.1% behind, and 0.4% behind the AMD Radeon RX 570X at 13,871. The Radeon 660M also sits 0.5% ahead of the AMD Radeon RX 7900 XT at 13,745, a curious data point that highlights how average scores can compress differences across very different hardware.

The RTX 3060 Ti's individual benchmark scores show consistency across API generations. It scores 20,349 in Passmark G3D, 10,006 in GPU compute, 989 in G2D, and 234 in DirectX 9. Lower scores in DirectX 10 (132), DirectX 11 (163), and DirectX 12 (78) reflect the Passmark test suite's weighting, not necessarily raw capability. The Radeon 660M has no comparable Passmark entries in the database, limiting direct comparison to the Geekbench tests.

Specification Differences

The two GPUs differ in nearly every measurable specification. The RTX 3060 Ti uses an 8 nm process at Samsung, while the Radeon 660M uses a 6 nm process at TSMC. Transistor counts are 17,400 million versus 13,100 million, and die sizes are 392 mm² versus 208 mm². Transistor density is higher on the AMD part at 63.0M per mm², versus 44.4M per mm² for NVIDIA.

Base clocks are 1410 MHz for the RTX 3060 Ti and 1500 MHz for the Radeon 660M, with boost clocks of 1665 MHz and 1900 MHz respectively. The AMD part runs faster at both points, but this does not compensate for the massive gap in compute resources. The RTX 3060 Ti has 4,864 shading units, 152 TMUs, 80 ROPs, 38 RT cores, and 152 tensor cores. The Radeon 660M has 384 shading units, 24 TMUs, 16 ROPs, and 6 RT cores, with no tensor cores listed.

Memory specifications are fundamentally different. The RTX 3060 Ti has 8 GB GDDR6 on a 256-bit bus with 448.0 GB/s bandwidth. The Radeon 660M uses system shared memory with system dependent bandwidth. The NVIDIA part has a fixed memory pipeline; the AMD part's performance depends on the host system's memory configuration.

Pixel rate is 133.2 GPixel/s for the RTX 3060 Ti versus 30.40 GPixel/s for the Radeon 660M, a 4.4x difference. Texture rate is 253.1 GTexel/s versus 45.60 GTexel/s, a 5.5x difference. FP32 throughput is 16.20 TFLOPS versus 1,459.2 GFLOPS. FP16 rates are 16.20 TFLOPS (1:1) for NVIDIA versus 2.918 TFLOPS (2:1) for AMD.

The RTX 3060 Ti has a 200 W TDP, dual-slot design, a 12-pin power connector, and a suggested 550 W PSU. The Radeon 660M has a 40 W TDP, IGP form factor, no power connectors, and no suggested PSU. The NVIDIA card measures 242 mm in length and 112 mm in height. The AMD part has no recorded dimensions, reflecting its integration into a host device.

The RTX 3060 Ti uses PCIe 4.0 x16, while the Radeon 660M uses PCIe 4.0 x8. Display outputs are 1x HDMI 2.1 and 3x DisplayPort 1.4a for NVIDIA, while the AMD part is portable device dependent. The NVIDIA card was released on 2020-11-30, the AMD part on 2022-01-03. The RTX 3060 Ti has a launch MSRP of 399 USD; the Radeon 660M has no listed MSRP.

Where Each One Wins

The RTX 3060 Ti wins in every benchmark category where both are tested. Its 513% lead in OpenCL and 224% lead in Vulkan indicate total dominance in compute-oriented workloads. The NVIDIA card's dedicated 8 GB GDDR6 memory with 448.0 GB/s bandwidth makes it suitable for tasks that require high memory throughput, such as large texture loads, complex shader work, and data-parallel compute. Its 16.20 TFLOPS FP32 throughput positions it for demanding graphics applications, including high-resolution gaming and GPU-accelerated content creation.

The Radeon 660M wins in power efficiency and physical integration. At 40 W TDP, it consumes one-fifth the power of the RTX 3060 Ti's 200 W. It requires no power connector and fits into portable devices as an IGP, making it the only choice for systems that cannot accommodate a discrete card. Its 6 nm process gives it a higher transistor density, suggesting better power-per-transistor characteristics, though the database does not include efficiency benchmarks to confirm this.

For desktop users with a PCIe x16 slot and a 550 W PSU, the RTX 3060 Ti is the clear choice for raw performance. For laptop or compact system builders where size, power, and thermal constraints dominate, the Radeon 660M is the only viable option among these two. The RTX 3060 Ti also has a fixed memory configuration that cannot be changed, while the Radeon 660M's performance scales with system memory, a variable the database does not capture.

The Verdict

From the recorded data, the NVIDIA GeForce RTX 3060 Ti is the superior GPU for any workload that prioritizes compute and graphics performance. Its average benchmark score of 16,129 is 16.8% higher than the Radeon 660M's 13,812, and its head-to-head wins are decisive at 513% and 224% margins. The RTX 3060 Ti also holds a higher percentile ranking at 59 versus 55 for the AMD part.

However, the comparison is not apples to apples. The RTX 3060 Ti is a discrete desktop card with a 200 W TDP, a 242 mm length, and a dedicated 8 GB GDDR6 buffer. The Radeon 660M is an integrated processor designed for portable devices, drawing 40 W with no power connector and relying on system shared memory. A buyer choosing between these two is not choosing between similar products; they are choosing between a full-size desktop solution and a mobile IGP.

For users with a desktop chassis, a 550 W PSU, and a PCIe 4.0 x16 slot, the RTX 3060 Ti is the only reasonable pick based on the benchmark data. Its performance ceiling is far higher, and its memory bandwidth is fixed and generous. For users constrained to portable devices or low-power systems, the Radeon 660M is the only option, and its 55th percentile ranking shows it is not a weak performer in its class, sitting within 0.5% of the RX 7900 XT in average score.

The database shows 2 wins for the RTX 3060 Ti and 0 for the Radeon 660M. That record, combined with the average score and percentile differences, leads to a clear verdict: the RTX 3060 Ti wins on raw performance, while the Radeon 660M wins on integration and power efficiency. Choose based on your hardware constraints, not on benchmark numbers alone, because no benchmark can erase the physical difference between a dual-slot card and an IGP.