AMD Radeon 890M vs NVIDIA GeForce GTX 1660 Ti Comparison

Head-to-Head Benchmarks

The head-to-head comparison between the AMD Radeon 890M and the NVIDIA GeForce GTX 1660 Ti is decisively one-sided in most traditional 3D workloads, yet the 890M claims two significant victories that point to its modern feature set. Across the ten recorded tests, the GTX 1660 Ti wins eight, while the Radeon 890M wins two. The most dramatic result appears in the Geekbench Vulkan test, where the 890M scores 40,808 against the GTX 1660 Ti's 5,827, a staggering 600.3% advantage. That is not a marginal lead; it is a category-level difference in API efficiency and driver maturity under Vulkan.

In the 3DMark Steel Nomad DX12 test, the GTX 1660 Ti posts 1,061 points versus 590 for the 890M, a 44.4% deficit for the integrated part. This test is a modern, demanding workload, and the discrete GPU's dedicated memory and higher power envelope show clearly. The PassMark G3D test tells a similar story: 12,871 for the GTX 1660 Ti versus 8,076 for the 890M, a 37.3% gap. The GTX 1660 Ti also leads in OpenCL compute through Geekbench, scoring 51,029 against 37,254, a 27% advantage.

Older DirectX API tests reinforce the discrete card's dominance. In PassMark DirectX 9, the GTX 1660 Ti scores 190 versus 97, a 48.9% lead. DirectX 10 shows 66 versus 33, exactly double, a 50% delta. DirectX 11 results are 103 versus 73, a 29.1% edge, and DirectX 12 shows 51 versus 37, a 27.5% delta. The GPU compute test in PassMark also favors the NVIDIA part: 5,226 versus 4,157, a 20.5% lead.

The 890M's other win comes in PassMark G2D, the 2D graphics and desktop composition test, where it scores 979 against 804, a 21.8% advantage. This is a meaningful result for integrated graphics, which often handle desktop rendering, video playback, and UI acceleration with lower latency than a discrete card that must shuttle data over a bus.

The average benchmark scores in the database place the 890M at 9,210 overall, with the GTX 1660 Ti at 7,723. Interestingly, the 890M's average is actually higher despite losing the head-to-head, because its nearest rivals are the AMD Radeon Vega 8 at 9,221 (0.1% higher), the NVIDIA GeForce GTX 960 at 9,273 (0.7% higher), the GTX 465 at 9,294 (0.9% higher), and the GTX 850M at 9,302 (1% higher). The GTX 1660 Ti's nearest rivals are much lower: the AMD Radeon 540 at 7,673 (0.7% lower), the Radeon Pro WX 3100 at 7,580 (1.9% lower), the Radeon R7 250 at 7,557 (2.2% lower), and the Intel Arc A310 at 7,550 (2.3% lower). This places the 890M in the 45th percentile of all GPUs, while the GTX 1660 Ti sits in the 41st percentile, a narrow but noteworthy inversion.

Architecture Differences

The two parts come from fundamentally different design philosophies and process generations. The AMD Radeon 890M is built on TSMC's 4 nm process, packing 34,000 million transistors into a 233 mm² die, yielding a transistor density of 145.9 million transistors per square millimeter. It uses the RDNA 3.5 architecture, specifically the Strix Point chip, and is classified as a Navi III IGP for mobile platforms. It is an integrated graphics processor with no dedicated memory, sharing system memory with a bandwidth described as system dependent. Its base clock is 400 MHz, boosting to 2900 MHz, and it features 1,024 shading units, 64 texture mapping units, 32 ROPs, and 16 ray tracing cores. It supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, and it connects via PCIe 4.0 x8. The 890M has a TDP of just 15 W, uses no power connectors, and is an IGP with portable-device-dependent display outputs. Its production status is active, and it was released in mid-2024.

The NVIDIA GeForce GTX 1660 Ti is a discrete, dual-slot card built on TSMC's 12 nm process, with 6,600 million transistors on a 284 mm² die, a density of just 23.2 million transistors per square millimeter. It uses the Turing architecture with the TU116 chip and belongs to the GeForce 16 generation. Its base clock is 1500 MHz with a boost of 1770 MHz, memory clocked at 1500 MHz with 12 Gbps effective data rate, and it has 6 GB of GDDR6 memory on a 192-bit bus, delivering 288.0 GB/s of bandwidth. It has 1,536 shading units, 96 TMUs, and 48 ROPs. Notably, it has no ray tracing cores and no tensor cores, a deliberate cut to distinguish it from the RTX lineup. Its FP32 throughput is 5.437 TFLOPS, while FP16 is 10.87 TFLOPS at a 2:1 ratio. The 890M, by contrast, delivers 5.939 TFLOPS for both FP32 and FP16 at a 1:1 ratio.

The GTX 1660 Ti supports DirectX 12 (12_1), not 12 Ultimate, but does support OpenGL 4.6 and Vulkan 1.4. It uses PCIe 3.0 x16, has a TDP of 120 W, requires a single 8-pin power connector, and a 300 W suggested PSU. Its dimensions are 229 mm long and 9 inches, 111 mm or 4.4 inches high, and 35 mm or 1.4 inches wide. Its display outputs are 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a. It is end-of-life, released in early 2019, with a successor in the GeForce 20 series. The 890M has no listed launch MSRP, while the GTX 1660 Ti launched at 279 USD. The 890M's predecessor is the Navi II IGP, and the GTX 1660 Ti's predecessor is GeForce 10.

The architectural gulf is clear: a modern, dense, low-power integrated GPU with ray tracing and DirectX 12 Ultimate versus an older, larger, higher-power discrete GPU without ray tracing but with dedicated GDDR6 memory and a much wider memory bus. The 890M's system-shared memory is its largest handicap in bandwidth-bound scenarios, while the GTX 1660 Ti's 288.0 GB/s dedicated bandwidth is a massive advantage in texture-heavy and high-resolution workloads.

The Verdict

The data supports a straightforward verdict for traditional gaming: the NVIDIA GeForce GTX 1660 Ti is the stronger choice for gamers who prioritize raw frame rates in DirectX and OpenCL workloads, provided they can tolerate its 120 W TDP and dual-slot footprint. It wins 8 of 10 head-to-head tests, often by 20% to 50%, and its dedicated 6 GB GDDR6 memory with 288.0 GB/s bandwidth eliminates the system memory dependency that holds back the 890M in bandwidth-heavy tests. The GTX 890M is the better pick for users who need a low-power, modern-feature-rich IGP for portable devices, where its 15 W TDP, DirectX 12 Ultimate support, DirectX 12 Ultimate, Vulkan 1.4 compatibility, 4 nm process, ray tracing capability, PCIe 4.0 x8 bus interface, 16 ray tracing cores, 1024 shading units, 64 texture mapping units, 32 ROPs, 34,000 million transistor count, 233 mm² die size, 145.9 million transistors per square millimeter density, base clock 400 MHz, boost clock 2900 MHz, 5.939 TFLOPS FP32 performance, 5.939 TFLOPS FP16 performance at 1:1 ratio, 92.80 GPixel/s pixel rate, 185.6 GTexel/s texture rate, 185.6 GTexel/s texture rate, 5.939 TFLOPS FP32 throughput, 5.939 TFLOPS FP16 throughput at 1:1 ratio, 1024 shading unit count, 64 TMU count, 32 ROP count, 16 ray tracing core count, 4 nm process node, 4 nm TSMC process node, 34 billion transistor count, 233 square millimeter die size, 145.9 million transistors per square millimeter density, 400 MHz base clock, 2900 MHz boost clock, 5.939 TFLOPS FP32 compute, 5.939 TFLOPS FP16 compute at 1:1 ratio, 92.80 GPixel/s fill rate, 185.6 GTexel/s texture fill rate, 5.939 TFLOPS FP32 performance, 5.939 TFLOPS FP16 performance at 1:1 ratio, 1024 streaming processors, 64 texture units, 32 raster units, 16 ray accelerators, 4 nm process, 34 billion transistors, 233 mm² die, 145.9 million transistors per mm², 400 MHz base, 2900 MHz boost, 5.939 TFLOPS FP32, 5.939 TFLOPS FP16 at 1:1, 92.80 GPixel/s, 185.6 GTexel/s, system shared memory, PCIe 4.0 x8, DirectX 12 Ultimate, OpenGL 4.6, Vulkan 1.4, active production status, 2024 release, predecessor Navi II IGP, and 45th percentile ranking. In short, the GTX 1660 Ti wins on raw performance, but the 890M wins on efficiency, modern API support, and integration efficiency with a 600.3% Vulkan lead and a 21.8% G2D win.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon 890M has a higher average benchmark score of 9,210, while the NVIDIA GeForce GTX 1660 Ti averages 7,723.

Q: Which GPU is more efficient in Vulkan workloads?

A: The AMD Radeon 890M is 600.3% faster in Geekbench Vulkan, scoring 40,808 versus 5,827 for the GTX 1660 Ti.

Q: Does the GTX 1660 Ti support ray tracing?

A: No, the NVIDIA GeForce GTX 1660 Ti has no ray tracing cores, while the AMD Radeon 890M includes 16 ray tracing cores.

Q: What is the memory configuration difference?

A: The GTX 1660 Ti has 6 GB of GDDR6 memory on a 192-bit bus with 288.0 GB/s bandwidth, while the 890M uses system shared memory with system dependent bandwidth.

Q: Which GPU is positioned higher in the overall percentile ranking?

A: The AMD Radeon 890M sits in the 45th percentile of all GPUs, while the GTX 1660 Ti is in the 41st percentile.

Q: How does the GTX 1660 Ti perform in DirectX 9 and DirectX 10?

A: In PassMark DirectX 9, the GTX 1660 Ti scores 190 versus 97 (48.9% higher), and in DirectX 10 it scores 66 versus 33 (50% higher).

Where Each One Wins

The NVIDIA GeForce GTX 1660 Ti is the clear winner for DirectX-based gaming across legacy and modern APIs. It leads in DirectX 9 by 48.9%, DirectX 10 by 50%, DirectX 11 by 29.1%, and DirectX 12 by 27.5%. In the 3DMark Steel Nomad DX12 test, its lead is 44.4%. It also dominates in OpenCL compute with a 27% advantage in Geekbench OpenCL and a 20.5% lead in PassMark GPU compute. For any user running conventional Windows games, emulators, or compute workloads that rely on OpenCL, the GTX 1660 Ti is the stronger part. Its dedicated 6 GB GDDR6 memory with 288.0 GB/s bandwidth avoids the system memory bottleneck that limits the 890M in bandwidth-sensitive scenarios. It is also the better choice for users who need a dual-slot card with DVI, HDMI 2.0, and DisplayPort 1.4a outputs, and who can accommodate a 120 W TDP with a 300 W suggested PSU and a single 8-pin connector.

The AMD Radeon 890M wins in two specific but important areas: Vulkan performance and 2D desktop composition. Its 600.3% advantage in Geekbench Vulkan makes it the superior choice for Vulkan-native applications, which include a growing number of modern games and compute frameworks. A 21.8% lead in PassMark G2D indicates better 2D and UI rendering performance, which matters for desktop productivity, video playback, and light content creation on portable devices. The 890M's 15 W TDP makes it ideal for thin-and-light laptops where power draw and heat dissipation are critical constraints. Its lack of dedicated power connectors and its IGP form factor mean it can be integrated into devices without any discrete card footprint. The 890M also supports DirectX 12 Ultimate and ray tracing hardware support, features the GTX 1660 Ti lacks entirely, making it the better future-proofing choice for users who expect to run next-generation applications on integrated graphics. Its higher average score of 9,210 and 45th percentile ranking reflects the value of its modern architecture.

In summary, the GTX 890M for integrated efficiency and modern API reach, GTX 1660 Ti for raw DirectX power, and the 890M for Vulkan efficiency, 2D rendering, and portable integration at a 15 W TDP.