NVIDIA GeForce GTX 1660 Ti vs NVIDIA GeForce MX330 Comparison

# NVIDIA GeForce MX330 vs NVIDIA GeForce GTX 1660 Ti

The data presents a fascinating contrast: the MX330, a 10-watt integrated-class part from the Pascal era, and the GTX 1660 Ti, a 120-watt Turing-based discrete GPU, occupy wildly different performance tiers despite sharing the NVIDIA brand. While the GTX 1660 Ti dominates in raw compute, the MX330 posts a surprising victory in one specific benchmark, making this comparison far more nuanced than a simple generational gap.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The MX330 has a higher average benchmark score of 8,458, compared to the GTX 1660 Ti’s 7,723. This is counterintuitive given the GTX 1660 Ti’s much larger die and memory subsystem, and it stems from the specific tests included in each GPU’s benchmark set.

Q: How do the two compare in OpenCL performance?

A: The GTX 1660 Ti is dramatically faster, scoring 51,029 in Geekbench OpenCL versus the MX330’s 7,896. That is a delta of -84.5% from the MX330’s perspective, meaning the GTX 1660 Ti is roughly 6.5 times faster in this workload.

Q: Does the MX330 win any benchmark outright?

A: Yes, the MX330 wins the Geekbench Vulkan test by a wide margin, scoring 9,019 against the GTX 1660 Ti’s 5,827. This represents a +54.8% delta in favor of the MX330, a surprising result given the architectural differences.

Q: What is the memory configuration difference?

A: The GTX 1660 Ti ships with 6 GB of GDDR6 memory on a 192-bit bus, delivering 288.0 GB/s bandwidth. The MX330 has just 2 GB of GDDR5 on a 64-bit bus, yielding 56.06 GB/s — a five-fold bandwidth gap.

Q: Which GPU has a higher transistor density?

A: The MX330’s GP108B chip has a transistor density of 24.3M / mm², slightly higher than the GTX 1660 Ti’s TU116 at 23.2M / mm², despite the former being fabricated on a 14 nm Samsung process versus the latter’s 12 nm TSMC node.

Q: What are the percentile rankings?

A: The MX330 sits at the 43rd percentile of all GPUs, while the GTX 1660 Ti ranks at the 41st percentile. Both are mid-pack performers, but the MX330’s higher percentile reflects its stronger showing in the limited benchmark set available.

Architecture Differences

The architectural gap between these two is generational. The MX330 uses the GP108B chip built on Pascal, a 14 nm design from Samsung, while the GTX 1660 Ti employs the TU116 chip on Turing, fabricated by TSMC at 12 nm. The transistor counts tell the scale of the difference: the GTX 1660 Ti packs 6,600 million transistors across a 284 mm² die, versus just 1,800 million on a 74 mm² die for the MX330. Interestingly, the MX330 has a slightly higher transistor density (24.3M / mm² vs 23.2M / mm²), but that is a reflection of its simpler design rather than a technological advantage.

The compute resources are vastly different. The GTX 1660 Ti fields 1,536 shading units, 96 texture mapping units, and 48 ROPs. The MX330 has only 384 shading units, 24 TMUs, and 16 ROPs. This means the GTX 1660 Ti has four times the shading and TMU count, and three times the ROP count. The FP32 compute rating confirms this: 5.437 TFLOPS for the GTX 1660 Ti versus 1,224.2 GFLOPS for the MX330 — a 4.4x advantage.

The FP16 story is equally lopsided but reveals different design philosophies. The GTX 1660 Ti delivers 10.87 TFLOPS with a 2:1 FP16 ratio, indicating a more balanced throughput. The MX330 manages only 19.13 GFLOPS at a 1:64 ratio, meaning its FP16 performance is negligible — a clear sign that Pascal’s consumer silicon was never optimized for half-precision workloads.

Memory architecture amplifies the gap. The GTX 1660 Ti’s 6 GB GDDR6 on a 192-bit bus at 12 Gbps effective yields 288.0 GB/s. The MX330’s 2 GB GDDR5 on a 64-bit bus at 7 Gbps effective produces just 56.06 GB/s. That 5.1x bandwidth difference is critical for any texture-heavy or large-frame-buffer workload. The GTX 1660 Ti also uses a PCIe 3.0 x16 interface, while the MX330 is limited to x4, further reducing data transfer capacity.

Power and physical design reflect their intended roles. The MX330 is an IGP (integrated graphics processor) with a 10 W TDP, no power connectors, and portable-device-dependent outputs. The GTX 1660 Ti is a dual-slot card drawing 120 W, requiring a single 8-pin connector and a 300 W suggested PSU, with output options including 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a.

Head-to-Head Benchmarks

The direct head-to-head data is sparse but revealing. Two tests were run on both GPUs: Geekbench OpenCL and Geekbench Vulkan. The OpenCL result is a landslide — the GTX 1660 Ti scores 51,029 against the MX330’s 7,896, a -84.5% delta. This aligns perfectly with the raw compute and memory advantages, confirming that in compute-heavy OpenCL workloads, the GTX 1660 Ti is in a completely different league.

The Vulkan result flips the script. The MX330 scores 9,019, beating the GTX 1660 Ti’s 5,827 by +54.8%. This is a strange outlier. Vulkan is a low-overhead API that often favors driver efficiency and command buffer handling over raw compute. The MX330’s older Pascal architecture, paired with a simpler memory subsystem, may exhibit lower latency in certain Vulkan dispatch patterns. It could also be that the GTX 1660 Ti’s Turing drivers are not optimized for this specific workload, or that the test itself is bound by a bottleneck the MX330 avoids. The data does not explain the mechanism, but it does show that the GTX 1660 Ti is not universally superior.

Looking at the broader benchmark portfolio, the GTX 1660 Ti has a much richer dataset: Passmark scores across DirectX 9, 10, 11, 12, plus G2D, G3D, and GPU compute, alongside a 3DMark Steel Nomad DX12 result. The MX330 only has the two Geekbench entries. This disparity means the MX330’s average benchmark score of 8,458 is based on a narrower set, while the GTX 1660 Ti’s 7,723 average is diluted by multiple Passmark sub-tests that may be less representative of gaming performance. The GTX 1660 Ti’s Passmark G3D score of 12,871 is strong, and its GPU compute score of 5,226 suggests solid compute throughput. But the DirectX 9 score of 190 and DirectX 12 score of 51 are oddly low, potentially dragging down the average.

Specification Differences

The two GPUs differ in nearly every measurable specification. The process node is 14 nm (Samsung) for the MX330 versus 12 nm (TSMC) for the GTX 1660 Ti. Transistor count is 1,800 million versus 6,600 million, and die size is 74 mm² versus 284 mm². The MX330’s base clock is 1531 MHz with a boost of 1594 MHz, while the GTX 1660 Ti runs a lower base of 1500 MHz but boosts higher to 1770 MHz. Memory clocks are 1752 MHz (7 Gbps effective) for the MX330 and 1500 MHz (12 Gbps effective) for the GTX 1660 Ti — the latter’s GDDR6 achieves higher bandwidth despite a lower clock.

Memory size, type, bus width, and bandwidth all favor the GTX 1660 Ti: 6 GB GDDR6, 192-bit, 288.0 GB/s versus 2 GB GDDR5, 64-bit, 56.06 GB/s. Shading units, TMUs, and ROPs are 384/24/16 for the MX330 and 1536/96/48 for the GTX 1660 Ti. Pixel rate is 25.50 GPixel/s versus 84.96 GPixel/s, and texture rate is 38.26 GTexel/s versus 169.9 GTexel/s. FP32 is 1,224.2 GFLOPS versus 5.437 TFLOPS. FP16 is 19.13 GFLOPS (1:64) versus 10.87 TFLOPS (2:1).

TDP is 10 W versus 120 W. Slot width is IGP versus dual-slot. Power connectors are none versus 1x 8-pin. Suggested PSU is absent for the MX330 but 300 W for the GTX 1660 Ti. Bus interface is PCIe 3.0 x4 versus x16. Display outputs are portable-device-dependent versus 1x DVI, 1x HDMI 2.0, 1x DisplayPort 1.4a. Dimensions are not listed for the MX330, but the GTX 1660 Ti measures 229 mm (9 inches) in length, 111 mm (4.4 inches) in height, and 35 mm (1.4 inches) in width. The MX330’s architecture is Pascal, while the GTX 1660 Ti is Turing. The GTX 1660 Ti’s series is GeForce 10-series, but its generation is listed as GeForce 16, with a predecessor of GeForce 10 and successor of GeForce 20.

Where Each One Wins

The GTX 1660 Ti wins decisively in raw compute throughput. Its OpenCL score is 6.5 times higher, its FP32 is 4.4 times higher, and its memory bandwidth is 5.1 times higher. Any workload that exercises shader math, texture filtering, or large data sets will favor the GTX 1660 Ti. The pixel and texture rates reinforce this: 84.96 GPixel/s and 169.9 GTexel/s versus 25.50 GPixel/s and 38.26 GTexel/s. For gaming at higher resolutions or with modern API features, the GTX 1660 Ti is the clear choice. Its Passmark G3D score of 12,871 indicates strong DirectX 11-era game performance, and the 3DMark Steel Nomad DX12 score of 1,061 suggests it can handle newer titles, albeit not at high settings.

The MX330 wins in the Vulkan benchmark, and that is its only direct head-to-head victory. It also carries a higher transistor density and a lower TDP, which makes it suitable for thin-and-light laptops where power and heat are constrained. Its percentile rank of 43 is slightly better than the GTX 1660 Ti’s 41, but this is likely an artifact of the benchmark set rather than a true performance advantage. In terms of nearest rivals, the MX330 sits near the AMD Radeon HD 8870M (delta 0%), AMD Radeon 880M (delta 0.3%), and NVIDIA GeForce GTX 675MX (delta 0.4%), while the GTX 1660 Ti’s rivals are the AMD Radeon 540 (delta 0.7%), AMD Radeon Pro WX 3100 (delta 1.9%), and AMD Radeon R7 250 (delta 2.2%). These rival clusters suggest the MX330 is competing with older mobile parts, while the GTX 1660 Ti’s nearest neighbors are low-end desktop GPUs.

The Verdict

The data is unambiguous for compute-heavy tasks: the GTX 1660 Ti is the superior GPU. Its 6 GB of GDDR6 memory, 288.0 GB/s bandwidth, and 5.437 TFLOPS FP32 make it a capable card for gaming and content creation. The MX330, with 2 GB of GDDR5 and 1,224.2 GFLOPS, is a basic solution for light workloads. The GTX 1660 Ti’s launch MSRP is 279 USD, which reflects its intended position as a mid-range discrete card, though the MX330 has no listed launch MSRP.

However, the Vulkan result complicates the narrative. The MX330’s +54.8% win in Geekbench Vulkan suggests that for certain low-level API workloads, the older Pascal architecture can outperform Turing. This is not a recommendation to choose the MX330 for gaming — its compute and memory limits are too severe — but it does raise questions about driver maturity and workload-specific optimization on the GTX 1660 Ti.

For a buyer seeking a discrete GPU for modern gaming or compute, the GTX 1660 Ti is the only rational choice from this data. For a buyer constrained by power envelopes — such as an ultraportable laptop user — the MX330’s 10 W TDP and IGP form factor make it the only viable option. The GTX 1660 Ti’s 120 W TDP and dual-slot footprint are simply not portable. The verdict hinges on use case: choose the GTX 1660 Ti for performance, the MX330 for efficiency and portability. The benchmark data does not support any other conclusion.