NVIDIA GeForce GTX 660 Ti vs NVIDIA RTX A2000 Mobile Comparison

The NVIDIA GeForce GTX 660 Ti and the NVIDIA RTX A2000 Mobile represent two very different eras of GPU design, separated by nearly a decade of architectural evolution. The data shows a clear generational shift, with the mobile Ampere part delivering vastly superior raw compute performance, though the older Kepler card retains a unique legacy in its specific benchmark profile.

Head-to-Head Benchmarks

The benchmark data presents a decisive victory for the RTX A2000 Mobile. In the Geekbench OpenCL test, the A2000 Mobile scores 56,518, while the GTX 660 Ti manages only 15,113. This translates to a 73.3% delta in favor of the newer card. The margin is nearly identical in Geekbench Vulkan, where the A2000 Mobile posts 53,146 against the GTX 660 Ti’s 15,012, a 71.8% difference. These are not incremental gains; they represent a fundamental leap in compute throughput.

The RTX A2000 Mobile’s win count is a clean 2-0 across the two shared benchmarks. However, the data also reveals a curious inconsistency in the A2000 Mobile’s profile. While it dominates in Geekbench compute tests, its Passmark scores are surprisingly low, with a DirectX 10 score of 57, DirectX 11 score of 68, and DirectX 12 score of 47. These figures are far below what its OpenCL and Vulkan numbers suggest, indicating that the A2000 Mobile’s performance is highly workload-dependent. The GTX 660 Ti, by contrast, has no Passmark scores listed, so its performance in legacy DirectX titles cannot be directly compared from this data.

When viewing the average benchmark scores, the picture becomes more nuanced. The RTX A2000 Mobile’s average score is 13,821, which is dragged down by its low Passmark results. The GTX 660 Ti’s average is 15,063, a figure that is actually higher than the A2000 Mobile’s average. This inversion highlights a critical point: while the A2000 Mobile is a compute monster in OpenCL and Vulkan, its overall average is lower than the GTX 660 Ti’s, meaning the GTX 660 Ti is more consistent across the aggregate of all tests. The GTX 660 Ti sits at the 57th percentile of all GPUs, while the A2000 Mobile sits at the 55th, reinforcing that the older card’s average performance is slightly better relative to the entire GPU landscape.

Architecture Differences

The architectural chasm between these two GPUs is vast. The GTX 660 Ti is built on the GK104 chip using NVIDIA’s Kepler architecture, manufactured on a 28 nm process at TSMC. It packs 3,540 million transistors into a 294 mm² die, yielding a transistor density of 12.0M per mm². In contrast, the RTX A2000 Mobile uses the GA107 chip on the Ampere architecture, fabricated by Samsung on an 8 nm node. It contains 8,700 million transistors on a smaller 200 mm² die, achieving a far higher density of 43.5M per mm². This density difference is the first sign of the A2000 Mobile’s modern design.

Core configurations diverge sharply. The GTX 660 Ti has 1,344 shading units, 112 TMUs, and 24 ROPs. The RTX A2000 Mobile nearly doubles the shading units to 2,560, but reduces TMUs to 80 while increasing ROPs to 48. More importantly, the A2000 Mobile introduces dedicated hardware that the Kepler card lacks entirely: 20 ray tracing cores and 80 tensor cores. This is a fundamental feature gap, as the GTX 660 Ti cannot accelerate ray tracing or AI workloads at all.

Clock speeds and memory technology also reflect the generational gap. The GTX 660 Ti runs at a base of 915 MHz with a boost of 980 MHz, while the A2000 Mobile boosts higher from a 1,215 MHz base to 1,687 MHz. The memory subsystem tells a similar story: the GTX 660 Ti has 2 GB of GDDR5 on a 192-bit bus, delivering 144.2 GB/s of bandwidth. The A2000 Mobile offers 4 GB of GDDR6 on a 128-bit bus, but achieves 192.0 GB/s due to faster 12 Gbps effective memory speed. The result is a higher bandwidth figure despite the narrower bus.

Compute rates are where the A2000 Mobile pulls away decisively. The GTX 660 Ti delivers 2.634 TFLOPS of FP32 performance, while the A2000 Mobile delivers 8.637 TFLOPS — more than three times the raw throughput. The A2000 Mobile also offers FP16 performance at a 1:1 ratio with FP32 (8.637 TFLOPS), a capability the GTX 660 Ti lacks entirely. Pixel and texture rates follow suit: the A2000 Mobile’s 80.98 GPixel/s and 135.0 GTexel/s dwarf the GTX 660 Ti’s 27.44 GPixel/s and 109.8 GTexel/s.

Power and interface differences are equally stark. The GTX 660 Ti is a dual-slot desktop card with a 150 W TDP and two 6-pin power connectors, requiring a 450 W power supply. The RTX A2000 Mobile is an IGP (integrated graphics processor) with a 95 W TDP and no power connectors, designed for portable devices. The bus interface also advances from PCIe 3.0 x16 on the GTX 660 Ti to PCIe 4.0 x16 on the A2000 Mobile. API support reflects the newer architecture: the A2000 Mobile supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the GTX 660 Ti is limited to DirectX 12 (11_0) and Vulkan 1.2.175.

Where Each One Wins

The RTX A2000 Mobile’s wins are unambiguous in compute-heavy workloads. Its Geekbench OpenCL score of 56,518 is 73.3% higher than the GTX 660 Ti’s, making it the clear choice for general-purpose GPU computing, scientific simulations, and any task that leverages OpenCL. Its Vulkan score of 53,146, also 71.8% higher, positions it as superior for modern Vulkan-based game engines and compute APIs. The presence of ray tracing and tensor cores means it can handle real-time ray tracing and AI acceleration, which the GTX 660 Ti simply cannot do. Its higher FP32 (8.637 TFLOPS) and FP16 (8.637 TFLOPS) throughput, combined with 192.0 GB/s of memory bandwidth, make it a far more capable platform for content creation and modern rendering pipelines.

The GTX 660 Ti, despite losing both head-to-head benchmarks, has its own arena of strength. Its average benchmark score of 15,063 is higher than the A2000 Mobile’s 13,821, indicating that it performs more consistently across a broader set of tests. This suggests that in legacy DirectX workloads — the Passmark DirectX 10 and DirectX 11 tests where the A2000 Mobile scores a mere 57 and 68, respectively — the GTX 660 Ti likely holds its own, though no direct comparison scores exist. The GTX 660 Ti’s dual-slot design and desktop form factor make it a drop-in card for older systems, whereas the A2000 Mobile is an integrated solution dependent on the host laptop’s design. For users running older software that predates modern compute APIs, the GTX 660 Ti’s consistent average performance is a meaningful advantage.

The Verdict

The data is unambiguous on raw compute: the RTX A2000 Mobile is the superior GPU. It wins both head-to-head benchmarks by over 70%, offers triple the FP32 throughput (8.637 vs 2.634 TFLOPS), and includes feature hardware — 20 ray tracing cores and 80 tensor cores — that the GTX 660 Ti lacks entirely. Its 8 nm Ampere architecture, with 8,700 million transistors and 4 GB of GDDR6, is a generation ahead of the 28 nm Kepler design. For any modern workload that uses OpenCL, Vulkan, ray tracing, or AI inference, the A2000 Mobile is the only rational choice. Its higher memory bandwidth (192.0 GB/s vs 144.2 GB/s) and larger memory pool (4 GB vs 2 GB) further cement its lead.

However, the GTX 660 Ti is not without merit in this comparison. Its average benchmark score of 15,063 exceeds the A2000 Mobile’s 13,821, and its 57th percentile rank versus the A2000 Mobile’s 55th means it sits higher in the overall GPU distribution. This is a statistical anomaly driven by the A2000 Mobile’s poor Passmark scores, but it indicates that the GTX 660 Ti maintains a more balanced profile across diverse test types. The GTX 660 Ti is also the only one of the two with a launch MSRP of 299 USD, making it a known-quantity desktop card with dedicated power connectors and display outputs (2x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2). The A2000 Mobile, being an IGP with portable-device-dependent outputs, is not a direct replacement.

The verdict depends on the use case. For a desktop build targeting older DirectX titles or a system that needs a reliable, well-rounded card for mixed legacy workloads, the GTX 660 Ti’s higher average score and desktop flexibility are worth noting. For any modern compute, gaming, or professional application that can leverage modern APIs, the RTX A2000 Mobile is overwhelmingly the better performer, despite its lower average benchmark score. The 73.3% and 71.8% deltas in OpenCL and Vulkan are decisive, and the architectural advantages — ray tracing cores, tensor cores, higher clock speeds, and greater memory bandwidth — are insurmountable.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The NVIDIA GeForce GTX 660 Ti has a higher average benchmark score of 15,063, compared to the NVIDIA RTX A2000 Mobile’s 13,821.

Q: What is the performance difference in Geekbench OpenCL?

A: The RTX A2000 Mobile scores 56,518 in Geekbench OpenCL, which is 73.3% higher than the GTX 660 Ti’s score of 15,113.

Q: Does the RTX A2000 Mobile support ray tracing?

A: Yes, the RTX A2000 Mobile has 20 ray tracing cores, while the GTX 660 Ti has no ray tracing cores.

Q: What are the memory configurations of each card?

A: The GTX 660 Ti has 2 GB of GDDR5 on a 192-bit bus with 144.2 GB/s bandwidth. The RTX A2000 Mobile has 4 GB of GDDR6 on a 128-bit bus with 192.0 GB/s bandwidth.

Q: How do their transistor densities compare?

A: The GTX 660 Ti has a transistor density of 12.0M per mm², while the RTX A2000 Mobile has a density of 43.5M per mm².

Q: Which GPU has more shading units?

A: The RTX A2000 Mobile has 2,560 shading units, compared to the GTX 660 Ti’s 1,344 shading units.