NVIDIA GeForce GTX 1650 GDDR6

The NVIDIA GeForce GTX 1650 GDDR6 is a dual-slot, 229 mm graphics card based on the TU117 chip and Turing architecture, fabricated on TSMC's 12 nm process. It was released on 2020-03-31 with a launch MSRP of 149 USD, and the data places it at the 50th percentile among all GPUs, indicating a squarely mid-pack position in the broader performance landscape.

Benchmark Performance

The GTX 1650 GDDR6 delivers 2.849 TFLOPS of FP32 compute, a figure that defines its baseline rasterization capability. This throughput is paired with a pixel rate of 50.88 GPixel/s and a texture rate of 89.04 GTexel/s, which together dictate how quickly the card can fill frames at lower resolutions. The benchmark percentile of 50 places this card exactly at the median of the entire GPU database, meaning half of all tracked graphics cards are faster and half are slower. This is not a top-tier performer, but it is not an entry-level outlier either; it sits in the middle of the pack, where it competes on efficiency and adequacy rather than raw speed.

The FP32 rate of 2.849 TFLOPS is the headline compute metric, but the card also supports a 2:1 FP16 ratio, yielding 5.699 TFLOPS for half-precision workloads. This doubling of throughput in FP16 can be leveraged in applications that support mixed-precision execution, though it does not elevate the card beyond its mid-pack classification. The base clock of 1410 MHz and boost clock of 1590 MHz are modest, and the data shows no benchmark scores to contextualize against other cards, leaving the percentile ranking as the primary comparative tool. For a card at the 50th percentile, users should expect playable frame rates at 1080p with settings tuned to medium or high, but the card will struggle with 1440p or high-refresh-rate targets where the pixel and texture rates become bottlenecks.

Memory Subsystem

The GTX 1650 GDDR6 is equipped with 4 GB of GDDR6 memory, a capacity that is adequate for 1080p gaming but increasingly constrained in modern titles that exceed this allocation. The memory operates across a 128-bit bus, which is a narrow interface for the capacity, but the GDDR6 type compensates with a memory clock of 1500 MHz, translating to 12 Gbps effective. This configuration produces a memory bandwidth of 192.0 GB/s, a figure that is respectable for the card's class and sufficient to feed the shading units at lower resolutions.

At high resolutions, the 4 GB capacity is the primary limitation. The 192.0 GB/s bandwidth is enough to sustain the GPU's compute throughput, but the capacity will force texture quality reductions or cause stuttering when the framebuffer exceeds 4 GB. The 128-bit bus width, while narrow, is matched to the card's compute capabilities, so bandwidth is not the bottleneck; capacity is. For users targeting 1080p, the memory subsystem is balanced, offering enough bandwidth to avoid starvation, but the 4 GB ceiling means that high-resolution textures or large draw distances will exceed the available VRAM, leading to performance dips that the compute hardware cannot mitigate.

Ray Tracing and Feature Set

The GTX 1650 GDDR6 does not include dedicated ray tracing cores or tensor cores, as these are null in the hardware specifications. This is a critical omission for modern feature support, as the card relies entirely on the shader units for any ray-traced effects, which will result in severe performance penalties. The absence of tensor cores also means no AI-accelerated features such as DLSS, limiting the card to traditional rendering methods.

The API support is robust for its generation, with DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The DirectX 12_1 feature level indicates support for tier-2 conservative rasterization and other advanced rasterization features, but it does not include the DirectX Raytracing (DXR) tier that is available on cards with dedicated RT cores. The Vulkan 1.4 support is notable, as it enables modern compute and rendering extensions, but again, without hardware acceleration for ray tracing, the card is relegated to rasterization-only workloads. The display outputs include 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a, which covers legacy and modern displays, but the absence of RT and tensor cores places this card firmly in the pre-ray-tracing era of gaming features.

Power and Cooling

The GTX 1650 GDDR6 has a TDP of 75 W, which is remarkably low and aligns with its mid-pack performance. This power envelope is supported by a suggested PSU rating of 250 W, meaning the card can be paired with modest power supplies without concern. The card requires no power connectors, drawing all its power from the PCIe 3.0 x16 slot, which simplifies installation in systems with limited PSU cabling. The dual-slot cooler, measuring 229 mm in length, 111 mm in height, and 35 mm in width, is adequate for the 75 W thermal load, and the absence of auxiliary power connectors indicates that the cooling solution is designed for quiet operation rather than extreme overclocking.

The low TDP is a significant advantage for this card, as it allows for deployment in pre-built systems or small form factor cases where power delivery is limited. The 12 nm process node and 4,700 million transistors on a 200 mm² die contribute to this efficiency, with a transistor density of 23.5M / mm². The thermal design is straightforward, and the lack of power connectors means there are no additional cables to manage, but the 75 W limit also caps overclocking headroom, as the card cannot draw beyond what the slot provides.

How It Compares

The data for the GTX 1650 GDDR6 lists no nearest rivals, no benchmark scores, and no delta percentages, meaning the comparative analysis must rely solely on the percentile and architectural characteristics. The 50th percentile ranking indicates that it is exactly average, with no specific rival card outperforming or underperforming it by a measurable margin in the provided data. This is a unique position, as most cards have a set of defined competitors, but the fact pack shows an empty array for this field, suggesting either a lack of comparable data or a deliberate isolation of this card's performance profile.

Without rival data, the GTX 1650 GDDR6 stands alone in the benchmark database at the median. Its 75 W TDP and 4 GB GDDR6 memory are its defining characteristics, but without a rival card to compare against, the analysis cannot state that it is 10% faster or 20% slower than any specific model. The card's predecessor is the GeForce 10 series, and its successor is the GeForce 20 series, but neither is listed as a direct comparator in the nearestRivals field, so no percentage deltas can be cited. The 50th percentile is the sole objective metric, and it anchors the card as a benchmark median, neither a standout value nor a laggard, but a reference point for mid-range performance in the database's historical context.