NVIDIA GeForce GT 120 Mac Edition

The NVIDIA GeForce GT 120 Mac Edition is a 55 nm Tesla-architecture GPU manufactured by TSMC with the G96C chip. The die contains 314 million transistors on a 121 mm² surface, for a transistor density of 2.6M/mm². It belongs to the GeForce 100 generation, was released on 2009-01-19, and sits between the GeForce 9 and GeForce 200 families in product succession. The specification table records 512 MB of GDDR3 memory on a 128-bit bus, a 50 W TDP, and a PCIe 2.0 x16 interface. Notably, the benchmark array is empty, the nearestRivals list is empty, and the average benchmark score is 0. This means the evaluation has to lean on the full specification list rather than measured performance data.

Power and Cooling

The power data in the fact pack is clear: TDP is 50 W, the suggested PSU is 250 W, and the power connector field is None. This means the card does not require an auxiliary PCIe power cable, so installation is limited to seating it in a PCIe 2.0 x16 slot. The physical dimensions match that simplicity: the card is single-slot and is 145 mm long, or 5.7 inches. A card that short is easy to fit in most case layouts, and the absence of power connectors removes one cable-management step. The 55 nm process and 314-million-transistor count are the background context for the 50 W TDP; with a modest transistor density of 2.6M/mm², the thermal load is concentrated less tightly than in denser designs. The fact pack does not specify a cooler, so the 50 W TDP is the best guide for thermal expectations. The production status is end-of-life, meaning this is a legacy part rather than a current purchase option, but the 250 W PSU recommendation keeps overall system power demand low.

Ray Tracing and Feature Set

The GT 120 Mac Edition has no RT core entries and no tensor core entries in the fact pack. Hardware ray tracing is therefore not part of its feature set, and no tensor-core acceleration is available. The API support is DirectX 11.1 with a 10_0 feature level, OpenGL 3.3, and no Vulkan support. The 10_0 feature level is important: the API string says DirectX 11.1, but the hardware feature ceiling is a DirectX 10-generation feature level. Software built for full DirectX 11 hardware features may not run at its intended level. OpenGL 3.3 offers a second path for applications that support it. Vulkan support is not listed, so Vulkan-only software has no supported path. The architecture is Tesla, which is consistent with the lack of newer RT and tensor core blocks. In practical terms, the feature set is aimed at compatibility and old workloads rather than advanced rendering techniques.

Memory Subsystem

The memory subsystem consists of 512 MB of GDDR3, a 128-bit bus, and 25.60 GB/s of bandwidth. The memory clock is 800 MHz, with an effective data rate of 1600 Mbps. At high resolutions, the 512 MB capacity is the first limit; large frame buffers, high-resolution textures, and multi-buffering can exceed that capacity quickly. The 128-bit bus and 25.60 GB/s bandwidth then determine how quickly data can move into and out of the frame buffer. With 32 shading units, 16 TMUs, and 8 ROPs, the rest of the pipeline is similarly modest. The listed pixel rate of 4.400 GPixel/s and texture rate of 8.800 GTexel/s reinforce the same picture: this is a low-bandwidth configuration. For anyone evaluating this card for higher-resolution work, the memory arrangement is the primary constraint.

How It Compares

The nearestRivals array in the fact pack is empty. There are no rival names, no scores, and no deltaPct values to compare against. This is a decisive absence: the database has not assigned a comparison set to this card, so any direct per-rival statement would be unsupported. The only relative field is percentileVsAllGpus, which is 50. However, avgBenchmarkScore is 0 and the benchmark array is empty, meaning the median placement has no measured score behind it. Product lineage provides a different kind of positioning: this is a GeForce 100 generation part, following the GeForce 9 family and preceding the GeForce 200 family. The G96C chip and 55 nm process locate it within a specific hardware generation, but not on a performance ladder. Because no nearestRivals data exists, the card cannot be ranked as faster or slower than any specific opponent. The comparison set is effectively undefined.

Benchmark Performance

The benchmark section of the fact pack is sparse. The benchmark array is empty, the average benchmark score is 0, and the nearestRivals list is empty. The sole ranking figure, percentileVsAllGpus, is 50. A 50th percentile would normally mean the GPU sits in the middle of the database, but with an average score of 0 there is no aggregated workload result to confirm it. This reads as an unverified database position rather than a measured median performance result. The only numeric performance indicators are the theoretical limits from the specification table. FP32 compute is 89.60 GFLOPS. Pixel fill rate is 4.400 GPixel/s. Texture fill rate is 8.800 GTexel/s. Those figures correspond to 32 shading units, 16 texture mapping units, and 8 render output units. The 512 MB GDDR3 memory and 25.60 GB/s bandwidth place additional practical caps on real-world performance. No deltaPct values are present, so exact percentage differences cannot be calculated.

FAQ

Q: What is the TDP and PSU recommendation?

A: The TDP is 50 W, and the suggested PSU is 250 W.

Q: Does the GT 120 Mac Edition require a power connector?

A: No. The power connector field is None, so no auxiliary power cable is listed.

Q: What display outputs does it provide?

A: It provides 1x DVI and 1x mini-DisplayPort.

Q: What APIs are supported?

A: DirectX 11.1 with a 10_0 feature level and OpenGL 3.3 are supported. Vulkan is not listed.

Q: Does the GPU have RT or tensor cores?

A: No. The fact pack lists RT cores and tensor cores as null.

Q: What is the memory configuration?

A: It uses 512 MB of GDDR3 on a 128-bit bus, with 25.60 GB/s bandwidth and an 800 MHz memory clock, or 1600 Mbps effective.