NVIDIA GeForce GTS 240 OEM

The NVIDIA GeForce GTS 240 OEM is a graphics card from the GeForce 200 generation, built on the G92B chip and the Tesla architecture. It was released on June 30, 2009, and is now end-of-life. The card packs 112 shading units, 56 texture mapping units, and 16 ROPs, paired with 1024 MB of GDDR3 memory on a 256-bit bus. Its memory clock is 1100 MHz, yielding 2.2 Gbps effective and 70.40 GB/s of bandwidth. The data sheet lists no benchmark scores, so the analysis below relies on the card’s theoretical rates and feature set.

Benchmark Performance

Without benchmark scores, the GTS 240 OEM’s performance must be inferred from its raw throughput numbers. The FP32 compute rate is 362.9 GFLOPS, a measure of single-precision floating-point capability. This figure is modest by modern standards but reflects the card’s Tesla-architecture design, which prioritizes traditional rasterization over compute-heavy workloads. The pixel rate of 10.80 GPixel/s and texture rate of 37.80 GTexel/s define the card’s fill-rate ceilings. The ratio of texture to pixel rate (37.8 / 10.8 ≈ 3.5) indicates a texture-heavy pipeline, which aligns with the 56 TMUs versus 16 ROPs.

The memory subsystem delivers 70.40 GB/s of bandwidth from a 256-bit interface and GDDR3 at 1100 MHz. This bandwidth is sufficient to feed the shading units and TMUs at lower resolutions, but it may become a bottleneck when textures and shader workloads grow. The card’s 1024 MB frame buffer is typical for its era, though modern titles with large texture sets could exceed it. The percentile field places this card at the 50th percentile among all GPUs in the database, suggesting a median position in overall performance—neither a low-end nor a high-end part. However, the average benchmark score is listed as 0, so this percentile is not backed by any measured result.

In the absence of direct rival scores, the theoretical rates offer a baseline. The FP32 figure of 362.9 GFLOPS is roughly one-tenth of what a mid-range card from a decade later would produce, but within the GeForce 200 family, it sits at a mid-tier level. The pixel and texture rates similarly indicate a card capable of 720p or 900p gaming with medium settings, though no resolution-specific claims can be made from the data pack. The card’s 55 nm process node and 754 million transistors on a 260 mm² die yield a transistor density of 2.9 million per square millimeter—a dense layout for its time, which helps explain the 120 W TDP.

Ray Tracing and Feature Set

The GTS 240 OEM has no ray tracing cores and no tensor cores, as indicated by the null entries for both. This means the card offers no hardware-accelerated ray tracing, no DLSS, and no AI-based features. Its API support is limited to DirectX 11.1 (with feature level 10_0) and OpenGL 3.3; Vulkan is not listed. The DirectX 11.1 designation with a 10_0 feature level is significant: the card can run DirectX 11.1 applications but only with the feature set of DirectX 10.0. This implies compatibility with older APIs but not with modern DirectX 12 or Vulkan titles. OpenGL 3.3 is also a legacy version, sufficient for many older games but not for recent releases that require OpenGL 4.5 or higher.

The display outputs consist of 2x DVI and 1x S-Video, which are standard for a card from 2009. There are no HDMI or DisplayPort connections, so connecting to modern monitors may require adapters. The bus interface is PCIe 2.0 x16, which is backward compatible with older slots but lacks the bandwidth of PCIe 3.0 or 4.0. The card’s single-slot design and 1x 6-pin power connector indicate a modest power draw of 120 W, with a suggested PSU rating of 300 W. This makes it suitable for older systems with limited power supplies.

The lack of ray tracing and tensor cores is a major limitation for any modern use case. The card cannot accelerate DirectX Raytracing (DXR) workloads, and its absence of Vulkan support further restricts its compatibility with current games and applications. The feature set is firmly rooted in the late-2000s, and any software that relies on newer APIs or hardware features will not run optimally, if at all.

How It Compares

The GTS 240 OEM sits between the GeForce 9 series (its predecessor) and the GeForce 400 series (its successor) within NVIDIA’s product stack. The GeForce 9 series used the same Tesla architecture but typically had fewer shading units and lower clock speeds. The GeForce 400 series, by contrast, introduced a new architecture (Fermi) with support for DirectX 11 and a fundamentally different shader design. The GTS 240 OEM, with its G92B chip, is essentially a refined version of the G92 architecture found in the GeForce 9 series, but the fact pack does not list any direct rival cards or benchmark deltas. Therefore, a precise percentage comparison cannot be made.

What the data does show is that the GTS 240 OEM is a mid-range part within its own generation. Its 112 shading units and 56 TMUs are higher than entry-level GeForce 200 cards, but its 16 ROPs and 256-bit memory bus are shared with many of its siblings. The 1024 MB frame buffer was a common size for mid-range cards at the time, and the 70.40 GB/s bandwidth is typical for a 256-bit GDDR3 setup. The card’s 50th percentile rank in the database suggests it performs in the middle of all GPUs ever tested, though again, no specific scores are provided.

Compared to its successor, the GeForce 400 series, the GTS 240 OEM lacks the newer features such as DirectX 11 support at the full feature level and improved tessellation. The GeForce 400 also introduced a new memory controller and a different thread scheduler. However, the GTS 240 OEM’s lower TDP of 120 W (versus the higher power draw of many GeForce 400 cards) may be an advantage in older systems. The predecessor, GeForce 9, likely had similar performance but with slightly lower clock speeds or fewer shaders, though no numbers are available to confirm this.

In the absence of nearest rival data, the GTS 240 OEM’s position is best described as a bridge between two architectural generations. It retains the Tesla architecture but with a denser 55 nm process and a higher transistor count (754 million) than many GeForce 9 cards. Its performance is unlikely to satisfy modern gaming demands, but it can handle legacy titles and basic 2D workloads without issue.

Who Should Consider It

Given the GTS 240 OEM’s specifications, it is only suitable for very old games, basic productivity, or as a display adapter for systems that do not require 3D acceleration. The card’s 1024 MB memory and 70.40 GB/s bandwidth are enough for 720p or 800x600 resolutions in games from its era, but modern titles would likely exceed the frame buffer or cause the card to run out of shader power. The lack of Vulkan support and the limited DirectX feature level (10_0) mean that many current games will not even launch, as they require DirectX 11 feature level 11_0 or higher.

Users with an older system that has a PCIe 2.0 x16 slot and a 300 W power supply could install this card as a replacement for a failed or inadequate GPU. Its single-slot design and 120 W TDP make it easy to fit in compact cases, and its 229 mm (9-inch) length is standard for the era. The 2x DVI and 1x S-Video outputs are sufficient for connecting to older monitors, though VGA or HDMI adapters may be needed for modern displays.

For anyone considering this card for gaming, the data clearly shows it is not a viable option for current titles. Even at low resolutions and settings, the card’s fill rates and compute power are far below what modern engines require. It could serve as a secondary card for multi-monitor setups (if the motherboard supports multiple GPUs) or as a test bench for retro gaming. However, its end-of-life status and lack of driver support for newer operating systems may further limit its usability.

FAQ

Q: Does the GTS 240 OEM support ray tracing?

A: No. The fact pack lists null for both RT cores and tensor cores, indicating no hardware ray tracing or AI acceleration.

Q: What is the maximum supported DirectX version?

A: The card supports DirectX 11.1, but only with the 10_0 feature level. This means it can run DX11.1 applications but with the capabilities of DirectX 10.0.

Q: How much memory does it have and what type?

A: It has 1024 MB of GDDR3 memory on a 256-bit bus, with a bandwidth of 70.40 GB/s.

Q: What is the power requirement?

A: The card has a TDP of 120 W and requires one 6-pin power connector. The suggested PSU rating is 300 W.

Q: Does it support Vulkan?

A: No. The API list does not include Vulkan; only DirectX 11.1 (10_0) and OpenGL 3.3 are supported.

Q: What display outputs are available?

A: The card offers 2x DVI and 1x S-Video outputs. There are no HDMI or DisplayPort connections.

Q: What is the card's length?

A: The length is 229 mm, which is 9 inches, and it occupies a single slot.