NVIDIA GRID K100
NVIDIA graphics card specifications and benchmark scores
At a Glance
NVIDIANVIDIA GRID K100 Specifications
GPU Core
Shader units and compute resources
The NVIDIA GRID K100 GPU core specifications define its raw processing power for graphics and compute workloads. Shading units (also called CUDA cores, stream processors, or execution units depending on manufacturer) handle the parallel calculations required for rendering. TMUs (Texture Mapping Units) process texture data, while ROPs (Render Output Units) handle final pixel output. Higher shader counts generally translate to better GPU benchmark performance, especially in demanding games and 3D applications.
GRID K100 Clock Speeds
GPU and memory frequencies
Clock speeds directly impact the GRID K100's performance in GPU benchmarks and real-world gaming. The base clock represents the minimum guaranteed frequency, while the boost clock indicates peak performance under optimal thermal conditions. Memory clock speed affects texture loading and frame buffer operations. The GRID K100 by NVIDIA dynamically adjusts frequencies based on workload, temperature, and power limits to maximize performance while maintaining stability.
NVIDIA's GRID K100 Memory
VRAM capacity and bandwidth
VRAM (Video RAM) is dedicated memory for storing textures, frame buffers, and shader data. The GRID K100's memory capacity determines how well it handles high-resolution textures and multiple displays. Memory bandwidth, measured in GB/s, affects how quickly data moves between the GPU and VRAM. Higher bandwidth improves performance in memory-intensive scenarios like 4K gaming. The memory bus width and type (GDDR6, GDDR6X, HBM) significantly influence overall GPU benchmark scores.
GRID K100 by NVIDIA Cache
On-chip cache hierarchy
On-chip cache provides ultra-fast data access for the GRID K100, reducing the need to fetch data from slower VRAM. L1 and L2 caches store frequently accessed data close to the compute units. AMD's Infinity Cache (L3) dramatically increases effective bandwidth, improving GPU benchmark performance without requiring wider memory buses. Larger cache sizes help maintain high frame rates in memory-bound scenarios and reduce power consumption by minimizing VRAM accesses.
GRID K100 Theoretical Performance
Compute and fill rates
Theoretical performance metrics provide a baseline for comparing the NVIDIA GRID K100 against other graphics cards. FP32 (single-precision) performance, measured in TFLOPS, indicates compute capability for gaming and general GPU workloads. FP64 (double-precision) matters for scientific computing. Pixel and texture fill rates determine how quickly the GPU can render complex scenes. While real-world GPU benchmark results depend on many factors, these specifications help predict relative performance levels.
Kepler Architecture & Process
Manufacturing and design details
The NVIDIA GRID K100 is built on NVIDIA's Kepler architecture, which defines how the GPU processes graphics and compute workloads. The manufacturing process node affects power efficiency, thermal characteristics, and maximum clock speeds. Smaller process nodes pack more transistors into the same die area, enabling higher performance per watt. Understanding the architecture helps predict how the GRID K100 will perform in GPU benchmarks compared to previous generations.
Power & Thermal
TDP and power requirements
Power specifications for the NVIDIA GRID K100 determine PSU requirements and thermal management needs. TDP (Thermal Design Power) indicates the heat output under typical loads, guiding cooler selection. Power connector requirements ensure adequate power delivery for stable operation during demanding GPU benchmarks. The suggested PSU wattage accounts for the entire system, not just the graphics card. Efficient power delivery enables the GRID K100 to maintain boost clocks without throttling.
GRID K100 by NVIDIA Physical & Connectivity
Dimensions and outputs
Physical dimensions of the NVIDIA GRID K100 are critical for case compatibility. Card length, height, and slot width determine whether it fits in your chassis. The PCIe interface version affects bandwidth for communication with the CPU. Display outputs define monitor connectivity options, with modern cards supporting multiple high-resolution displays simultaneously. Verify these specifications against your case and motherboard before purchasing to ensure a proper fit.
NVIDIA API Support
Graphics and compute APIs
API support determines which games and applications can fully utilize the NVIDIA GRID K100. DirectX 12 Ultimate enables advanced features like ray tracing and variable rate shading. Vulkan provides cross-platform graphics capabilities with low-level hardware access. OpenGL remains important for professional applications and older games. CUDA (NVIDIA) and OpenCL enable GPU compute for video editing, 3D rendering, and scientific applications. Higher API versions unlock newer graphical features in GPU benchmarks and games.
GRID K100 Product Information
Release and pricing details
The NVIDIA GRID K100 is manufactured by NVIDIA as part of their graphics card lineup. Release date and launch pricing provide context for comparing GPU benchmark results with competing products from the same era. Understanding the product lifecycle helps evaluate whether the GRID K100 by NVIDIA represents good value at current market prices. Predecessor and successor information aids in tracking generational improvements and planning future upgrades.
About NVIDIA GRID K100
The NVIDIA GRID K100 is an end-of-life, dual-slot server-oriented graphics solution built on the 28 nm Kepler architecture, specifically the GK107 chip fabricated by TSMC. It integrates 1,270 million transistors on a 118 mm² die, resulting in a transistor density of 10.8M per mm², and it was released on 2013-06-27 with a launch MSRP of 63 USD. The data shows this card occupies the 50th percentile among all GPUs in the database, with an average benchmark score of zero, indicating it is not designed for general-purpose rendering workloads but rather for specific virtualized environments.
Benchmark Performance
The GRID K100’s benchmark data is stark: its average benchmark score is 0, and it has no nearest rivals listed in the database. This absence of comparative scores is itself revealing. The card’s FP32 compute is 326.4 GFLOPS, a figure that places it far below any modern consumer or workstation GPU. To contextualize, the shading units number 192, with 16 texture mapping units and 16 raster operation pipelines. The pixel rate is 3.400 GPixel/s, and the texture rate is 13.60 GTexel/s. These numbers suggest a card that was never intended to push high frame rates in conventional applications. Instead, the 50th percentile ranking is likely a statistical artifact of the database’s inclusion criteria, not a measure of gaming capability. The lack of any rival scores means there are no deltaPct values to report; the card stands alone in its category, which is a niche server virtualization product rather than a performance part. When interpreting these figures, the key takeaway is that raw performance metrics are secondary to the card’s intended function—grid computing and virtual desktop infrastructure, where individual virtual machines receive a fraction of the hardware’s resources. The 326.4 GFLOPS FP32 throughput, while minuscule by today’s standards, was sufficient for basic 2D and lightweight 3D workloads in a multi-user server context. No clock speeds are provided for the base or boost states, further underscoring that this is not a product measured by peak frequency but by sustained, low-power operation across many concurrent sessions.
Ray Tracing and Feature Set
The GRID K100 has no dedicated ray tracing cores and no tensor cores, as these were not part of the Kepler architecture. The API support includes DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The DirectX 12 (11_0) designation is important: it indicates that while the hardware can expose a DirectX 12 interface, the feature level is capped at 11_0, meaning it lacks the full DirectX 12 feature set such as mesh shaders, variable rate shading, or hardware-accelerated ray tracing. Vulkan 1.2.175 provides modern low-level API access, but again, without RT cores, any ray tracing would have to be done via compute shaders, which would be impractically slow given the 326.4 GFLOPS compute budget. OpenGL 4.6 support is solid for legacy applications. The absence of tensor cores means no AI-accelerated features like DLSS or neural rendering are possible. The card’s display outputs are listed as "No outputs," confirming that it is a headless server part—all rendering is performed off-screen and streamed to clients over the network. The feature set, therefore, is oriented toward API compatibility rather than state-of-the-art visual effects. For virtualized environments, this level of API support allows legacy DirectX 11 and OpenGL applications to run, but modern titles requiring DirectX 12 Ultimate features will not work. The lack of RT and tensor cores is expected for a 2013 product, but the data shows that even the API support is limited to the 11_0 feature level, which is a significant constraint for any modern workload.
Memory Subsystem
The GRID K100 comes equipped with 256 MB of DDR3 memory on a 128-bit bus, yielding a memory bandwidth of 28.51 GB/s. The memory clock is 891 MHz, with an effective data rate of 1782 Mbps. This is an extraordinarily small memory capacity and bandwidth by any standard, even for 2013. The 256 MB frame buffer is insufficient for any modern game at 1080p, let alone higher resolutions. The 128-bit bus width limits the amount of data that can be transferred per clock cycle, and the 28.51 GB/s bandwidth is roughly an order of magnitude lower than what even entry-level consumer GPUs of that era offered. For high-resolution workloads, this memory subsystem is a severe bottleneck. The card’s target use case—server-side rendering for virtual desktops—typically involves low resolution (e.g., 1024x768 or 1280x1024) and low color depth to keep memory usage within the 256 MB limit. At such settings, the 28.51 GB/s bandwidth can handle basic 2D desktop composition and light 3D acceleration, but anything more demanding would cause texture thrashing and stuttering. The DDR3 type is also slower than GDDR5, which was common on consumer GPUs of the same generation. In a multi-user environment, the memory is partitioned across several virtual machines, meaning each user gets a fraction of 256 MB—often just 64 MB or less—which is comparable to early 2000s integrated graphics. The data shows no room for expansion or high-resolution textures. Therefore, the memory subsystem is adequate only for legacy operating systems and simple productivity apps, not for media consumption or 3D modeling.
Power and Cooling
The GRID K100 has a thermal design power (TDP) of 130 W, which is modest for a dual-slot card. The suggested power supply unit is 300 W, indicating that the card draws a reasonable amount of power for a server environment. There are no power connectors listed, meaning the card likely draws all its power from the PCIe 3.0 x16 slot, which can supply up to 75 W. Given the 130 W TDP, this is a discrepancy—the slot alone cannot provide 130 W, so there must be an additional connector, but the FACT PACK does not specify one. The dual-slot design suggests a passive cooling solution or a low-profile active cooler, typical for server cards that rely on chassis airflow. The 28 nm process node helps keep power consumption in check, but the 130 W TDP is still notable for a card with only 192 shading units. The power draw is likely dominated by the memory subsystem and the fixed-function hardware for virtualization. In a server rack with dozens of such cards, the cumulative power draw and heat output are significant, but the 300 W PSU recommendation implies a low per-card overhead. The lack of display outputs means the card does not need to drive a monitor, saving some power. The cooling solution, while unspecified, must be sufficient to dissipate 130 W in a dual-slot form factor. For system integrators, the key data points are the 130 W TDP and 300 W PSU suggestion, which allow for accurate power budgeting in dense server configurations. No clock boost information is provided, suggesting the card runs at a fixed clock to ensure predictable performance across all virtual machines.
Who Should Consider It
The GRID K100 is not for gamers or workstation users. Its 256 MB memory and 326.4 GFLOPS compute are far too low for any modern resolution or settings. Instead, the data points to a specific use case: virtual desktop infrastructure (VDI) where multiple users need basic 2D desktop acceleration and light 3D support for legacy applications. The card’s 50th percentile ranking and zero benchmark score confirm that it is not evaluated on standard gaming metrics. For organizations running Windows 7 or Windows 10 virtual machines with Office productivity suites, web browsing, and perhaps older CAD or GIS software, the GRID K100 can provide hardware acceleration that offloads the CPU. The DirectX 12 (11_0) support means that any application requiring DirectX 11 features will run, but DirectX 12-only titles will not. The Vulkan 1.2.175 support could enable some modern Linux applications, but the low memory and bandwidth will limit complexity. The 130 W TDP and 300 W PSU recommendation make it feasible to deploy in existing server chassis. However, given the end-of-life status and the minuscule memory, this card is only suitable for very low-resolution virtual desktops (e.g., 1280x1024 or lower) with reduced color depth. For any user expecting to run multiple monitors or 4K resolution, the 28.51 GB/s bandwidth and 256 MB VRAM are insurmountable barriers. In summary, this card is for niche, low-intensity virtualized environments, not for any form of modern gaming or professional rendering.
FAQ
Q: What is the maximum API level supported by the GRID K100?
A: The card supports DirectX 12 (11_0), meaning it exposes a DirectX 12 interface but only up to the 11_0 feature level. It also supports OpenGL 4.6 and Vulkan 1.2.175.
Q: Does the GRID K100 support ray tracing?
A: No, the card has no ray tracing cores and no tensor cores. The Kepler architecture does not include hardware acceleration for ray tracing or AI workloads.
Q: How much video memory does the GRID K100 have?
A: It has 256 MB of DDR3 memory on a 128-bit bus, with a memory bandwidth of 28.51 GB/s. The effective memory clock is 1782 Mbps.
Q: What power supply is recommended for the GRID K100?
A: The suggested PSU is 300 W, and the card has a TDP of 130 W. It is a dual-slot card with no specified power connectors.
Q: Can the GRID K100 output video to a display?
A: No, the card has no display outputs. It is designed for server-side rendering, with results streamed to clients over a network.
Q: What is the production status of the GRID K100?
A: The production status is end-of-life. It was released on 2013-06-27 with a launch MSRP of 63 USD.
Detailed benchmark scores and charts for the NVIDIA GRID K100 are below.
Benchmark Scores
No benchmark data available for this GPU.
Compare with Other GPUs
Select another GPU to compare specifications and benchmarks side-by-side.
Browse GPUs