AMD Radeon Pro V340 16 GB

The AMD Radeon Pro V340 16 GB is a professional workstation graphics card built on the Vega 10 chip and the GCN 5.0 architecture. Fabricated on a 14 nm process at GlobalFoundries, it integrates 12,500 million transistors into a 495 mm² die, yielding a transistor density of 25.3M per mm². Released on 2018-08-25, the card is now end-of-life, and its dual-slot design measures 267 mm (10.5 inches) in length and 111 mm (4.4 inches) in height. The specification sheet lists 3584 shading units, 224 texture mapping units, and 64 raster operation units, with a base clock of 852 MHz and a boost clock of 1500 MHz.

How It Compares

The fact pack for this card lists no nearest rivals, no benchmark scores, and an average benchmark score of 0. The percentile versus all GPUs is 50, which places it exactly at the median of the database's GPU population. Without a benchmark score, that percentile is a positional marker rather than a measured performance result. The card's predecessor is the Radeon Pro Polaris, and its successor is the Radeon Pro Navi, but the fact pack provides no comparative performance data for either. Consequently, the data cannot support any direct head-to-head percentage comparisons against specific rivals.

The 50th percentile indicates that, in terms of the database's ranking, this card sits in the middle of the pack, but the absence of a score means that ranking is provisional. The only performance metrics available are the raw specifications: FP32 throughput of 10.75 TFLOPS and FP16 throughput of 21.50 TFLOPS (2:1). These figures define the card's theoretical compute ceiling, but without rival scores or deltaPct values, there is no way to contextualize them against competing products. The card's position is therefore defined by its architecture and memory subsystem rather than by any measured benchmark data.

Memory Subsystem

The Radeon Pro V340 16 GB is equipped with 16 GB of HBM2 memory on a 2048-bit bus. The memory clock is 945 MHz, translating to 1890 Mbps effective, and this configuration yields a memory bandwidth of 483.8 GB/s. For high-resolution workloads, the combination of a 2048-bit bus and 483.8 GB/s bandwidth means that large textures and data sets can be moved quickly between the GPU and memory. The 16 GB capacity is substantial for professional applications that load large models or high-resolution frames, and the wide bus allows the memory controller to access data across a very large address space in parallel.

In practice, this bandwidth is well-suited for rendering scenes with complex geometry or processing large compute buffers. The HBM2 type is known for its high bandwidth characteristics, and the 483.8 GB/s figure is the sole bandwidth metric available. The card's memory clock of 945 MHz is the operating frequency for the HBM2 stacks, and the effective rate of 1890 Mbps is the data transfer rate. The 2048-bit bus width is a key differentiator, as it provides a broad pathway for data movement. For high-resolution output, the 16 GB capacity reduces the need for texture streaming, while the bandwidth ensures that the shading units are not starved for data.

Ray Tracing and Feature Set

The fact pack lists no dedicated ray tracing cores and no tensor cores for this GPU. The architecture is GCN 5.0, which predates dedicated RT hardware. The card's API support includes DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3. These APIs are the software interfaces through which the card's features are exposed. Without dedicated RT cores, ray tracing workloads would rely on the general-purpose compute units, but the fact pack does not provide any specific ray tracing performance data. The absence of tensor cores means that machine learning acceleration is not a hardware-specific feature of this card.

The feature set is therefore defined by its compute capabilities and API compliance. The DirectX 12 (12_1) support indicates compatibility with modern graphics features, while Vulkan 1.3 and OpenGL 4.6 provide cross-platform access. The card's pixel rate is 96.00 GPixel/s and its texture rate is 336.0 GTexel/s, which are the rasterization throughput metrics available in the pack. These rates, combined with the 64 ROPs, define the card's fill-rate capabilities for traditional rendering. The single mini-DisplayPort 1.4a output is the only display interface, which limits physical display connectivity to one monitor.

Who Should Consider It

The Radeon Pro V340 16 GB is positioned as a professional workstation card, given its "Radeon Pro" branding. Its 16 GB of HBM2 memory and 483.8 GB/s bandwidth make it suitable for large datasets and high-resolution textures. The FP32 performance of 10.75 TFLOPS and FP16 performance of 21.50 TFLOPS (2:1) indicate strong compute throughput for scientific or engineering workloads that can leverage the 2:1 FP16 ratio. However, the single mini-DisplayPort 1.4a output means it is not designed for multi-display setups or typical desktop use. The card's end-of-life status and release date of 2018-08-25 suggest it is a legacy product.

Users who need a compute-focused card with high memory capacity and bandwidth, and who do not require multiple display outputs, might consider it. The card's 50th percentile ranking among all GPUs in the database, while lacking a benchmark score, indicates it is not an outlier in either direction. The 230 W TDP and 550 W suggested PSU requirement are the power constraints to note. For high-resolution rendering, the 16 GB memory is a strong asset, but the single output limits physical display configuration. The card's dual-slot design and 267 mm length (10.5 inches) mean it requires adequate chassis space, and the 111 mm (4.4 inches) height is within standard ATX specifications.

Benchmark Performance

The fact pack provides no benchmark scores for this card. The average benchmark score is 0, and the percentile versus all GPUs is 50. This means the database has no measured performance data to analyze. The only performance metrics available are the raw specifications. The FP32 throughput is 10.75 TFLOPS, and the FP16 throughput is 21.50 TFLOPS, with the 2:1 ratio indicating that FP16 operations are exactly double the FP32 rate. The pixel rate is 96.00 GPixel/s, and the texture rate is 336.0 GTexel/s. The base clock is 852 MHz, and the boost clock is 1500 MHz, showing a significant boost range. The shading units number 3584, with 224 TMUs and 64 ROPs.

These numbers define the card's theoretical peak performance. Without rival scores or deltaPct values, the data cannot support any comparative percentage claims. The 50th percentile is the only positional reference, but it is not derived from a benchmark score. The card's compute density is reflected in its transistor count of 12,500 million on a 495 mm² die, yielding 25.3M transistors per mm². The 14 nm process at GlobalFoundries is the manufacturing node. The absence of benchmark data means that any performance assessment must rely solely on the specification sheet. The card's memory bandwidth of 483.8 GB/s, combined with its FP32 rate, suggests a balanced compute-to-memory ratio, but this is an interpretation of the specs rather than a measured result. The boost clock of 1500 MHz is the maximum frequency under load, while the base clock of 852 MHz is the sustained floor.

Power and Cooling

The Radeon Pro V340 16 GB has a thermal design power (TDP) of 230 W. The suggested power supply is 550 W. Power is delivered via two 8-pin connectors. The card is a dual-slot design, which affects cooling and chassis compatibility. Its physical dimensions are 267 mm (10.5 inches) in length and 111 mm (4.4 inches) in height. The 230 W TDP is a fixed figure from the fact pack, and the 550 W PSU recommendation is the minimum suggested. The two 8-pin connectors are required for power delivery. The dual-slot form factor indicates a substantial cooler, but the fact pack does not specify the cooler type.

The length of 267 mm (10.5 inches) means it will fit in most mid-tower cases, but the height of 111 mm (4.4 inches) is within standard ATX specifications. The card's power requirements are modest for a professional GPU of this era, but the 550 W PSU is a hard requirement. The end-of-life status means that power efficiency is not a selling point; the 230 W TDP is the figure to plan around. The memory type HBM2 is typically more power-efficient than other types, but the pack does not provide comparative power data for memory. The card's boost clock of 1500 MHz and base clock of 852 MHz are the operating frequencies that influence power draw, but the TDP is the official rating. The dual-slot design means the cooler occupies two expansion slots, which is a consideration for dense workstation builds.