AMD Radeon RX Vega 11 Embedded

The AMD Radeon RX Vega 11 Embedded is a 14 nm integrated GPU from AMD, part of the Vega IGP (Raven Ridge) generation. It combines GCN 5.0 architecture, a Raven chip, and a GlobalFoundries 14 nm process into a 210 mm² die containing 4,940 million transistors, for a transistor density of 23.5M per mm². The GPU is listed with a 300 MHz base clock and a 1251 MHz boost clock. Its fixed-function resources include 704 shading units, 44 texture mapping units, and 8 ROPs, with a pixel rate of 10.01 GPixel/s and a texture rate of 55.04 GTexel/s. The record also lists an FP32 throughput of 1.761 TFLOPS and an FP16 throughput of 3.523 TFLOPS. This is an end-of-life part released on 2018-04-18T17:00:00.000Z.

Who Should Consider It

The data points toward a modest integrated graphics solution rather than a high-throughput discrete part. With 704 shading units and FP32 performance of 1.761 TFLOPS, the RX Vega 11 Embedded sits in a class where lighter rendering workloads and reduced-detail scenarios are more plausible than heavy high-resolution processing. The 8 ROPs and 10.01 GPixel/s pixel rate set a hard boundary for fill-rate-bound tasks; the 44 TMUs and 55.04 GTexel/s texture rate are enough to feed a relatively small number of pixels, but not enough to sustain large, texture-heavy scenes at high quality.

Because the bandwidth is listed as System Dependent, the practical performance of this IGP depends heavily on the host platform’s memory configuration. That makes it a candidate for systems where memory bandwidth is predictable and the workload is kept within modest resolution and detail ranges. No benchmark entries are present to confirm specific game settings, so the recommendation cannot be tied to measured scores. Instead, the specification envelope suggests that users should treat this as an integrated part for lower-resolution, lower-detail use, not as a high-resolution performer.

Memory Subsystem

All memory fields in the record are labeled System Shared. The memory size is System Shared, the memory type is System Shared, and the bus width is System Shared. There is no dedicated VRAM pool, no fixed frame buffer capacity, and no fixed memory bus figure. Bandwidth is listed as System Dependent, meaning the actual throughput available to the GPU changes according to the system memory used by the host.

This has direct implications for high resolutions. A shared memory subsystem competes with the CPU for the same memory resources, so the GPU cannot rely on a private high-bandwidth frame buffer. The absence of a fixed bandwidth number also means there is no guaranteed throughput for large textures or high-resolution surfaces. In practical terms, the memory subsystem is a variable rather than a designed constant. For high-resolution rendering, that variability is a limiting factor; the data does not provide any fixed bandwidth assertion to suggest otherwise.

Ray Tracing and Feature Set

The record lists no RT cores and no tensor cores. Neither hardware ray tracing blocks nor tensor-oriented accelerators are present in the specification. That means the feature set is reliant on the GCN 5.0 architecture’s general-purpose shader resources and the APIs stated in the record. The API support includes DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3. DirectX support is qualified at feature level 12_1, which defines a specific set of shader and resource capabilities but does not indicate a more advanced feature level. OpenGL 4.6 and Vulkan 1.3 provide modern graphics entry points, though the absence of dedicated RT or tensor hardware limits specialized acceleration paths.

Power and Cooling

The thermal envelope is low: TDP is listed as 25 W. The slot width is IGP, and the power connectors field is None, meaning there are no separate auxiliary power connector requirements. The suggested PSU is not listed, so no recommended power supply wattage can be stated from the data. Display outputs are described as Motherboard Dependent, which indicates that the physical connector arrangement is determined by the motherboard rather than by the GPU itself. The bus interface is also listed as IGP, consistent with an integrated design. A 25 W TDP with no auxiliary power connectors points to a motherboard-driven power solution and a modest cooling requirement, although the record does not specify a cooler.

How It Compares

The nearestRivals array in the data is empty, so there are no rival names, scores, or percentage deltas to report. This analysis cannot assign a lead or deficit against competing GPUs because the record simply does not include those entries. What is available is lineage data: the predecessor is listed as GCN 3.0 IGP, and the successor is Vega II IGP. That places the RX Vega 11 Embedded between those two IGP generations in the product sequence. The architecture is GCN 5.0, manufactured on a 14 nm node, so the structural comparison is to an older GCN 3.0 IGP before it and a Vega II IGP after it. No numeric comparison beyond that lineage is present in the pack.

FAQ

Q: Does the AMD Radeon RX Vega 11 Embedded have dedicated VRAM?

A: No. The memory size, type, and bus width are all listed as System Shared, and bandwidth is listed as System Dependent.

Q: What TDP and power connector requirements are listed?

A: The TDP is 25 W, the slot width is IGP, and the power connectors field is None. No suggested PSU is listed.

Q: Which graphics APIs are supported?

A: The listed APIs are DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3.

Q: Does it include ray tracing cores or tensor cores?

A: Neither RT cores nor tensor cores are listed in the specification.

Q: What are the core counts?

A: The record lists 704 shading units, 44 texture mapping units, and 8 ROPs.

Q: What is the production status?

A: The production status is end-of-life, with a release date of 2018-04-18T17:00:00.000Z.

Benchmark Performance

The benchmarks array in the record is empty. The only aggregate metrics are avgBenchmarkScore of 0 and percentileVsAllGpus of 50. Because there are no benchmark entries and no nearest-rival entries, exact percentage deltas against other GPUs cannot be computed. The 50th percentile figure places the part at the midpoint of the all-GPU distribution in the record, but the accompanying score of 0 means the percentile is not tied to a positive measured score in this dataset. In other words, the aggregate result does not provide a usable performance baseline.

The remaining performance indicators are fixed-function throughput figures. The pixel rate is 10.01 GPixel/s, the texture rate is 55.04 GTexel/s, FP32 throughput is 1.761 TFLOPS, and FP16 throughput is 3.523 TFLOPS. These numbers describe a low-throughput integrated GPU. The 8 ROPs constrain pixel output, while the 704 shading units and 44 TMUs provide modest shader and texture capacity. With memory bandwidth dependent on the system, the measured throughput of the GPU itself is not the only variable; memory behavior can further constrain performance. The overall picture is one of conservative, low-power integrated performance rather than competitive headroom.