AMD Radeon HD 8400E

The AMD Radeon HD 8400E is an integrated graphics processor built on GCN 2.0 architecture, manufactured by AMD at TSMC on a 28 nm process. Its chip is listed as Kalindi, and the generation field identifies it as “GCN 2.0 IGP (Kabini Mobile).” It was released on 2013-04-22 and is marked End-of-life. The database records TeraScale 3 IGP as its predecessor and GCN 3.0 IGP as its successor. The GPU contains 128 shading units, 8 TMUs, and 4 ROPs. Its pixel rate is 2.400 GPixel/s, its texture rate is 4.800 GTexel/s, and FP32 compute is 153.6 GFLOPS. Memory size, type, and bus width are all System Shared, and memory bandwidth is System Dependent. The entry has no benchmark entries, an avgBenchmarkScore of 0, and an empty nearestRivals array, leaving percentileVsAllGpus = 50 as the only comparative field.

Benchmark Performance

The benchmarks array is empty, and the avgBenchmarkScore field is 0. As a result, the database contains no per-game or per-workload scores for the HD 8400E. The single positional figure is percentileVsAllGpus, which is 50; this places the part at the midpoint of the database’s GPU distribution. However, with no nearestRivals entries and no deltaPct values, that median position cannot be translated into a percentage lead over or deficit to any named competitor.

The fixed throughput values define the hardware’s theoretical limits. The pixel rate is 2.400 GPixel/s, produced by the 4 ROPs. The texture rate is 4.800 GTexel/s, produced by the 8 TMUs. FP32 compute is measured at 153.6 GFLOPS from the 128 shading units. These are the only performance-related numbers in the pack. Clock speeds are null in all three fields — base, boost, and game — so there is no clock frequency to relate these rates to. Memory bandwidth is also System Dependent, meaning it is not a fixed measurement and can vary with the host platform. That dependency further limits direct benchmark-style reasoning: the same GPU listing could behave differently in different Kabini Mobile systems if the platform’s memory configuration changes.

Because nearestRivals is empty, there is no stored rival score to subtract or divide in order to produce an exact percentage delta. The empty benchmark array also prevents any comparison of average scores against other products. In this state, the data can be interpreted only as a design specification: 128 shaders, 8 TMUs, 4 ROPs, 153.6 GFLOPS, 2.400 GPixel/s, 4.800 GTexel/s, and a 50th-percentile rank with no supporting rival data attached.

Who Should Consider It

Resolution/settings recommendations require measured scores, and this record has none. What the data does identify is an IGP for the Kabini Mobile platform. The slot width is IGP, and the bus interface is IGP. Display outputs are listed as Portable Device Dependent, which means the display connection is tied to the portable device rather than to a discrete GPU slot. Memory size, type, and bus width are all System Shared, and bandwidth is System Dependent. These facts frame the product as an integrated solution whose output and memory behavior are governed by the surrounding host system.

Given no benchmark scores, any statement about a specific resolution or quality setting would go beyond the FACT PACK. The raw throughput figures — 2.400 GPixel/s and 4.800 GTexel/s — are the only quantitative render limits, and they describe the entire pipeline, not a tested game condition. FP32 compute is 153.6 GFLOPS. The lack of a dedicated memory bus is visible in the data: every memory field is System Shared, so the GPU does not carry its own frame buffer specification. Memory bandwidth being System Dependent means this listing cannot promise a fixed level of memory performance.

This is therefore a part for legacy Kabini Mobile systems, not for a user seeking a standalone GPU recommendation. The end-of-life production status and the GCN 3.0 IGP successor further narrow the context to older portable platforms. A user considering it would be working within an existing portable device, not choosing a new expansion card. The FACT PACK does not support recommendations for high-load settings because no tested settings are recorded anywhere in the entry.

Power and Cooling

The TDP is 25 W. The entry lists no suggested PSU and no power connector, so no external power connector requirement is recorded. The slot width of IGP and the bus interface of IGP indicate that this is not an add-in card; the power connector field being null is consistent with that integrated positioning. The 25 W TDP is the only power-related numeric value in the record.

The silicon details are also present in the FACT PACK. The GPU is fabricated on a 28 nm process at TSMC. The die size is 110 mm², and the transistor count is 1,178 million. This yields a transistor density of 10.7M / mm². These figures describe the physical design but do not alter the TDP field. No dimensions are recorded for the card, so length, height, and width cannot be used to assess physical cooling compatibility. Likewise, the suggested PSU field is null, so the database makes no power-supply recommendation. All power and cooling information in the record reduces to the 25 W TDP, the IGP form factor, and the absence of connector/PSU specifications.

How It Compares

The nearestRivals array is empty. There are no rival names, no rival scores, and no deltaPct values in the FACT PACK. Consequently, the page cannot report that the HD 8400E is a certain percentage ahead of or behind any other GPU. The only comparative value is percentileVsAllGpus = 50, which is a database-wide rank, not a rival-specific delta. With no nearest rival records, it is not possible to construct the usual comparison paragraphs for this product.

The product’s lineage fields provide architectural context rather than performance comparison. Its predecessor is TeraScale 3 IGP, and its successor is GCN 3.0 IGP. The generation label, “GCN 2.0 IGP (Kabini Mobile),” identifies the mobile integrated segment. The avgBenchmarkScore of 0 and the empty benchmarks array mean there is no measured score to place next to any competitor, even if a rival were listed. The database record simply does not include the comparison metadata required to say where this GPU sits relative to a named product. Every prospective rival relationship is unspecified.

Ray Tracing and Feature Set

The rtCores field is null, so no ray tracing core count exists in the data. The tensorCores field is also null, so no tensor core count exists. Feature support is instead listed through APIs: DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.2.170. These API versions are recorded as supported. The hardware implements the GCN 2.0 pipeline with 128 shading units, 8 TMUs, and 4 ROPs.

The render rates are unchanged from the specification: 2.400 GPixel/s and 4.800 GTexel/s. FP32 performance is 153.6 GFLOPS. Manufactured at TSMC on 28 nm, the chip contains 1,178 million transistors on a 110 mm² die, with a density of 10.7M / mm². DirectX 12 (12_0) is the DirectX feature level listed; OpenGL 4.6 and Vulkan 1.2.170 accompany it. Since no RT or tensor core numbers are present, the feature set is defined by the traditional shader/texture/ROP rates and the three API entries, not by dedicated ray tracing or tensor hardware.

FAQ

Q: Does the HD 8400E support DirectX 12?

A: Yes. The API list includes DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.2.170.

Q: Does the HD 8400E have ray tracing cores?

A: The rtCores field is null, so no ray tracing core count is listed. The tensorCores field is also null.

Q: What is the TDP of the HD 8400E?

A: The TDP is 25 W. No suggested PSU and no power connector are listed.

Q: What is the FP32 performance?

A: The FP32 throughput is 153.6 GFLOPS, with a pixel rate of 2.400 GPixel/s and a texture rate of 4.800 GTexel/s.

Q: What is the memory configuration?

A: Memory size, type, and bus width are System Shared, and memory bandwidth is System Dependent.

Q: When was this product released?

A: It was released on 2013-04-22. The production status is End-of-life, with TeraScale 3 IGP as predecessor and GCN 3.0 IGP as successor.