ATI Radeon IGP 320

ATI Radeon IGP 320 is an integrated processor from AMD, built around the Rage 6 architecture and the RS100 chip. The database places it in the Radeon IGP (300) generation, fabricated on a 180 nm process with 30 million transistors in a 73 mm² die. The production status is end-of-life, and the successor is TeraScale IGP.

Benchmark Performance

The benchmark record for this product contains an empty benchmarks array and an empty nearestRivals list. The average benchmark score is 0, and the percentile field against all GPUs is 50. Because no rival names, rival scores, or deltaPct values are stored, the data set cannot support any "x% faster than y" statement. There is no measured score to rank, and there are no competitors to compare against.

The only quantitative performance figures recorded are the fillrates. The pixel rate is 160.0 MPixel/s, the texture rate is 160.0 MTexel/s, the TMU count is 1, and the ROP count is 1. These four values constitute the full numeric performance profile of the entry. With a single texture mapping unit and a single ROP, the record shows a pipeline where texturing output and pixel output are both listed at 160.0. The database does not provide FP32 or FP16 throughput values, and the shading unit count is null, so there is no shader-throughput figure to analyze.

The 50th percentile is the only placement marker, but it is uncalibrated within this record. An empty benchmark array means no application-level result supports that percentile. The average benchmark score of 0 reinforces the absence of recorded tests. In this state, the fillrates are the only data points that describe workload capacity; a workload that exceeds the 160.0 MPixel/s or 160.0 MTexel/s ceiling would hit the limit of the 1 TMU and 1 ROP. No rival data exists to say whether that limit is high or low relative to any other product, so the correct reading of the benchmark section is that this GPU is undocumented in terms of measured performance.

Ray Tracing and Feature Set

The RT-core field is null, and the tensor-core field is null. Accordingly, the record lists no dedicated ray tracing acceleration hardware and no tensor processing hardware. The architectural identifier is Rage 6, paired with the RS100 chip, and the generation field is Radeon IGP (300).

The API support list is DirectX 7.0 and OpenGL 1.4. Vulkan is absent from the data, so the software interface is limited to those two listed APIs. The feature set has no programmable shading-unit count in the database; the only execution resources specified are the single texture mapping unit and the single ROP. The die is fabricated on a 180 nm process, integrates 30 million transistors, and measures 73 mm², with a transistor density of 411.0K / mm². The display outputs are listed as Motherboard Dependent, meaning the physical connectors are determined by the host board rather than by this GPU itself. The slot width is IGP, confirming that the product is an integrated graphics solution rather than a discrete expansion card.

Without RT cores and without Vulkan, the record provides no path toward modern ray-traced workloads. The API ceiling of DirectX 7.0 and OpenGL 1.4 bounds the generation of software that can address the device. The feature set, as documented, is a small fixed-function integrated design from the early 2000s, built around the Rage 6 architecture.

Power and Cooling

The thermal design power field is null; the database does not specify a TDP for this product. The suggested PSU field is also null, so no power supply recommendation is recorded. Power connectors are listed as None, which means no auxiliary power connector is required. The slot width is IGP, identifying the unit as an integrated GPU that resides on the motherboard.

The absence of a TDP value and a suggested PSU is consistent with the integrated form factor. Because the power connector field is None, the part is not documented as drawing power through any external connector. The AGP 4x bus interface is the system connection, while display outputs are Motherboard Dependent. The 180 nm process, 30 million transistor count, and 73 mm² die size are the only physical data that could relate to heat, but the database does not convert those values into a thermal rating. Cooling hardware is not specified for this entry, and no temperature data appears in the record.

Who Should Consider It

The Radeon IGP 320 is an end-of-life integrated GPU communicating over AGP 4x. It uses System Shared memory, and the memory bandwidth is System Dependent. The API list is DirectX 7.0 and OpenGL 1.4, which defines the software generation this device addresses. Users considering this part would be working with an AGP 4x motherboard that has integrated graphics, and they would need to operate within the listed API level.

The documented performance limits are 160.0 MPixel/s and 160.0 MTexel/s, with 1 TMU and 1 ROP. Those numbers indicate a device that is not suited to unconstrained modern rendering loads. There is no benchmark data to support settings-specific guidance such as a particular resolution or quality level; the database does not contain any tested resolution or frame-rate results. What the data does permit is a scoping of the product: AGP 4x interface, DirectX 7.0 / OpenGL 1.4 feature level, shared system memory, and 160.0 fillrate ceilings. Because the memory subsystem is System Shared, performance will be influenced by the host system memory, but the record quantifies none of that influence. The part is also end-of-life, so the intended environment is legacy hardware rather than a new system.

Memory Subsystem

The memory size field is System Shared, the memory type is System Shared, and the memory bus width is System Shared. The memory clock field is also listed as System Shared. Bandwidth is System Dependent. This means the GPU has no dedicated VRAM, no independent memory bus width, and no fixed bandwidth number in the database.

The only fixed numeric rates belong to the pixel and texture pipelines: 160.0 MPixel/s and 160.0 MTexel/s respectively. Whether those rates can be sustained in practice depends on the System Dependent bandwidth, because the GPU shares system memory with the CPU. The database does not record a memory clock speed and does not provide a frame buffer size. The transistor density is 411.0K / mm², but that figure describes the die, not the memory path.

For high-resolution work, the absence of dedicated memory is the defining characteristic. Since bandwidth is System Dependent, the memory ceiling is whatever the host platform provides. The database contains no high-resolution test results, so the effect of this shared-memory design on pixel throughput at higher resolutions is not quantified. The record shows a device whose memory behavior cannot be rated from the GPU alone; it is a function of the surrounding system.

FAQ

Q: What is the production status of the ATI Radeon IGP 320?

A: The production status is End-of-life. The release date is 2002-10-04, the successor is TeraScale IGP, and no predecessor is listed.

Q: What APIs does the Radeon IGP 320 support?

A: DirectX 7.0 and OpenGL 1.4 are listed. Vulkan is not present in the record, and no RT cores or tensor cores are listed.

Q: How much dedicated memory does this GPU have?

A: It has no dedicated memory. The memory size, type, and bus width are all System Shared, and the bandwidth is System Dependent.

Q: Does the Radeon IGP 320 require power connectors?

A: No. The power connectors field is None. The TDP and suggested PSU fields are not specified.

Q: What is the manufacturing process of the RS100 chip?

A: The process node is 180 nm, with 30 million transistors, a 73 mm² die size, and a transistor density of 411.0K / mm².

Q: Is there a benchmark score for this GPU?

A: No. The benchmarks array is empty, the nearestRivals list is empty, and the average benchmark score is 0.