ATI Xbox 360 S GPU

The ATI Xbox 360 S GPU is a console GPU from AMD, built on the TeraScale architecture and using the Xenos Vejle chip. TSMC fabricates the die on a 45 nm process; the 372-million-transistor die measures 168 mm² and reaches a transistor density of 2.2M / mm². The generation field is “Console GPU (Microsoft)”, production status is End-of-life, and the release date is 2010-06-18. The record has no series, codename, predecessor, or successor entries. There are no base, boost, or game clocks; the only clock figure is memory at 700 MHz / 1400 Mbps effective. The benchmarks array is empty, nearestRivals is empty, average benchmark score is 0, and percentileVsAllGpus is 50. Launch MSRP is 299 USD.

How It Compares

With no nearestRivals entries, no rival names and no deltaPct values exist to anchor comparisons. This is a positional record rather than a competitive one. The one positional number is percentileVsAllGpus: 50. That places the GPU at the midpoint of the database’s distribution of GPUs. The average benchmark score of 0 sits alongside it, but the empty benchmarks array means the score is not a measured result. The generation label “Console GPU (Microsoft)” is the only product-class context. Since no rival list is present, the data cannot support statements such as “ahead of” or “behind” a specific next-best card. The database has simply not associated this GPU with any adjacent entries, and no deltaPct values are available to translate into percentage deltas.

The absence of rival entries is itself significant. The database has no comparison row for this part, meaning there is no automatic nearest-neighbor ranking to inspect. The only rank-like field is the global percentile of 50, which is a broad placement rather than a head-to-head result. There is also no bus interface listed, so the record does not even indicate how this GPU connects to a host system. Those missing fields leave the comparison section limited to the percentile value and the generation context. A reader should therefore treat any competitive conclusion as unsupported by this FACT PACK.

Memory Subsystem

The memory subsystem is one of the most complete parts of this record. It lists 512 MB of GDDR3, a 128-bit bus, and 22.40 GB/s of bandwidth. The memory clock is 700 MHz with a 1400 Mbps effective rate. That effective rate is the transfer-level figure tied to the 700 MHz clock. For high-resolution workloads, the capacity and bus width matter more than the raw memory clock. A 512 MB framebuffer cannot hold an unlimited number of high-resolution render targets, and the 128-bit bus restricts how much data can cross between the GPU and memory. The 22.40 GB/s bandwidth is the resulting throughput ceiling.

This is a compact memory configuration. The 512 MB capacity is the only memory size listed, and GDDR3 is the only memory type listed. No alternative variants are present in the record. At high resolutions, the practical constraint is twofold: the framebuffer may run out of space for additional surfaces, and the 22.40 GB/s ceiling may limit how quickly those surfaces can be updated. The effective 1400 Mbps data rate provides the transfer speed, but it does not change the narrow 128-bit datapath. The combination of 512 MB, 128 bit, and 22.40 GB/s defines a memory subsystem aimed at a fixed console profile rather than a scalable PC configuration.

Who Should Consider It

Because the benchmarks array is empty, there is no score-based resolution/settings recommendation in the database. The numbers that are present describe the hardware ceiling: 240 shading units, 16 TMUs, 8 ROPs, 240.0 GFLOPS FP32, 4.000 GPixel/s, and 8.000 GTexel/s. These numbers suggest a part for a defined console workload rather than a broad PC product. The 1x HDMI 1.2 output limits connection to a single HDMI display. The API list is DirectX 9.0c (9_3), with no OpenGL or Vulkan versions recorded, so software that requires newer or alternative APIs has no listed support. Given End-of-life status and the 2010-06-18 release date, the likely audience is someone maintaining or studying hardware of that specific console generation. The data only supports lower-resolution, lower-settings expectations; no score evidence exists for high-resolution, high-settings use.

The fixed nature of the hardware reinforces that position. With 16 TMUs and 8 ROPs, texture and pixel throughput are modest relative to the FP32 figure of 240.0 GFLOPS. The record does not contain any measured application results, so it cannot identify playable frame rates or quality levels. A user who specifically needs this GPU’s exact output support—1x HDMI 1.2—and its DirectX 9.0c (9_3) feature set would be the appropriate audience. Users expecting modern API coverage or high-resolution performance would find no support in this database entry.

FAQ

Q: What chip is inside the ATI Xbox 360 S GPU?

A: The chip is the AMD Xenos Vejle, built on the TeraScale architecture. It is fabricated by TSMC on a 45 nm process with 372 million transistors, a 168 mm² die, and a transistor density of 2.2M / mm².

Q: How much memory does it have?

A: It has 512 MB of GDDR3 on a 128-bit bus. The memory clock is 700 MHz / 1400 Mbps effective, and total bandwidth is 22.40 GB/s.

Q: What are its compute and throughput rates?

A: The GPU has 240 shading units, 16 TMUs, and 8 ROPs. It lists FP32 at 240.0 GFLOPS, pixel rate at 4.000 GPixel/s, and texture rate at 8.000 GTexel/s.

Q: What is the power requirement?

A: The TDP is 133 W. The database does not specify power connectors, a suggested PSU, or slot width.

Q: What display and API support does it have?

A: There is 1x HDMI 1.2 display output. The API entry is DirectX 9.0c (9_3); no OpenGL or Vulkan version is listed.

Q: When was it released and what did it launch at?

A: The release date is 2010-06-18, the production status is End-of-life, and the launch MSRP is 299 USD.

Benchmark Performance

Benchmark performance is not represented in this record. The benchmarks array is empty, so there are no measured workloads, frame rates, or synthetic test indices. The average benchmark score is 0, which is the default value that appears when no scores are listed. nearestRivals is also empty, so no rival products and no deltaPct values are available. This means exact percentage deltas cannot be calculated from the FACT PACK. The only quantitative performance data are the fixed throughput rates: FP32 at 240.0 GFLOPS, pixel rate at 4.000 GPixel/s, and texture rate at 8.000 GTexel/s. These are ceiling figures, not measured results.

The global percentile of 50 places the record at the midpoint of the database, but without underlying benchmark scores, that percentile cannot be converted into a comparison percentage. Any statement such as “30% ahead” or “25% behind” would require deltaPct values, and none are present. The record also lacks base, boost, and game clocks, so the operating frequency range is not part of the dataset. That leaves the fixed rates as the only performance-related numbers. The FP32 throughput of 240.0 GFLOPS is the compute ceiling, while 4.000 GPixel/s and 8.000 GTexel/s are the pixel and texture ceilings. A reader should interpret this as a specification-only entry with no benchmark-based performance tier.

Power and Cooling

The only power figure in the record is TDP: 133 W. There is no suggested PSU value, so the database does not state a minimum power supply rating. There are no power connectors listed, and no slot width is specified. The physical size is listed as 260 mm / 10.2 inches in length, 270 mm / 10.6 inches in height, and 78 mm / 3.1 inches in width. Those dimensions define the envelope that a cooling solution must fit, although the record does not name a cooler. Since the production status is End-of-life, the power and cooling characteristics are fixed at the 133 W TDP and the unspecified PSU/connector fields. A user cannot derive connector or PSU requirements from this database entry.

The absence of a suggested PSU is a gap in the record, not a signal that no supply is needed. The 133 W TDP is the only power-related number, and it appears without a slot-width companion. The dimensions provide a physical footprint, but they do not describe the thermal solution itself. The record does not list a cooler size, fan configuration, or mounting layout. For anyone using this data to plan a system, the only actionable power figure is 133 W; the mechanical and electrical interface details remain unlisted.