Xbox 360 Architecture

1
Xbox 360Architecture

• Presenter Ataç Deniz Oral
• Date 30/11/06

2
Overview

• The Xbox
• What kind of computation?
• Architectural details
• Decisions / Trade-offs
• Conclusion
• Discussion

3
The Xbox
4
Computation

• Decompression kernel
• Game World Geometry
• Data streaming
• also
• AI software
• Audio synthesis

5
Why not use a PC?
?Dot product implementation ?Support for D3D
formats
6
IBM PowerPC core
4 KB two-way set-associative BHT
SIMD Vector unit
Floating Point Unit
Fixed Point Unit
Load/Store Unit
7
The Cache
8
Decisions / Trade Offs

• Why multiple cores? (CMP versus SMP)
• Cost-effective!
• Enables shared L2 implementation (therefore
  reduces communication latency)

9
Decisions / Trade Offs (cont.)

• Shared L2 Cache
• To adapt to varying workloads
• i.e. Scene management vs. audio processing

10
Decisions / Trade Offs (cont.)

• In-order instruction issuance cores
• Simplifies logic ? Reduced die area ? Reduced
  cost and power consumption
• Out-of-order issuance requires
• Additional pipeline stages to meet clock period
  timing
• Rename registers and completion queues
• In-order instruction execution
• Claimed to be justified by two SMT (Symmetric
  MultiThreading) hardware threads per core

11
Computation

• Decompression kernel
• Game World Geometry
• Data streaming

12
CPU Data StreamingWrite Streaming

• Enable data streaming
• But do not thrash private cache or shared cache
• Write-through L1 caches

13
CPU Data StreamingWrite Streaming

• Enable data streaming
• But do not thrash private cache or shared cache
• Write-through L1 caches
• Uncached write gathering buffers in shared L2 for
  each core (for later dumping to FSB)

14
The Cache
15
CPU Data StreamingWrite Streaming

• Enable data streaming
• But do not thrash private cache or shared cache
• Write-through L1 caches
• Uncached write gathering buffers in shared L2 for
  each core (for later dumping to FSB)
• Cacheable write gathering buffers (for data
  transformation workloads)

16
The Cache
17
CPU Data StreamingRead Streaming

• Custom prefetch instruction
• ? separates read streaming from write streaming
• ? L2 cache is not thrashed

18
Conclusion
19
Discussion

• The End
• Any Questions?

20
References

• Application Customized CPU Design,
  http//www-128.ibm.com/developerworks/power/librar
  y/pa-fpfxbox/index.html, 2005.
• J. Andrews, N. Baker, Xbox 360 Architecture,
  IEEE Macro, vol. 26, no. 2, pp. 25-37, 2006.
• PowerPC Wikipedia, the free encyclopedia,
  http//en.wikipedia.org/wiki/Powerpc, 2006.
• Xbox 360 Architecture, www.hotchips.org/archives/h
  c17/3_Tue/HC17.S8/HC17.S8T4.pdf, 2006.