CS 7810 Lecture 23 Maximizing CMP Throughput with Mediocre Cores J. Davis, J. Laudon, K. Olukotun Proceedings of PACT-14 September 2005.

CS 7810 Lecture 23

Maximizing CMP Throughput with Mediocre Cores

J. Davis, J. Laudon, K. OlukotunProceedings of PACT-14

September 2005

Niagara

• Commercial servers require high thread-level throughput and suffer from cache misses

• Sun’s Niagara focuses on: simple cores (low power, design complexity, can accommodate more cores) fine-grain multi-threading (to tolerate long memory latencies)

Niagara Overview

SPARC Pipe

No branch predictorLow clock speed (1.2 GHz)One FP unit shared by all cores

Thread Selection

• Round-Robin

• Threads that are speculating on a load-hit receive lower priority

• Threads are unavailable if they suffer from cache misses, long-latency ops

Register File

• Each procedure has eight local and eight in registers (and eight out registers that serve as in registers for the callee) – each thread has eight such windows

• Total register file size: 640! 3 read and 2 write ports (1 write/cycle for long and short latency ops)

• Implemented as a 2-level structure: 1st level contains the current register windows

Cache Hierarchy

• 16KB L1I and 8KB L1D, write-thru, read-allocate, write-no-allocate

• Invalidate-based directory protocol – the shared L2 cache (3MB, 4 banks) identifies sharers and sends out the invalidates

• Rather than store sharers per L2 line, the L1 tags are replicated – such a structure is more efficient to search through

Next Generation: Rock

• 4 cores; each core has 4 pipelines; each pipeline can execute two threads: 32 threads

Design Space Exploration: Methodology

• Workloads: SPEC-JBB (Java middleware), TPC-C (OLTP), TPC-W (transactional web), XML-Test (XML parsing) – all are thread-oriented

• Sun’s chip design databases were examined to derive area overheads of various features (primarily to evaluate the overhead of threading and ooo execution)

Pipelines

8-stage pipelinesScalar proc is fine-grain multi-threadedSuperscalar proc is SMTFrequency not more than ½ of the max ITRS-projected frequency

400mm2 die25% devoted to off-chip interfaces: mem controllers, I/O, clocking11% devoted to the inter-core xbar

Of the remaining area, 25-75%are allocated to cores/L2-cache

Area Effect of Multi-Threading

• The curve is linear for a while – study is restricted to such designs• Multi-threading adds a 5-8% area overhead per thread (primary caches are included in the baseline)• A thread is statically assigned to an IDP – multiple threads can share an IDP

Design Space Exploration

Single Core IPC

4 bars correspond to 4 different L2 sizes

IPC range for different L1 sizes

Aggregate IPC

C1: 2p4t with 64KB L1 cachesC2: 2p4t with 32KB L1 caches*L1 latencies are always constant

Maximal Aggregate IPCs

Maximal Aggregate IPCs

Observations

• Scalar cores are better than ooo superscalars

• Too many threads (> 8) can saturate the caches and memory buses

• Processor-centric design is often better (medium sized L2s are good enough)

PACT 2001 Paper on CMP Designs

• Different workload: SPEC2k (multi-programmed)

• Private L2 caches (no cache coherence)

Effect of L2 Size

Effect of Memory Bandwidth

Optimal Configurations

Title

• Bullet