HP ProLiant DL380p DDR3 memory technology - Page 13

Achieving best performance with DDR3 memory

Page 13 highlights

handling queued requests, and memory interleaving. Loaded latency gives a truer representation of the memory subsystem's capabilities in real-world environments. Table 5 compares the unloaded and loaded latencies of DDR2 and DDR3 memory in Intel-based 2P ProLiant servers. Table 5. Memory latency of DDR2 and DDR3 in Intel-based ProLiant servers Unloaded latency Loaded latency 2P ProLiant G5 DDR2 at 667 MT/s 126 ns 2P ProLiant G6 DDR3 at 800 MT/s 80 ns 147 ns 140 ns 2P ProLiant G6 DDR3 at 1333 MT/s 70 ns 100 ns Achieving best performance with DDR3 memory DDR3 memory can achieve significantly improved performance over DDR2 memory. With the new NUMA architectures, the manner in which DDR3 DIMMs are installed in the system is important. Maximizing system throughput The key to maximizing system throughput is to have as many of the system memory channels populated as possible. This helps ensure that the memory bandwidth of all the channels is available to system. With 2P ProLiant G6 servers based on the Intel Xeon 5500 series processors, this means installing a minimum of six DIMM modules, one in each memory channel. Minimizing memory latency Memory latency, particularly loaded memory latency, can be optimized by running at the highest data rate, which is 1333 MT/s for DDR3. For the new systems that are capable of supporting DDR3 at 1333 MT/s, achieving this memory speed is dependent on the number of and the rank of the DIMMs installed in each channel. Using balanced memory configurations For almost all application environments, the optimal configuration for DDR3 memory is to balance installed memory both across memory channels and across processors. Balancing installed memory across memory channels on a processor optimizes channel and rank interleaving, ensuring maximum memory throughput. Balancing the installed memory across the processors ensures consistent performance for all threads running on the server. If more memory is installed on one processor, threads running on that processor will have significantly better performance than threads on the other processor. A performance imbalance can result in degradation of overall system performance, particularly in virtualization environments. The white paper DDR3 Configuration Recommendations for HP ProLiant G6 Servers, downloadable from http://h20195.www2.hp.com/v2/GetPDF.aspx/c01750914.pdf, provides a detailed discussion of DDR3 memory configuration considerations for HP ProLiant G6 servers. 13

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14

13
handling queued requests, and memory interleaving. Loaded latency gives a truer representation of
the memory subsystem’s capabilities in real-world environments.
Table 5 compares the unloaded and loaded latencies of DDR2 and DDR3 memory in Intel-based 2P
ProLiant servers.
Table 5.
Memory latency of DDR2 and DDR3 in Intel-based ProLiant servers
2P ProLiant G5
DDR2 at 667 MT/s
2P ProLiant G6
DDR3 at 800 MT/s
2P ProLiant G6
DDR3 at 1333 MT/s
Unloaded latency
126 ns
80 ns
70 ns
Loaded latency
147 ns
140 ns
100 ns
Achieving best performance with DDR3 memory
DDR3 memory can achieve significantly improved performance over DDR2 memory. With the new
NUMA architectures, the manner in which DDR3 DIMMs are installed in the system is important.
Maximizing system throughput
The key to maximizing system throughput is to have as many of the system memory channels
populated as possible. This helps ensure that the memory bandwidth of all the channels is available to
system. With 2P ProLiant G6 servers based on the Intel Xeon 5500 series processors, this means
installing a minimum of six DIMM modules, one in each memory channel.
Minimizing memory latency
Memory latency, particularly loaded memory latency, can be optimized by running at the highest
data rate, which is 1333 MT/s for DDR3. For the new systems that are capable of supporting DDR3
at 1333 MT/s, achieving this memory speed is dependent on the number of and the rank of the
DIMMs installed in each channel.
Using balanced memory configurations
For almost all application environments, the optimal configuration for DDR3 memory is to balance
installed memory both across memory channels and across processors. Balancing installed memory
across memory channels on a processor optimizes channel and rank interleaving, ensuring maximum
memory throughput.
Balancing the installed memory across the processors ensures consistent performance for all threads
running on the server. If more memory is installed on one processor, threads running on that
processor will have significantly better performance than threads on the other processor. A
performance imbalance can result in degradation of overall system performance, particularly in
virtualization environments.
The white paper DDR3 Configuration Recommendations for HP ProLiant G6 Servers, downloadable
from
,
provides a detailed
discussion of DDR3 memory configuration considerations for HP ProLiant G6 servers.