SPEC CPU2017 Platform Settings for HPE ProLiant Intel-based systems

Operating System Tuning Parameters

OS Tuning

Drive Write Cache

The Drive Write Cache is an option that can be enabled or disabled in the Array Configuration Utility, CLI version. The default value for the Drive Write Cache is set to Disabled, and in order to change this the Array Configuration Utility, CLI version needs to be installed. When the Drive Write Cache option is enabled on a Smart Array Controller in a system, it can allow the Smart Array Controller to help make drive writes more efficient.

Accelerator Ratio

The Accelerator Ratio is an option that can be set to different percentages (in 25% increments) in the Array Configuration Utility, CLI version. The default value for the Accelerator Ratio is set to 0% Read and 100% Write. In order to change this the Array Configuration Utility, CLI version needs to be installed. Changing the Accelerator Ratio allows the array installed on the Smart Array Controller to adjust how the priority of reads and writes.

ulimit -s [n | unlimited] (Linux)

Sets the stack size to n kbytes, or unlimited to allow the stack size to grow without limit.

vm.max_map_count-n (Linux)

The maximum number of memory map areas a process may have. Memory map areas are used as a side-effect of calling malloc, directly by mmap and mprotect, and also when loading shared libraries.

Zone Reclaim:

Zone reclaim allows the reclaiming of pages from a zone if the number of free pages falls below a watermark even if other zones still have enough pages available. Reclaiming a page can be more beneficial than taking the performance penalties that are associated with allocating a page on a remote zone, especially for NUMA machines.

Disabled unused Linux services via stop_services.sh script.

The following unused Linux services were disabled before the run in simple shell scirpt via the command "service {name} stop": abrt-ccpp, abrt-oops, abrtd, acpid, atd, auditd, autofs, avahi-daemon, cgconfig, cpuspeed, crond, cups, haldaemon, irqbalance, kdump, libvirt-guests, mcelogd, mdmonitor, messagebus, portreserve, postfix, rhnsd, rhsmcertd, rpcbind, rpcgssd, rpcidmapd, certmonger, lvm2-monitor, netfs, and sysstat.

Firmware / BIOS / Microcode Settings

Firmware Settings

One or more of the following settings may have been set. If so, the "Platform Notes" section of the report will say so; and you can read below to find out more about what these settings mean.

Intel Hyperthreading Options (Default = Enabled):

This feature allows enabling/disabling of logical processor cores on processors supporting Intel's Hyper-Threading Technology. When enabled, each physical processor core operates as two logical processor cores. When disabled, each physical core operates as only one logical processor core. Enabling this option can improve overall performance for applications that benefit from a higher processor core count.

Power Profile (Default = Balanced Power and Performance):

Values for this BIOS setting can be:

• Balanced Power and Performance: Provides the optimum settings to maximize power savings with minimal impact to performance for most operating systems and applications.
• Minimum Power Usage: Enables power reduction mechanisms that may negatively affect performance. This mode guarantees a lower maximum power usage by the system.
• Maximum Performance: Disables all power management options that may negatively affect performance.
• Custom:Allows the user to customize power management options independent of the defaults that each preset Power Profile sets.

Power Regulator (Default = Dynamic Power Savings Mode)

This option can only be configured if the Power Profile is set to Custom. Values for this BIOS setting can be:

• Dynamic Power Savings Mode: Automatically varies processor speed and power usage based on processor utilization. Allows reducing overall power consumption with little or no impact to performance. Does not require OS support.
• Static Low Power Mode: Reduces processor speed and power usage. Guarantees a lower maximum power usage for the system. Performance impacts will be greater for environments with higher processor utilization.
• Static High Performance Mode: Processors will run in their maximum power/performance state at all times regardless of the OS power management policy.
• OS Control Mode: Processors will run in their maximum power performance state at all times unless the OS enables a power management policy.

Minimum Processor Idle Power Core C-State (Default = C6 State):

This option can only be configured if the Power Profile is set to Custom. This feature selects the processor's lowest idle core power state (C-state) that the operating system uses. The higher the C-state, the lower the power usage of that idle state (C6 is the lowest power idle state supported by the processor). Values for this setting can be:

• C6 State: While in C6, the core PLLs are turned off, the core caches are flushed and the core state is saved to the Last Level Cache. Power Gates are used to reduce power consumption to close to zero. C6 is considered an inactive core.
• C3 State: While in C3 the core PLLs are turned off, and all the core caches are flushed. C3 is considered an inactive core.
• C1E State: C1E is defined as the enhanced halt state. While in C1E no instructions are being executed. C1E considered an active core.
• No C-states: No C-states is defined as C0, which is defined as the active state. While in C0, instructions are being executed by the core.

Minimum Processor Idle Power Core C-State (Default = C6 State):

This option can only be configured if the Power Profile is set to Custom. This feature selects the processor's lowest idle package power state (C-state) that the operating system uses. The higher the C-state, the lower the power usage of that idle state (C6 is the lowest power idle state supported by the processor). Values for this setting can be:

• C6 State: While in C6, the core PLLs are turned off, the core caches are flushed and the core state is saved to the Last Level Cache. Power Gates are used to reduce power consumption to close to zero. C6 is considered an inactive core.
• C3 State: While in C3 the core PLLs are turned off, and all the core caches are flushed. C3 is considered an inactive core.
• C1E State: C1E is defined as the enhanced halt state. While in C1E no instructions are being executed. C1E considered an active core.
• No C-states: No C-states is defined as C0, which is defined as the active state. While in C0, instructions are being executed by the core.

Minimum Processor Idle Power Package C-State (Default = Package C6(retention) State):

This option can only be configured if the Power Profile is set to Custom. This feature selects the processor's lowest idle package power state (C-state) that is enabled. The processor will automatically transition into the package C-states based on the Core C-states, in which cores on the processor have transitioned. The higher the package C-state, the lower the power usage of that idle package state (Package C6(retention) is the lowed power idle package state supported by the processor). Values for this setting can be:

• Package C6 (retention) State: All cores have saved their architectural state and have had their core voltages reduced to zero volts. The LLC retains context, but no accesses can be made to the LLC in this state, the cores must break out to the internal state package C2 for snoops to occur.
• Package C6 (non-retention) State: All cores have saved their architectural state and have had their core voltages reduced to zero volts. The LLC does not retain context, and no accesses can be made to the LLC in this state, the cores must break out to the internal state package C2 for snoops to occur.
• No Package State: All cores are in an active state and have not entered any power saving state.

Energy/Performance Bias (Default = Balanced Performance):

This option can only be configured if the Power Profile is set to Custom. This option configures several processor subsystems to optimize the processor's performance and power usage. Values for this BIOS setting can be:

• Balanced Performance: Provides optimum performance efficiency and is recommended for most environments.
• Maximum Performance Mode: Should be used for environments that require the highes performance and lowest latency but are not sensitive to power consumption.
• Balanced Power: Provides optimum power efficiency
• Power Savings Mode: Should only be used in environments that are power sensitive and are willing to accept reduced performance.

Collaborative Power Control (Default = Enabled):

This BIOS option allows the enabling/disabling of the Processor Clocking Control (PCC) Interface, for operating systems which support this feature. Enabling this option allows the Operating System to request processor frequency changes even when the server has the Power Regulator option configured for Dynamic Power Savings Mode.

For Operating Systems that do not support the PCC Interface or when the Power Regulator Mode is not configured for Dynamic Power Savings Mode, his option has no impact on system operation.

QPI Snoop Configuration (Default = Home Snoop):

This option allows the user to choose the snoop mode used by the processor and QPI (Quick Path Interconnect) bus. Snooping is a process where the individual caches monitor address lines for accesses to memory locations that have cached. Changing this option can have an impact on memory performance. The performance of various workloads will vary depending on the mode selected and the order in which they are listed here does not reflect any technical ordering or ranking. The order seen here reflects the order as currently seen in the BIOS. Values for this setting can be:

• Home Snoop: The snoop is always spawned by the home agent in this mode. This mode is best for workloads that are bandwidth sensitive and need maximum local and remote memory bandwidth. This mode provides high memory bandwidth in an average NUMA environment.
• Early Snoop: Since the snoop is sent from the caching agent, this mode has lower latency vs. Home Snoop mode. This mode is best for workloads that are latency sensitive, or workloads that benefit from fast cache-to-cache transfer latencies from the remote socket. This mode will decrease memory latency but may also result in lower overall bandwidth as compared to the other modes.
• Cluster on Die: This mode logically splits a socket into 2 NUMA domains that are exposed to the OS. Each NUMA domain has half the cores, last level cache and one home agent with an equal number of cache slices in each NUMA domain. A directory cache in the home agent is used along with in memory directory bits to determine whether a snoop needs to be sent from the home agent. This mode has the best local latency and local memory bandwidth and is best for highly NUMA optimized workloads. This mode reduces average last level cache hit and local memory latencies. Although selectable in unsupported configurations, Cluster on Die will be set and occur only when a system has one socket and two socket processors SKUs with two Home Agents (ex. E5-2600 v3 processors with 10+ cores). On an ProLiant Gen9 system, when selecting Cluster on Die with a processor that doesn't support this mode, the QPI Snoop Configuration will fall back to Home Snoop.

Thermal Configuration (Default = Optimal Cooling):

This feature allows the user to select the fan cooling solution for the system. Values for this BIOS option can be:

• Optimal Cooling: Provides the most efficient solution by configuring fan speeds to the minimum required to provide adequate cooling.
• Increased Cooling: Will run fans at higher speeds to provide additional cooling. Increased Cooling should be selected when non-storage controllers are cabled to the embedded hard drive cage, or if the system is experiencing thermal issues that cannot be resolved in another manner.
• Maximum Cooling: Will provide the maximum cooling available by this platform.

Processor Power and Utilization Monitoring (Default = Enabled):

This BIOS option allows the enabling/disabling of iLo Processor State Mode Switching and Insight Power Management Processor Utilization Monitoring.

When set to disabled, the system will also set the Power Regulator mode to Static High Performance mode and the Power Profile mode to Custom. This option may be useful in some environments that require absolute minimum latency.

Memory Patrol Scrubbing (Default = Enabled):

This BIOS option allows the enabling/disabling of Memory Periodic Patrol Scrubber. The Memory Periodic Patrol Scrubber corrects memory soft errors so that, over the length of the system runtime, the risk of producing multi-bit and uncorrectable errors is reduced.

Memory Refresh Rate (Default = 2x Refresh):

This BIOS option controls the refresh rate of the memory controller and may affect the performance and resiliency of the servers memory.

When set to 1x Refresh, the memory refresh rate will be decreased, the Power Regulator mode will be set to Static High Performance mode, and the Power Profile mode to Custom. This option may be useful in some environments that require absolute minimum latency.

When set to 3x Refresh, the memory refresh rate will be increased, the Power Regulator mode will be set to Static High Performance mode, and the Power Profile mode to Custom.

Last updated May 25, 2023.