This measurement provides a way to compare the energy efficiency of servers and determine the amount of power servers require at different levels of utilization. The
benchmark consists of measuring power consumption in 13 intervals at different levels of system utilization. First, the system runs three intervals (usually 240 seconds each) at maximum system capacity. These three intervals calibrating the benchmark for the 100% performance level of the system and are not counted in the final benchmark results. From there, the workload is throttled and runs intervals in 10% steps from 100% to active idle, measuring system performance and power consumption.
Additional info
SPEC CPU2006 represents rich compute intensive environments in mathematics and the sciences, with a small focus on business application workloads. The SPEC CPU2006 benchmark is intended to stress the computer processor (CPU), the memory architecture, the compilers, and the chipset/front side bus. SPEC CPU2006 uses workloads that represent real-world problems of R&D environments or a highly specialized field such as weather forecasting. Some of the more common applications are ray tracing, data compression, video compression, PERL language interpretation and XML processing. SPEC CPU2006 results are useful in the following situations:
- To compare disparate systems (different processor platforms for example)
- To see the performance impact of a specific hardware component (i.e. performance delta between two different processor speeds in the same server)
- When the target environment has a CPU intensive, likely scientific workload
- When there are no other benchmark results available on the systems that you want to compare
Additional info
The TPC-C benchmark simulates an Online Transaction Processing (OLTP) database environment. The performance of a system is measured when the system is tasked with processing numerous short business transactions concurrently. The TPC-C workload simulates a tiered environment wherein users interact with web pages to enter business transactions. The transactions and their percentage of the transaction mix are:
- New Order transaction (~45%): a new order entered into the database
- Payment transaction (~43%): a payment recorded as received from a customer
- Order Status transaction (~5%): an inquiry as to whether an order has been processed
- Stock Level transaction (~5%): an inquiry as to what stocked items have a low inventory
- Delivery transaction (~5%): an item is removed from inventory and the status of the order is updated
Transactions are entered by simulated users, business logic and queuing of the transactions are
handled by a middle tier server, and then the transactions are passed to the TPC-C database server for processing.
Additional info
The VMmark benchmark is intended to measure the performance of virtualized servers on a system under test (SUT) so that customers can compare the capabilities of different platforms for virtualization. VMmark represents the performance of virtual machines within a server running VMware ESX and a set combination of operating systems and specially tuned applications reflecting a typical datacenter environment. VMmark uses a collection of ‘sub-tests’ derived from commonly used load-generation tools as well as from benchmarks developed by the Standard Performance Evaluation Corporation (SPEC®). VMmark is an open standards effort that is agnostic toward hardware platforms and different virtualization software systems. VMmark uses workloads that represent common applications in datacenters. It is important to note that VMmark is designed to benchmark the performance of the virtualization software and the hardware, and is not designed as a benchmark of any other software component.
Additional info
TPC-H is a benchmark that simulates a Decision Support System or Business Intelligence database environment. The performance of a system is measured when the system is tasked with providing answers for business analyses on a dataset. These analyses include:
- Pricing & Promotions Analysis
- Supply & Demand Management Analysis
- Profit & Revenue Management Analysis
- Customer Satisfaction Studies
- Market Share Studies
- Shipping Management Analysis
The server system runs a read-intensive Decision Support System (DSS) style database to provide the results for the business analyses. The DSS database is designed to mimic a repository of commercial order-processing Online Transaction Processing Databases. The analyses are performed are 100GB, 300GB, 1000GB, 10,000GB, 30,000GB or even 100,000GB scale factor datasets.
The TPC-H benchmark runs the business analyses in two different ways – it performs a Power Test and a Throughput Test.
The TPC-H Power Test measures the query execution power of the system when connected with a single user. It runs the analyses in a serial manner – the queries and update functions run one at a time and the elapsed time is measured.
The TPC-H Throughput Test measures the ability of the system to process the most queries (and update functions) in the least amount of time in a multi-user environment. Each simulated user runs its own version of the TPC-H Power Test simultaneously. The number of simulated users for this test is selectable with a minimum number of users or “streams” detailed by the TPC-H specification for each scale factor.
Additional info
