INCA V7 Performance s And Tricks
Copyright The data in this document may not be altered or amended without special notification from ETAS GmbH. ETAS GmbH undertakes no further obligation in relation to this document. The software described in it can only be used if the customer is in possession of a general license agreement or single license. Using and copying is only allowed in concurrence with the specifications stipulated in the contract. Under no circumstances may any part of this document be copied, reproduced, transmitted, stored in a retrieval system or translated into another language without the express written permission of ETAS GmbH. Copyright 2013 ETAS GmbH, Stuttgart The names and designations used in this document are trademarks or brands belonging to the respective owners. Document Performance s And Tricks R02 EN 03.2013 2
Contents Contents 1 Introduction... 5 1.1 Definitions and Abbreviations... 5 2 PC Hardware Configuration... 6 2.1 INCA System Requirements... 6 2.2 Individual PC Hardware Components... 6 2.2.1 Processor... 6 2.2.2 Main Storage (RAM)... 7 2.2.3 Hard Disk / Solid State Disk... 7 2.2.4 Graphic Board... 8 2.2.5 Exemplary Performance Comparisons... 8 3 Software Environment... 10 3.1 Operating System... 10 3.2 Energy Saving Mode... 10 3.3 Hard Disk Data Encryption... 10 3.4 Impact of Other Applications Running in Parallel... 11 3.5 Virus Scanner... 11 4 Operation Hints for INCA... 12 4.1 Data Access from the Company Network... 12 4.2 ASAP2 Files... 12 4.2.1 COMPU_METHOD "FORM"...12 4.2.2 Errors and Warnings...12 4.3 ASAP3 Latency / Bandwidth... 13 4.4 Opening the Experiment Environment in Off-line Mode... 13 4.5 Dynamic Loading of Experiments... 13 4.6 Pre-configuration of Experiment Environment Instruments... 13 4.7 Number of Variables in the Experiment... 14 4.8 Borrowing of INCA Licences... 14 5 ETAS Contact Addresses... 16 INCA V7 Performance s and Tricks 3
Figures Figures Figure 1 - Performance comparison SSD / HDD... 8 Figure 2 - Performance comparison of different PC configurations... 9 Figure 3 - Performance comparison MS Windows 7 / XP... 10 INCA V7 Performance s and Tricks 4
Introduction 1 Introduction INCA supports engineers in the automotive industry to perform their measurement, calibration, validation and diagnostic tasks. These tasks come by default with high performance requirements for a PC system. For example huge description files like ASAP2 (ASAM MCD-2 MC) are used, a large amount of measurement data needs to be recorded and analysed, and a large amount of signals have to be monitored and visualized. During the development of INCA these performance requirements are taken into account, so that INCA can address the daily tasks of a calibration engineer on a standard Windows PC. However there are other environmental effects, which cannot be influenced by the design and the development of INCA, but have impact on INCA s performance. Those are: the PC s hardware configuration on which INCA is running; the software environment, under which INCA is running, i.e. the operating system and other software applications running in parallel to INCA; the individual usage and configuration of INCA by the user. The goal of this document is to provide some best practices and tips & tricks on how to deal with these environmental effects to get the best possible performance when operating INCA. As the given tips & tricks strongly depend on the individual used PC hardware and SW environment, they can only be seen as general hints. Therefore each hint has to be evaluated under the actual given environment of the user. So there might be hints, which bring performance improvements under the used environment and others which do not. 1.1 Definitions and Abbreviations AuSy Automation System CPU Central Processing Unit of a PC, shortly processor. GDI Graphical Device Interface GUI Graphical User Interface HDD Hard Disk Device OS Operating System. In this document only Windows operating systems are considered. SSD Solid State Disk INCA V7 Performance s and Tricks 5
PC Hardware Configuration 2 PC Hardware Configuration As for any software product the performance of INCA strongly depends on the PC hardware in use. This chapter provides you with recommendations which PC Hardware to choose when operating INCA. 2.1 INCA System Requirements For the use of INCA two generic PC hardware configurations are recommended. One defining the minimum required system, which is mandatory to run INCA, and a second, recommended system configuration, resulting in an acceptable performance for standard INCA use cases. For the current system requirements please refer to the Release s of your INCA version. The new Oscilloscope, which will be released in INCA V7.1, will require Graphic Boards supporting DirectX 9.29 or higher and equipped with sufficient RAM (256MB recommended) on board. 2.2 Individual PC Hardware Components The above given recommendations assume that there are no other resource consuming applications running in parallel to INCA and that the user is working with projects and experiments of medium size 1. If this is not the case, exceeding the specified system requirements helps to ensure a good INCA performance. The performance of a PC system can t be derived from the performance of each single hardware component but also depends on their smooth interaction. Therefore, before defining a standard PC configuration for a department or company it is recommended to evaluate the performance of such a PC configuration under real working conditions including INCA. 2.2.1 Processor The CPU is the calculation engine of your PC and therefore has high impact on the performance of INCA. However, the price/performance quotient for CPUs is not linear, which leads to the fact that choosing cutting-edge CPU models provides only a minor performance improvement at a significant higher price compared to state of the art CPUs. As a best practice it can be recommended to choose the second latest CPU core generation, which typically provides a good performance at a reasonable price. Multi-core CPUs: As of version 7.0 INCA supports single and dual core CPUs. In case of a dual core CPU, one core is used for INCA and the second one for the target server, the INCA component which is handling the low level HW access. INCA also runs on CPUs with more than two cores, however only two of the cores will be utilized by INCA. 1 Example of a medium-sized project and experiment: A project with ~15.000 variables, 2 ECU devices and an experiment with ~10 layers and ~ 500 selected variables. INCA V7 Performance s and Tricks 6
PC Hardware Configuration There are plans for INCA, especially for the target server, to support more than two CPU cores in the future. Mobile CPUs: For notebooks mobile CPUs are offered. They are optimized for low power consumption, resulting in longer operating hours under battery mode. However, mobile CPUs typically provide lower performance than standard CPUs. Mobile CPUs are only recommended when the user operates his notebook frequently for longer hours on battery mode. For use in vehicles, 12V DC/AC adapters should be considered as an alternative. 2.2.2 Main Storage (RAM) The main storage of a PC is used to store parts of the running operating system, running applications and their data in use. Compared to a hard disk the RAM has much faster access times. In case the PC runs out of available RAM, the operating system extends the RAM by swapping parts of the data to the hard disk. Due to the slower access times, this swapping has negative impact on the systems performance. Therefore it is recommended to equip the PC with sufficient RAM. Utilizable RAM under a 32bit OS A 32bit Windows operating system can only utilize a total of ~3GByte of RAM for the operating system and running applications. Of these 3GByte a single application can utilize up to 2GByte. Therefore extending the RAM to more than 3GByte brings no additional performance improvement for INCA. Utilizable RAM under a 64bit OS Under a 64bit Windows operating system far more then 3GByte of RAM can be utilized. As INCA is a native 32bit application it still can only use ~2GByte of RAM. However, under a 64bit OS INCA, the INCA target server, MDA and other applications running in parallel can each utilize up to ~2GByte of RAM (if there is sufficient RAM installed in the system). Due to the relatively low market prices for RAM it is recommended to equip a PC with 3GByte of RAM in case of a 32bit OS, and with at least 4GByte of RAM in case of a 64bit OS. 2.2.3 Hard Disk / Solid State Disk INCA can be installed either on a Hard Disk or on a Solid State Disk. Solid State Disk Typically the HDD (Hard Disk) is one of the slowest PC components and therefore can be a performance bottleneck of the system, especially when using huge data files. The performance of a hard disk depends on several factors (e.g. rotation speed, cache-buffer size, bus interface), which are out of scope of this document. But all the parameters result in two main key performance indicators: The Sustainable Data Rate and the (Data) Access Time, which can be found in the data sheet of the HDD. The storage capacity of the HDD is only limiting the amount of data which can be stored on a device, and has no impact on the performance. The storage capacity of state of the art HDDs is more than sufficient for INCA. INCA V7 Performance s and Tricks 7
PC Hardware Configuration When selecting a new PC, it is recommended to compare the HDD access times of different offers. Changing the HDD of an existing PC just for performance reasons makes less sense, unless more storage capacity is needed anyway. Solid State Disk The market prices for SSDs (Solid State Disks) have dropped significantly in the last years, and so they are a real alternative to HDDs today. ETAS internal comparisons have shown a 10% performance increase in average for INCA compared to a typical HDD. However, the actual possible performance increase heavily depends on the use cases performed under INCA (e.g. opening an experiment for the first time is 14% faster when using a SSD, but when opening the same experiment a second time, no performance increase can be observed due to caching algorithms of the OS). Figure 1 - Performance comparison SSD / HDD 2.2.4 Graphic Board As of INCA V7.0 the performance of the graphic board has almost no impact on the performance of INCA. Only the resolution and the color depth have to be considered, as they have impact on the appearance of INCA. For the usage of INCA almost any state-of-the-art graphic board can be used. ETAS recommends a graphic board with a resolution of at least 1280x1024 pixel, a 16 or 32bit color depth and at least DirectX 7 support. The new oscilloscope, which will be released in INCA V7.1, will not be executed on the PC s core, but on the graphic board. This requires a graphic board supporting DirectX 9.29 or higher and sufficient RAM (256MB recommended) on board. 2.2.5 Exemplary Performance Comparisons To provide a rough indicator what impact the PC hardware configuration has on the performance of INCA, ETAS compared two different PC configurations with the following results. INCA V7 Performance s and Tricks 8
PC Hardware Configuration Configuration 1 Configuration 2 Laptop Lenovo T61 Intel Core 2Duo CPU T7300@2.00GHz 1.96GByte RAM SSD Kingston SSDNow V+100 128GByte Windows XP Desktop HP Compaq 8100 Elite Intel Core i7 CPU Quad Core i7-860@2.80ghz 3.43GByte RAM SSD Kingston SSDNow V+100 128GByte Windows XP Configuration 1 is used as reference (i.e. 100%, explicitly shown only in the last line). The percentages given for Configuration 2 for the different use cases indicate the relation to Configuration 1. Any value below 100% means that Config.2 is faster than Config.1; the lower the percentage, the better the performance in relation to Config.1. Figure 2 - Performance comparison of different PC configurations INCA V7 Performance s and Tricks 9
Software Environment 3 Software Environment 3.1 Operating System As of version V7.0 INCA supports Windows XP 32bit, Windows Vista 32bit, Windows7 32bit and Windows7 64bit (32bit compliance mode). The impact of 32bit and 64bit operating systems on the RAM utilization has already been discussed in the chapter Main Storage (RAM) on page 7. Besides this effect ETAS internal INCA performance comparisons between Windows XP 32bit and Windows7 32bit have shown for Windows7 a ~10% higher performance in average over the tested use cases. Figure 3 - Performance comparison MS Windows 7 / XP 3.2 Energy Saving Mode To save battery power, some notebooks provide an option to reduce the performance of system components when the notebook is not connected to the AC power supply. This can have impact on the INCA performance as well, for example when the processor speed gets reduced when in battery mode. For maximum performance connect your PC to a power supply or configure the Windows Energy Saving Options in a way that maximum performance is granted even under battery mode. 3.3 Hard Disk Data Encryption For data security reasons, modern PCs provide an automated encryption of data stored on the HDD. This encryption forces the system to encrypt / decrypt any data access from/to the hard drive, which has negative impact on the system performance. The magnitude of the performance decrease depends on the PC hardware / BIOS configuration in use. In case of unexpected low system performance we recommend that you investigate the impact of the hard disk encryption under the PC hardware /BIOS configuration in use. If it turns out that the encryption is the root cause, use a hard disk password instead of hard disk encryption. However, as this might lead to a lower data security, such a step should be discussed with the company s IT department. INCA V7 Performance s and Tricks 10
Software Environment 3.4 Impact of Other Applications Running in Parallel Running other applications in parallel to INCA means that these applications have to share the system resources (RAM, CPU) with INCA. Depending on the applications in use and the PC configuration, this can have significant impact on the performance of INCA. This effect can easily be evaluated by monitoring the system resources in the Windows Task Manager. In case of a performance decrease due to running applications in parallel to INCA you have the following options: Reduce the number of applications running in parallel Should this not be possible you should consider increasing the overall system performance by upgrading the PC hardware configuration or single hardware components (see chapter Individual PC Hardware Components on page 6). 3.5 Virus Scanner Virus scanners can have a significant impact on the system performance. They typically offer three different scanning modes: Manual scanning of files, directories or entire partitions Scheduled scanning of files, directories or entire partitions Online scanning While in the first two modes the scanning is only performed when triggered by the user or at a predefined time, in the third mode every file immediately gets scanned when accessed by the system. As INCA accesses a lot of files when running, the impact of online scans on the performance of INCA can be significant. To avoid negative impact on the performance of INCA do the following: Perform manual and scheduled scans at times when you are not using INCA (e.g. during lunch break) Exclude the file paths of INCA from the online scans. Those are: C:\Program Files\ETAS * C:\ETAS * D:\ETASData * * The exact path may vary on your PC, as it can be configured during installation. INCA V7 Performance s and Tricks 11
Operation Hints for INCA 4 Operation Hints for INCA Besides the PC hardware configuration and additional applications running in parallel, the way of handling INCA and project files can have an impact on its performance. 4.1 Data Access from the Company Network The file access to network shares is in most cases significantly slower than the access to local hard drives. Therefore frequently accessed files should be stored on the local PC. Make sure that the INCA database is located on the local hard disk. Copy project files that are frequently accessed by INCA to the local hard disk. If necessary, configuration management tools (e.g. SubVersion) can help to keep the original project files on the network share and local copies on the local hard disk consistent. 4.2 ASAP2 Files Due to the increasing number of measurement and calibration variables in a project, the size of ASAM MCD 2MC files (ASAP2) has increased significantly in the last years. Therefore reading and parsing these files becomes more and more time consuming in INCA. However, there are ASAP2 elements, which are more critical for the parsing performance than others. These will be discussed in the following sections. 4.2.1 COMPU_METHOD "FORM" In an ASAP2 file the conversion formulas COMPU_METHOD can be defined by using the following types: IDENTICAL FORM LINEAR RAT_FUNCTION While INCA can calculate the formulas of the types IDENTICAL, LINEAR and RAT_FUNCTION online during operation, it has to convert the type FORM into a lookup table when parsing the ASAP2 file. This is a time and memory consuming task. Try to avoid COMPU_METHODS of the type "FORM" in your ASAP2 files. Often they can be replaced by the type RAT_FUNCTION. 4.2.2 Errors and Warnings When INCA detects ASAP2 constructions that lead to errors or warnings in INCA, a time consuming handling is necessary. Try to update / correct the A2L description file that provokes the errors and warnings. INCA V7 Performance s and Tricks 12
Operation Hints for INCA 4.3 ASAP3 Latency / Bandwidth Optimizing the data transmission between the Automation System and INCA consists of finding the best trade-off between bandwidth and latency. For this purpose INCA offers a user option which allows to tune the measurement data transmission from maximum bandwidth to minimum latency. Use the INCA user option Experiment / Measure / General / Cycle time for measure data polling [ms] to tune the measurement data transmission from INCA to the Automation System. For maximum bandwidth use 500ms, for minimum latency use 20ms. 4.4 Opening the Experiment Environment in Off-line Mode When opening the Experiment Environment INCA automatically searches for connected hardware and initializes it. However, there are uses cases, where you just want to access the Experiment Environment without the need for any hardware access (e.g. you want to prepare an experiment at your desk for later use in the vehicle). For such cases you can open the Experiment Environment in an off-line mode and skip the search for hardware and the hardware initialization. You can toggle between on-line and off-line access to the Experiment Environment by clicking on the experiment icon in the toolbar of the INCA Database Manager: Experiments will be opened in on-line mode Experiments will be opened in off-line mode Even if the Experiment Environment is opened in off-line mode, you can still switch explicitly to the on-line mode by pressing <F3> (initialize hardware). 4.5 Dynamic Loading of Experiments INCA V7.1 or higher loads experiments dynamically. This means that when opening the experiment, not all layers of an experiment are loaded, but only the active layer. Whenever the user switches between layers, INCA unloads the components of the current layer and loads the components of the layer selected by the user. This dynamic loading/unloading of layers results in a reduced consumption of Windows resources (memory, GDI resources) and a faster initial opening time of the experiment. However the dynamic loading of layers also leads to a short delay when switching between layers. In case a layer contains a high number of elements, this delay would become recognizable for the user. To avoid this, INCA does not unload layers with a high number of elements any more, once they were loaded. This minimizes the delay when activating this layer, but results in a slightly higher consumption of Windows resources. When working with huge experiments you can reduce the required Windows resources (memory, GDI resources) by distributing layers with a high number of elements to several smaller layers. 4.6 Pre-configuration of Experiment Environment Instruments All measurement and calibration instruments newly added to the Experiment Environment are opened with a pre-defined style (background color, font color, font size,.). Later on INCA V7 Performance s and Tricks 13
Operation Hints for INCA the style of each instrument can be adopted individually in the Properties dialog <ALT+RETURN> of the instrument or in the Display Configuration Window <F4> of the Experiment Environment. You can save configuration time by adopting the predefined style of each instrument type in the INCA user options to your preferred settings (e.g. the default background color of oscilloscopes). Use the INCA user options below Experiment / Measure and Experiment / Calibration to tune the default settings of Experiment Instruments to your personal preferences. 4.7 Number of Variables in the Experiment For displaying GUI elements, the Microsoft Windows operating system uses so called GDI objects. Microsoft limits the number of GDI objects which can be used by a single application to ~10'000 objects. Every GUI element in INCA requires a certain number of these GDI objects. When using experiments including a huge number of instruments it can happen that INCA runs short of GDI objects. INCA indicates this by a warning, and it is not possible to add further instruments to the experiment. You can increase the number of measurement and calibration variables which can be displayed in an experiment by using the instrument types "Measurement Table" and "Calibration Table" instead of "Measurement Windows" and "Calibration Windows", as the tables require less GDI objects than the classic windows. It is possible to increase the maximum number of GDI objects that a single application can use by editing the registry of the operating system. However this can lead to unexpected system crashes when the overall number of GDI objects for all applications is used up. Therefore ETAS does not recommend increasing the maximum number of GDI objects in the registry of the operating system. 4.8 Borrowing of INCA Licences INCA supports three different software license types: PC-based, machine-named (for a single computer) Server-based, user-named (administrated by a server and linked to a named user) Server-based, floating (administrated by a server, not linked to a specific user) While the licence file for PC-based licences is stored locally on the PC, the licence file for the other two licence types is stored on a licence server in the company s network. When starting INCA the validity of the licence is checked. As a network access takes longer than an access to a locally stored licence file, the validation of user-named and floating licences takes a little longer than the validation of a PC-based licence. Delays during the licence validation can also occur if the auto-borrow mechanism is used. When installing INCA you can define that the software licence shall automatically be borrowed from the server for a predefined period of time ( BorrowExpiryInterval ). Furthermore you can define how many days ( ExecuteBorrowAutomaticExtensionInterval ) before the expiration date the borrowing of the licences shall be renewed. If these two periods are almost equal, license borrowing will be executed frequently, without any specific need. Especially after holiday time (e.g. Christmas) this can lead to a high borrowing activity at the licence server and subsequent significant delays, as the INCA installations of all users will re-borrow their licence on the first working day. INCA V7 Performance s and Tricks 14
Operation Hints for INCA For more details about the configuration of the auto borrow mechanism please refer to the INCA Manual. INCA V7 Performance s and Tricks 15
ETAS Contact Addresses 5 ETAS Contact Addresses ETAS HQ ETAS GmbH Borsigstraße 14 Phone: +49 711 89661-0 70469 Stuttgart Fax: +49 711 89661-106 Germany WWW: www.etas.com ETAS Subsidiaries and Technical Support For details of your local sales office as well as your local technical support team and product hotlines, take a look at the ETAS website: ETAS subsidiaries WWW: www.etas.com/en/contact.php ETAS technical support WWW: www.etas.com/en/hotlines.php INCA V7 Performance s and Tricks 16
Index Index 3 32bit OS 7 6 64bit OS 7 A Applications Running in Parallel 11 ASAM MCD 2MC 12 ASAP2 12 ASAP3 13 B Bandwidth 13 BorrowExpiryInterval 14 C Calibration Table 14 Calibration Windows 14 COMPU_METHOD 12 CPU 6 D Dual Core CPU 6 E Energy Saving Mode 10 ExecuteBorrowAutomaticExtensionInterval 14 G GDI objects 14 Graphic Board 8 H Hard Disk 7 Hard Disk Data Encryption 10 HDD 7 I INCA System Requirements 6 L Latency 13 Licences File 14 M Main Storage 7, 10 Measurement Table 14 Measurement Windows 14 Mobile CPU 7 Multi-core CPU 6 O Off-line Mode 13 Operating System 10 P PC Hardware Components 6 PC Hardware Configuration 6 Processor 6 R RAM 7, 10 S Software Environment 10 Solid State Disk 7 SSD 8 System Requirements 6 V Virus Scanner 11 INCA V7 Performance s and Tricks 17