Fire Alarms Systems

enumeration is

Clarification

marks are for professional, right? Gordon Mah Ung disappears into the mouth of the Maximum PC Lab for days, and emerges to tell the world if the CPU or chipset next questions. I often crouched in the basement laboratory of the House of cases endless series of tests 3D games and graphics cards to find the perfect card for your low budget.

Sometimes, however, it is necessary to verify the performance of

your

System

. Maybe it seems to work slowly. Maybe you just got a new graphics card, double your installed RAM. So you want to make some quick performance tests to see if your system is actually slower than before, or faster than the upgrade.

What we do is to run the appropriate benchmark. The benchmarks are simply standardized methods for testing performance. Independent applications can be designed specifically to test the performance of a particular component (such as a graphics card) or the entire system. Another type of reference points using a real application in the test, but often only tell you how your system or component with an application behaves.

This is not a complete tutorial on how to run benchmarks to obtain reproducible results, if you want to know the truth about benchmarking methodologies, Gordon read the article on this about. Instead, we'll dive into the world of quick and dirty analysis: a test of the system as a quick way to see the impact of these changes, or as troubleshooting tools

addition, it is benchmarking on a budget. We will use benchmarks or applications available at the best price: free. In some cases, performance tests may be a smaller version of a more solid proof, if so, let's talk. Our journey by rail of reference were also divided into categories. CPU, graphics, storage and system tests

But before diving into the details of the individual, let's take a look at two main reasons for each day a user may want to run the benchmarks. Let's start, shall we?

something has changed

Something has changed

on your system. Maybe it was an intentional change on your part, adding more DRAM, fell into a second graphics card, and finally dropped a few coins for a hard earned and shiny new SSD.

Maybe you change your system, trying to overclock to the maximum stable configuration. In this case, you need to make a change at a time and do some benchmarks. To test for maximum stability, you can run several hours.

Or maybe the change is to be noticed. Your system suddenly takes longer to load applications, or is it just your imagination? Its frequency range of images from soft butter into pieces of mono-or maybe it's just that the game again?

Whatever the change, it is time to run some benchmarks. The correct reference point can say if this is really new SSD faster than the previous Raptor 10,000 RPM RAID 0. Or you could say that something went wrong with the graphics settings, if the 3D landmarks in the tank suddenly over a few weeks ago.

new reference

You just finished installing Windows 7, with 3542 updates. All drivers are updated, and new system works nice. What is the first application should install?

Why a reference point, of course. Better yet, install more.

Some say that you should install some type of dedicated application burn. However, I discovered over the years a good reference point, or sometimes several benchmarks

executed at the same time

burns are excellent applications.

also want to establish a performance baseline. To do this, you must run a reference system, a storage test and a reference point in 3D. According to the applications run more often, may be worth running the benchmarks that are replicated as these applications are the work of reference, if you're a PC player or a reference point if you like Cinebench is in modeling and 3D rendering.

you want to run the tests before you clutter your system with applications, many of which can run applets substantive or DLL files pre-loaded on system startup. Thus, to establish a baseline for the performance of your new system. So if you reach a point where the system seems slow, you can run the same tests and compare them with the original results. My general rule is that if my heavily loaded system, which has been active for six months or more, working less than 10% slower than the bottom line, I'm still good. If the overall performance than start looking for ways to delete items from the system.

Well, you have your basic performance measured and recorded. But what criteria should be used for reference? Let's look at a few.

---

benchmark system

system benchmarks typically generate a score which is a set of performance indicators, including CPU, memory, storage, graphics and tests that reflect the combination of individual sub-systems, such as reading video. There are many references available on the system, costs more or less, and pedigree.

benchmarks for the entire system, gradually I settled on the Futuremark PC Mark. Not perfect, but the free basic version of PCMark last 7 years of your system pretty well. I've used other tests, like Passmark and Novabench but none that really seem to trash the system. PCMark can be an accident if the PCMark test because something is not reliable with the system, not the reference point, as the reference point has been properly installed and maintained.

PCMark 7. The free basic version only works after PCMark and generates a single score.

PC Mark is considered a synthetic benchmark, it is not a real application, but is built with the actual recorded instructions for the use of integrated applications that are included with Windows.

If you want to upgrade to advanced, however, that costs $ 40. It's certainly cheaper than the cost of $ 250 for PC World a single-user version of PC WorldBench 6. The advantage of spending the $ 40 is also access to one of the best brands of disk-based applications at all. (The basic version does not provide access to the storage test.) PCMark 7 The storage test is based on the original reference RankDisk Intel, and uses the recorded behavior of real applications to negotiate the units.
Free Another useful test system is PC Wizard, the same people that bring you the CPU-Z, CPU and popular tool for system identification. PC Wizard also give a series of individual tests for the various subsystems that can be useful.

PC Wizard generates a graphical output based on a reference system.

There are other guides available free, as indicated, but most do not really perform a Windows 7 very well. Phoronix construct a benchmark that runs on Linux and Windows, but as with many similar projects open source, it requires a certain amount of manual work and writing scripts to make it work.

CPU and memory benchmarks

Sometimes it's worth trying just the CPU. Once realized that my system suddenly seemed to run slowly. After running a few tests CPU, and comparing it with my online tests, CPU seemed to work almost 50% slower. At first I suspected heat regulation, but the core temperature is about 40 degrees C at rest. Then I discovered that the updated BIOS restart I had moved to their lowest speed processor clock. That would never happen with a retail processor, the new BIOS did not know how to handle the sample spinning CPU engineering, and put the clock speed of the slowest known products.

• 
  There are several useful points of reference CPU specific interest to use. An old remedy is Prime95. Prime95 is a reference point rather geek, spitting a series of results for different sets of fast Fourier transforms and execution of different thread and plant before the end. It is probably more useful as a stress test. A test of torture that often work when burned in a system to perform a combination of Prime95 in blend mode when running the 3DMark 2006 benchmark level. If the entire system can run these two reference points in the concert for a couple of hours, the system is almost certain to be stable for gaming.

Prime95 is rare, but useful for testing torture.

wizard built into the processor test PC, above, is also a good test, rather quickly. It is so useful as evidence of torture, but sometimes you just need fast performance validation. I'm not a big believer in memory benchmarks. The results are often not translate well into actual performance. However, memory tests and latency in the PC Wizard, such as CPU are good tests to validate that your memory is made.

PC Wizard