I3 processor (notebook) slower than Celeron 356 (desktop)

When buying my new computer, I opted for a notebook, after all in college they deactivated the computer lab and now only students have their own equipment to be able to study, and somewhat desperate in the purchase, I opted for one that had the Core i3 processor. The model is the Lenovo Z460, which I thought could replace my desktop, however this will not be possible yet.

I've had several desktop computers, as well as several notebooks and netbooks, and among them, an HP was my best decision, at the time it was a Sempron processor, but it worked like a ninja (and let's say better than Core i3, despite single core).

The Core i3 by the way, is a processor that really shows to be very robust, its platform is new and challenging. Integrating many components within a single package and providing memory and even video card control capabilities, Core i3 would be the most cost-effective for manufacturers, as it would dispense with the use of a bold chipset on a motherboard to make it work.

Although the Core i3 has 2 cores and HT technology, it has a serious problem in this line of Lenovo.

Have you ever thought about seeing the flash plug-in crash when watching a video in full screen? Do you know the game Citiville on facebook? You know that 3D game from the Windows 98 era that runs on any current PC without any delay? These are things that this new processor is unable to do.

Irritated by the portion of multimedia that went into space, I then tested the processing issue further, starting to create a C language program, containing the sum and division using floating point numbers. So I ran it on the notebook and the desktop computer.

Of course, the desktop, as a desktop, may even be unfair to challenge with Core i3 processing, but we are talking about a new architecture that promises to be faster, and the desktop that was used was 2: Celeron 356 with 3.33 Ghz speed, 512 Kb of cache and 2 Gb of RAM. The other desktop, was a Pentium Dual Core 2.2 Ghz model E2200 (first in the series before Core 2 Duo processors appeared) with 3 Gb of RAM.

The test consisted of running a program with a loop of 1 million accounts, and to our surprise, the program ran much faster on desktops, using the full load of 100% processor on Celeron, and 50% load on Pentium Dual Core. (This is due to the fact that the C language limitation works on only 1 core, after all there was no treatment for multi-core at the time).

To better understand how the processor utilization system works, a processor core is responsible for running the application in the appropriate programming language. Language C, because it is not object oriented, is a language, let's say "sequential", it must follow the steps one by one until reaching the final result. To be able to use the other cores, just open a program for each core, and the system will automatically address this software to the next core, and this is done automatically and completely transparent to the user from the Pentium Dual Core line.

To be more exact, to use, for example, 100% of a quad-core using C language, you must divide your program in at least 4 independent processes, which work by themselves, which can be called via shell also in independent ways. Only in this case is it possible to use the full potential of the processor.

Just 2 programs in C language to take care of the Pentium Dual-Core desktop to work 100%.

However, this does not happen in the new line of processors from Intel.

Core i3 behaved like no processor in the architecture behaved. The program ran extremely slow, and the big difference was in the use of the processor, which did not exceed 3% for each process.

It was possible to execute more than 60 processes with finite loops that were practically running without ending (I had to close the programs by hand, they didn't end).

Unlike the desktop, where the processing goes smoothly to 100% with some processes, in Core i3 it is quite different. Many processes are required to achieve a mere 30% processor load. Apparently, this suggests that the threads are reused by the other programs, decreasing the total processing load of the system, however the programs run extremely slower.

This on the one hand is good, and on the other hand it sucks. Perhaps it would be possible to use the entire system load with millions of processes running in parallel, or with an object-oriented system that creates countless instances for processing.

It is a relatively good system, let's say, for parallel processing between several programs, this would be very useful on servers, not on home computers, moreover, apparently, during the execution of 60 test programs, the processor had a minimum of 5% utilization , that is, it is practically not even used for its task itself.

I suggest that you have a limiter on the processor, where it limits the processing of each process, looking for threads in the cache that have already been processed, before going through the processor or any kernel effectively.

This makes the processing of each program much slower, to the point that a Windows 98 platform game, which runs normally on both desktops, practically does not run on Core i3, in addition to the video card deficiency.

Video acceleration must be enabled in order to be able to play the game, or else you will see a screen full of pink dots and the game's coloring is totally distorted.

Now we have a big dilemma: The flash plug-in works only in the first core, that is, Adobe did not create a flash suitable for parallel programming, that is, from now on, or we will see Adobe working hard to create a new system to use these features of parallel programming, or we will see many people complaining about the flash and ignoring it completely, avoiding installing it on their machines.

Taking advantage of the flash case, Youtube, the Google video portal is already preparing a version of the site soon using HTML5, that is, we will no longer need the flash plug-in to see our favorite videos. In addition, Google intends to create a video format exclusive to Google Chrome, making it more difficult to download and use the videos by third parties, but because these are open standards, I'm sure we will soon be able to use the new encoded videos just like the FLV format right at the beginning.

Despite the subject of Google, we have in mind that we have a wide range of options ahead, that this new platform is robust in parallel programming, but totally ineffective for the multimedia that the computer user needs day after day.