AMD vs. Intel

[Miquel]

I always get confused while selecting the processors. Let me tell you that I am having only some basic knowledge about it. I am posting here, so that I can get some help from your side. Please tell me which is better AMD or Intel? Please explain me with the detailed information. Any suggestions will be appreciable. Also want to know about the configuration that would be required for respective processors. I think that you might have got my point. Respond me as soon as possible.

[Rover]

Over the whole benchmark course ran a mixed picture: while the processor is a burst with and without Hyper-Threading under parallel load to the finish without any additional hurdle. With Hyper-Threading performance of this loss was, however, much lower. The problem is no place for the setting up of such a test environment that is often tainted by chance, chance in meaningful benchmarks but nothing that significantly more work required with respect to the clinical patch test environment that is wound in the program after the program down. In the last two years we have therefore limited ourselves after our first foray into the world of the parallel benchmark environment effectively prevented return to the serial tests, not least the ever more short-term provision of the test products by the manufacturer for a return to more extensive testing methods. Even if we are in the truest sense of the word actually occurred repeatedly funny if was restarted before benchmarks the system cumbersome system tools were disabled and one on its personal computer a few minutes later from the overloaded with programs Windows environment with ICQ, MSN, Outlook, mIRC, antivirus and firewall started a game, we forgot or rejected then our findings have finally - the time allowed for no other test method.

[Shutdown]

Hyper-Threading Technology with the Pentium 4 3.06 GHz introduced, it was clear that a feature which favors the parallel processing of multiple tasks that can not be tested conventionally. And so we moved test for test concluded that AMD's CPUs are undefeated in games currently in the lead and even in multimedia applications, an eternal Pentium best discipline, in part, against the blue giant landed and went around the world with virtually any web and print publication compliant. The approach contrary was also, it was now the final idea had, taken the trouble, time taken and and Intel processors through a benchmark parcours hunted AMD selected trying the real operating conditions to be found in the majority of readers and customers should be specific effects. In addition to the actual benchmarks were therefore mIRC, Norton Anti Virus 2003, ICQ Lite, ZoneAlarm, Firefox and Outlook Express on the test system. In addition, one has the usual resolution of 1024 by the next larger counterpart replaced with 1280 and also AF and FSAA recourse, as with current high-end cards is now no longer a problem. The fact that many players still emulate the frame rates in the triple digits, you could hold against this decision.

[Jevin]

At first glance, the results obtained from impressive and one is inclined to sign the conclusion of the HardTecs4U that currently well but an Intel CPU would be the better desktop Porzessor. In particular, Hyper-Threading help the Pentium 4 in the work environment to a victory over the AMD counterpart. Ultimately, this must be decided by then, but - as so often - the perspective from which we look at the results. This allows Intel in the course of HardTecs4U save in a few games a narrow victory against AMD and thus make the results of our clinical trials with 1024x768 completely on its head. Ultimately, however, all CPUs are tested within a few frames-per-second. Crucial to the flap away from the background applications seems much more likely to be too high graphics superb, the graphics card with great difficulty will put on the screen and must clear a bottleneck. Practical? Here, yes, there no. Like the Intel CPUs were cut off due to hyperthreading in lower resolutions, was unfortunately not considered.

[Howerski]

In the end, only the vertebral question of how the practical performance of a processor now defines and creates a benchmark course, the definition of justice is this. "Should we slow down the processor in games with high resolutions, because the average user runs this setting and the service should then be equivalent?", A more specific question might be. If you end up performance and utility of this power away? True masses of comparative values of various resolutions and, ultimately, graphics cards, memory configurations and motherboards (especially for processors that do not have an integrated memory controller varies the speed of the chipset to chipset) should be used for such purpose. You notice the perfect, practical test seems impossible, because the practice is much too diverse and you will be able to cover only one aspect in an ever so extensive review, which leads to a conclusion that is only as good as the way there allows. But awakened by the product of their colleagues and in memory of our own experiments with a more complex benchmark course we will also make us nevertheless once thought to this issue that will win in particular to introduce the first dual-core CPUs to explosiveness.

[Kane01]

One fact seems off much speculation today that that is already certain: If the time window between providing the test sample and the case is not the embargo changed in favor of the editors and troughs do not flood the monthly new announcements, many of the ideas and approaches in practice difficult to apply. Whether this is in the interest of manufacturers, may not - be questioned - especially in view of the promotional and prestigious dual-core CPUs. In other areas, it seems has always been welcome. Solutions can be found with certainty only in co-operation. A first approach or an extension of the traditional approach, but also emphasize that such an effort to realize time is not on. Ultimately, all roads lead well if it is their own definition to the target and not the perfect practical test embody. The only question is whether one can pick out the best way or a complex image at the end can be characterized only by referral to all paths.