Pentium III: an immature mike?
Yield: from 0 to 100 (%) in only 10 tests
The previous paragraph was ending with an affirmation that will have seemed somewhat incredible to many, after seeing, reading, listening and even dreaming of the "shy" advertizing Intel campaign promoting the Pentium III. Since not at all, remitámonos to the facts... or rather to the numbers:
The 10 previous tests have been realized by the very same Intel, so we can be sure that they are almost objective, apart from avoiding me obtaining a mike for 100.000 pesetas only to certify that he does not convince me too much...
Let's analyze the results as the intention of every test:
Nevertheless, in tests optimized for the instructions SSE an incredible one reaches 73,7 % of progress; yes, an incredible increase, which we must put in his place. "3D Lighting and Transformation Test" is only a part of much more complex 3D WinBench 99. If we take the result of the finished test, as for example big Thomas Pabst does of www.tomshardware.com, we would see that the difference is only of 6,2 %... a number undoubtedly less attractive for a publicist.
To end we would have the classic synthetic tests, the CPUmark and the FPU Winmark. The first one measures the "brute force" of the mike in not hard mathematical applications not multimedia (the proper Windows, those of office computerization...) and the second one the brute force in mathematical operations of floating comma (for CAD or not optimized games). There are usually taken as indicative of the advanced of the internal technology that uses the mike, according to which 3,3 % and 0,4 % do not look like sufficient arguments to change Pentium "II" to "III"... but I do not know about marketing, skylight is.
The serial number, Internet and the Big Brother
Intel announces the Pentium III as a microprocessor "designed thinking about Internet". And what demons does that mean? Since they were demonstrating the previous tests, it must not be a speed increase, since a 1,3 % does not seem to indicate that they should have thought very much about Internet...
The big innovation for Internet is not, not any more nor less, that a serial number implanted in the microprocessor, which can be read by the Internet programs. It is supposed that this brilliant idea serves to impel the commerce in Internet, allowing a better identification of the user, although some questions might appear:
And the most important thing: Does it have any risk or additional problem? Since yes. The first one of them, the scarce or void safety of this method. Oh, Intel argues that it is not possible to read the serial number without we authorizing it, but in fact most of the people can be cheated to do it, apart from the fact that the safety configuration of the navigators is a cryptic and incomprehensible topic for the common one of the mortal ones...
... and the hackers are not normal users; for many it would be a very simple game the habitual barriers skip, to obtain the serial number and to use it as if it was proper. Some of them assure to know how already to do it. On the other hand, the privacy of the navigator would be minor: the track left by the network might be used to control the users of a lot of and immoral forms, worthy of totalitarian countries or of the Big Brother described by George Orwell in "1.984"... we are warned.
Clear it is that the serial number might be used, for example, to form the computers of a company of remote form or other similar questions of administration of networks... but for that it is not necessary to have a serial number; in fact it has been years since it has being done without problems.
To end with this topic, to say that we will obtain an increase in the Internet yield it cannot be considered any more than like publi-demagogy. The speed will be given fundamentally, how always, for the speed of the modem and the quality of the connection (provider, line saturation, Infovía...); the special only thing that we will be able to do with a P3 is to visit a few tens pages with spectacular worlds 3D, a little in what many people are interested... perhaps 1 % of the Internet users. And even for that, a Pentium II can be sufficient.