What I said above is the clock rate is not of your key concern here: memory starvation & cache-size is. Equivalently clocked 4 core system will have worse per-core performance than a 2-core system.

Sure, if you heavily multitask and your tasks are all CPU-bound (some of which you mentioned, are not, by the way), then you'll have overall better system "snappiness" on a 4-core system. But each of those tasks will be running slower than otherwise.

If trying to swing things in your interests by means of crafty presentation makes you a "Stalin", what's to be said of the media who spun the entire issue exactly as much, but in the opposite direction?

It's almost impossible to find an objective source of information these days. It's all well-tuned to someone's very specific interests.

I admire the depth if your technical knowledge on the matter.

The irony here is that with most CPUs on sale right now, 4-6 cores have worse individual performance than a 2-core system. Many games need strong core-individual performance.

For example, a typical 4-core system performs 25% slower than 2-core.

If you enable hyperthreading you lose another 40% of your single-thread performance, as each core is split into two virtual threads (without optimization you'd lose 50% but thankfully CPUs are smarter than this).

So to recap, 4-cores: 25%. Hyperthreading: 40%. Total loss compared to a two-core system with no hyperhtreading: 55%

If you bought a 6-core system for gaming, the numbers would be even funnier.

Semi-knowledgable sounds like "knowing enough to be dangerous". I have counted at least four people who are friends, or I work with, who think more cores is better, without any regard for the type of task they'd use the machine for. Then surprise when their older machine runs their browser twice faster than their new expensive 4-core machine.

For most desktop tasks, which by their nature depend mostly on strong linear performance, two cores is the line, that, if crossed, you start to lose performance, rather than win one. Not everything can be parallelized. The problem is these cores still need to share access to your RAM. With 2 cores that works great, but 4-6 cores means you starve each core for memory bandwidth.

The rule is: do you plan to heavily crunch numbers in parallel, i.e. use the machine for: 1) a server 2) or encode/edit plenty of video 3) render 3D (GPU-driven games not counting, I mean professional rendering software that pushes the CPU)? Then you will benefit from 4/6 cores. Otherwise, you'll pay more and be hurt in performance.

Cores and clock speed are just two factors of CPU performance, and you need to consider also the cache size, core interconnect architecture, memory bus performance and so on. You may as well have a meaningless number as your model.

Your best best is to browse around for good benchmarks and see where the best price/performance ratio is, according to the types of tasks you need.

I just want to let the quotes say few things first:

I had the same exact observations few months ago when they demonstrated their new mobile OS for the first time. Looks like Microsoft's attempt at making an interface that's easier and more innovative than Android/iPhone ends up "complicated" and "cumbersome".

If their goal was to make a complex post-modern interface targeted to a small niche of geeks willing to get involved with such a taxing concepts as their scrolling clipped hub views, that'd be fine.

But they're achieving exactly the opposite of what they want, which is tragic. It means very likely their marketing will go "mainstream" and it'll be largely ignored, just like their Kin series (which, by the way have mostly the same GUI as Win 7 Mobile).

To quote Wikipedia on "disinformation":

Jobs hit all the right notes on both disinformation, and traditional propaganda in the span of one short presentation.

Emotional support: [talking to the press] "we make so many great products, I thought you guys trust us"; "we maybe shouldn't take it personally, but we do, and it really hurts us"; "we have worked out asses off to satisfy every last customer"

At rational level he tweaked and made those antenna video demos (also see http://www.apple.com/antenna/. He used reframing techniques to make the problem appear common in the industry, blurring the differences between the iPhone specific antenna issues and general signal attenuation.

I don't believe a word Jobs says. He has a long history of using these techniques to sell and brand his company, it's how the "reality distortion field" joke came to be.

But you gotta admit: he's so good at it, even when it's apparent he's tweaking facts and inserting little lies here and there, it's hard not to be sympathetic to his side. Which may be largely why he succeeds, even if many won't take his presentations at face value.

Sounds easy. Since the researchers apparently said Python/Perl/scripts and so on are easier to hack than C++, let's take typical factors in script execution.

You have an ASM-level operation that takes nanoseconds, and you have the following factors that take a random amount of milliseconds (millions of nanoseconds):

1) garbage collection 'jitter'
2) dynamic optimization of runtime (tracing, JIT, caching) and so on 'jitter'
3) parallel execution of thousands of tasks per second jitter (you measure just one random call of these at a time, you have no idea how loaded the server is or what's the structure of the tasks queue)
4) I/O calls in parallel delays jitter (same as 2. except much slower)
5) network jitter

Also I'm probably missing another 10 to 20 factors I could list above.

So, how many millions or billion of calls would I have to make to be able to "average" all those, factor them out and time some ASM-level operation in there?

And even once that's done, all I have is whether one character possibility I sent is wrong or not. Repeat for every printable ASCII character, and every single character of the password.

And still, you say, it's possible, though it'll take billions of calls.

Well, most services would deny you further checks after the 10 or so wrong logins and lock the account. Good luck using a timing attack on that service.

Sure, scripts are slower than C. They're slower in general, but when you compare two binary strings, it's still mostly the same C memcmp call that's being called. You also have semi-random events like mark and swipe garbage collection, dynamic optimizations, I/O delays.

This means scripts may be slower and more random, while the password check call still isn't, how the heck would that make is easier to hack scripted sites?

I'm not even mentioning the fact most web apps don't have passwords hardcoded, but actually do the check in an external agent, say SQL database, making his observation even less logical.

He claims he can filter out all those variables, including network jitter and pinpoint the exact time of some ASM-level operation on a string. I say, show us.

Someone is taking the phrase "Money talks" a bit too literally, huh.

Although, I've always wanted every time someone says "that's my 2 cents", that he's charged 2 real cents.

Ok, I'll use my vast imagination to imagine a bunch of geeks group together to sue Motorola because they can't root their phone. As if it was listed in the bullet list of features. But anyway.

Motorola allows rooting and now you imagine 4 months later a nice little virus, designed to root the phone and steal vasts amounts of private and business data off innocent users' phones.

Damned if you do, damned if you don't. But.. I forgot I'm imagining all this. People are just fine not being able to root their phone, so none of the above actually is about to happen.

So, wait, are you saying Opera is more secure? It's hard to imagine too many people with the time, inclination and skills, would volunteer to sift through the thousands of boring bad code on the Mozilla Addons site.

Dunno about what you said, but every time Bill Gates showed up in front of the camera, Milo would run and cry. Must be some bug.

If I had the game, first thing that pops to mind, is shooting a "Milo switches to Mac" ad.

Also, given Milo can apparently do homework, I'll just teach him C++ and let him work for me under threat of cutting the power off.

Staged demos aside, though.

I wonder how easy it'd be to interact with that thing in real life. If I asked him "How was your day" and all it can say back (in a distinct Stephen Hawking voice) is: "Dear aunt, let's set so double the killer delete select all"... that'll be just plain disappointing.

That's not the Uncanny Valley causing it. It's your basic animal fear that your own species has just been made obsolete by some basic gaming software.

But I might be wrong. You can test it. Get the DVD of Pixar's "The Incredibles" and play some. Do you feel the impending doom, or you're entertained? If the former, it's the Uncanny Valley (also check that with your doctor).

As for the fear-inducing-AI in the demo: it's fake, so you'll be able to sleep well tonight.

I love the detective work on detecting the subtle visual clues that it's fake.

I guess Microsoft (or anyone in the world at all) having casually developed this AI, speech recognition and a virtually flawless speech synth, solely for the purpose of making a casual role-playing console game, doesn't seem suspicious to anyone.