Linus Torvalds: "Geekbench is SH*T"

[cytg111]

Linus Tourette Torvalds

Sure as hell glad Im not working with the guy.

 

[Ajay]

Yes, I also liked his articles on RWT. But it seems his love for Intel Itanium turned him into an angry and embittered man. He now keeps on bashing people that don't agree with him. Note that he still makes some interesting posts, but he somehow reminds me of people who were great, are no more great, and keep on talking about better times, failing to notice things changed. He recently basically wrote that everything had been done in compiler technology, how can a smart guy say that unless he's blinded by his hate or love for a company?

Thanks for the info! Wow, that's sad visa vi de Mone. I just read comments on the RWT forums. I haven't remained current enough on CPU architectures to have to add to the conversations.

The last CPU I studied in any detail was Intel's P4 (as part of a graduate course on computer architecture). As an embedded developer I needed in depth knowledge of the i960 and PPC 750 CPUs and made sure I knew more than most. As a hobby I read up quite a bit on Alphas since they were the coolest thing since sliced bread at the time. After seeing what Apple's A7 could do on less than 2 watts, I'm starting to read up on ARMv8, but I have so much I need to relearn that I don't see myself posting anywhere but here, and even then, not really on the deep down details of the architecture or any implementations (unless I really decide to put my mind to it for some time).

 

[Idontcare]

People, even great people, never become infallible. Put anyone on a pedestal and they are sure to, eventually, disappoint you in some manner.

Give these guys some room to make mistakes, and credit for getting on with their lives despite the unknown real-life challenges and dilemmas they are facing in the background of their forum posting.

 

[Ajay]

People, even great people, never become infallible. Put anyone on a pedestal and they are sure to, eventually, disappoint you in some manner.

Give these guys some room to make mistakes, and credit for getting on with their lives despite the unknown real-life challenges and dilemmas they are facing in the background of their forum posting.

Good points, thanks IDC.

 

[Loki726]

He recently basically wrote that everything had been done in compiler technology, how can a smart guy say that unless he's blinded by his hate or love for a company?

I'm constantly surprised by how much we are still learning about a field that has so much history. The application of perfect elimination orders to graph coloring was discovered as a theoretical result in 2003 (http://www.cis.uoguelph.ca/~sawada/papers/chordal.pdf), and the implications of this result are just starting to show up in compilers.

 

[beginner99]

Linus, what a legend.

http://www.youtube.com/watch?v=iYWzMvlj2RQ

I think this is funny. Personally I would love to be in a spot were I can just say fuck you all I will just say what I think. Just go look at the company were you work. At least half of the just plain suck, not as persons but at their job. However as a lowly employee you can't really say that too loud or at all but if you were the boss...

And in this video common, it's obvious he is actually playing with his reputation of bad behavior. He isn't really extremely agitated and out of control, the opposite is the case and I'm sure that was planned...

 

[Arachnotronic]

I think Linus is a funny guy. It's as though he just does/says what he wants to. Must be good to be on top like that.

 

[Nothingness]

I think Linus is a funny guy. It's as though he just does/says what he wants to. Must be good to be on top like that.

As long as you don't abuse your position, yes. IMHO he's been abusing it a few times, but as IDC says everyone is fallible

BTW Paul de Mone is back to defending Itanium on RWT, funny... or not.

 

[cytg111]

People, even great people, never become infallible. Put anyone on a pedestal and they are sure to, eventually, disappoint you in some manner.

Give these guys some room to make mistakes, and credit for getting on with their lives despite the unknown real-life challenges and dilemmas they are facing in the background of their forum posting.

Infallible. It is a basic human psychological construct, a feedback mechanism, if noone reminds you from time to time that you are, in fact, fallible.. Man will begin to think that he is infallible. I suppose you could say it is our fault for putting them on the pedestal to begin with but the fact remains that some of these guys needs a heavy dose of reality, reality as defined by the majority of two-legged-walkers.

 

[cytg111]

I think Linus is a funny guy. It's as though he just does/says what he wants to. Must be good to be on top like that.

Yea .. and once in a while you will encounter one of these groupies that acts talks and walks like their big idol, wanting to be all that .. when in fact they're shit, write poor code, poor social skills and dont get the big picture. Someone like Linus is a rolemodel .. and he just sets a f* poor example.. f* him .

edit : I admire the guy, agrees with many of his views, for example http://article.gmane.org/gmane.comp.version-control.git/57918 .. I am laughing/agreeing when reading something like that but it has to be with a humorous angle .. you cant say and *mean* idiot to someone in their face cause they prefer c++. You just cant.

 

[Dresdenboy]

So what some of us assumed all along is true. The Apple hype is...hype.

Apple simply creates the right "aura" around their products (no prob with that since I don't care), and has some strong quality control and requirements towards the look and feel. Since many people judge by appearance, this fits exceptionally well.

But what Apple doesn't use to sell mobile devices are CPU benchmarks. Instead they simply need faster CPUs and GPUs to meet their look'n'feel requirements at high screen resolutions. Or to improve Siri and whatnot.

So if a Coremark or Geekbench result shows very fast A7 cores, this is just an indication of the truth, and not a discovered guerilla marketing attempt.

I already mentioned in the past, that both Apple and Samsung are up to something with all those microarchitecture designers.

 

[ShintaiDK]

So if a Coremark or Geekbench result shows very fast A7 cores, this is just an indication of the truth, and not a discovered guerilla marketing attempt.

Yet its still not an indication of real world performance. So how can that be truth? :awe:

 

[Nothingness]

Yet its still not an indication of real world performance. So how can that be truth? :awe:

What is an indication of real world performance according to you?

 

[Dresdenboy]

Yet its still not an indication of real world performance. So how can that be truth? :awe:

Correct, that's why I didn't say real world performance.

The results show how fast the given cores run some specific small or large code, depending on compiler optimizations, platform, core microarchitecture, OS, power management, clock frequency, boost modes, and available ISA extensions. And unless the timer counters weren't accessed incorrectly, the results tell us something without being created using some PRNG.

But we don't know the correlations of these results with real life apps and user experience.