Jukka Larja (roskakori2006@gmail.com) on 9/26/09 wrote:
---------------------------
>AM (myname4rwt@jee.male) on 9/25/09 wrote:
>---------------------------
>
>> It's pretty hard for me to believe that your
>> claims like "of course in-order system will
>> clock higher, everything else being equal"
>> or "Out-of-order adds complexity, so how
>> could it clock faster?" root in theory.
>
>David Kanter actually already answered this, but just to be clear, what of following do you dispute:
>
>1) Out-of-order adds more transistors[1]
>2) More transistors adds power
>3) More power equals lower clock

No, higher power doesn't *necessarily* mean lower clock in general. I suggest you should reread the thread, for some of this was probably discussed several times, and worse, my answers to your questions are likely to generate even more questions (which were likely discussed as well), which is rather useless.

So why don't you show the theory that formed your understanding above (I'm asking because you had referred to theory)?

>There is more, of course, like complexity in general, die size, signal propagation
>etc., but I think the power issue is the clearest one.
>
>And sorry, I don't have any references. These are such a fundamental claims that
>if you dispute them, I don't think there's any way for me to continue, with or without
>references, and we just have to agree to disagree. This isn't my area of expertise.

No, what you say is pretty far from being established truth (where are the examples I provided, btw? Deleted?) If, for some reason, you do believe it to be true, feel free to ask others if you happen to not trust what I say. If these "fundamental claims" are from theory you referred to, I'm curious to see the sources.

>[1] Not in general case, of course, but in everything else being equal case. You
>can always pick horribly complex in-order core, throw most of it away and add simple
>OoO to what is left, if you want make that point.

Yes, >xistors and >power. But >power is not necessarily lower speed.

>>> So, really, stop implying that I have some
>>> fanboyish agenda in this discussion. If any,
>>> I'm a fanboy of theory over practise.
>>
>> No implications; I don't take you for anything
>> else other than what your writing suggests
>> (and hardly anyone will). Theoretical fanboyism
>> in my books is not any better than any other
>> fanboyism, that's all.
>
>Please keep in mind that you're probably reading me wrong (at least seem so to
>me). You can blame me for that (for being a bad writer), but don't ignore it.
>
>As for implications, you're right. You didn't imply, but pretty much said it straight. Just didn't use the word fanboy.
>
>As for fanboyishim, in my book anyone who thinks he or someone else is not a fanboy
>of anything (e.g. claims to be completely objective) is delusional or lying. But
>I just probably used word fanboy wrong. Luckily everybody here understands that I'm not native English speaker.

A fanboy of theory is how you characterized yourself, so what's the problem with that?

>> I lost count of how many times he adjusted
>> his argument
>
>You see adjusted arguments, I see people having trouble understand each other.
>You do understand that there's nothing bad in "adjusting arguments" as such? Especially
>if your not writing in your own language, your bound to make mistakes and use inprecise wordings.
>
>I'm not going to go reading through this thread again, analyzing everything and
>pondering whether some "adjustment" was because of bad wording or because writer
>changed his mind but doesn't want to acknowledge that. What would be the point?

If you're not going to reread the thread, not going to refer to sources, what do you seek in this discussion then?

>> Of course OoOE is not bound to constrain clock
>> or result in inferior power efficiency as
>> exemplified by *real* designs (as opposed to
>> mysterious theory you refer to, which might
>> say something different). But if that's the
>> case indeed, I would strongly like to see some
>> references.
>
>So what's the deal with GPUs? I think you are referring to general purpose processors
>here. You do understand that I haven't? I don't think there is low-level theoretical general purpose anyway :).

Again: do you seek answers to your questions or want to hear something from me? I have already told you my view.

>>> I don't know the right number if that's what
>>> you mean. My comment was mostly a funny way
>>> (see, a smiley) to point out the difficulty
>>> of discussing practise.
>>
>> OK, so does it mean you have little interest
>> in tech side of the discussion?
>
>Well, I am interested, but I don't think there's any change of coming to some clear
>conclusion. I don't think we have enough of the right kind of data. If you are having
>trouble agreeing on number of pipeline stages, what change do you think there is to agree on low-level details?

Interesting. I didn't measure pipeline length myself, and I gather neither did Jouni. However, only one of us was making statements which were repeatedly found to be incorrect in this discussion. Who do you think it was? :)

>> Actually, so far noone has seen a *single*
>> reference to theory from you to support the
>> "in-order system will clock higher" and other
>> claims.
>
>Well, I didn't believe someone would challenge the basic truths (I finally wrote
>them down at the beginning of the message so we'll see if the problem lies there).
>This line of conversation feels a bit like someone asking me to post some proof
>that gravity exists. I wouldn't know what to do.

What you say is very far from being "basic truth".

>> A factless discussion is a territory of *bigots*,
>> theoretical fanboys would post some references.
>
>I actually just checked from a dictionary what bigot means and now I'm highly insulted.
>
>Just what do you think I'm being bigoty about? I honestly still have hard time
>believing that the facts are the problem. But if you disagree with the three I wrote
>to the beginning of the message, then I stand corrected and admit that the facts really were and are the problem.

Actually I didn't call *you* a bigot, I simply stated what they tend to do (more detailed explanation can be found in dictionary).

If you want a fact-based discussion, then contribute something; if you're a theoretical fanboy, show some studies; if you're bigot, this is not necessary.

>>Btw, it was stated by me at least once in this very thread
>
>Yes. That's exactly the problem (or one of the problems, if taken the view of whole
>thread). I can't fit all your statements into the same universe, ergo I don't understand
>what you mean or your not making any sense.

What exactly doesn't fit? It was stated long ago and I didn't change the argument.

>> or at least spend some time rereading the thread
>> instead of making others repeat the same to you?
>
>Re-reading or repeating the same doesn't help. If your not willing to explain from
>a different angle or perhaps just with different words and examples, don't expect me to suddenly see the light.

I think I answered clearly what you asked: Forgetting the broken logic in (3), the "waste" is no benefit in those special cases when static machine can execute given schedule as fast as one with dynamic scheduler would do.

The most interesting thing here is that whereas you accuse me of "not willing to explain" something, you simply delete those parts where *you* are asked to show some references from "theory" you keep referring to:

>Is it just that you find theoretical discussion
>insulting or are you actually saying that even on the low-level the theory is wrong.

Please show me "the theory" that states what you say. I'm very much interested in seeing the references.

>> I myself never felt a need for strict definition
>> of adjectives like "embedded" or
>> "general-purpose" (this is the first time I see
>> such question asked) and I doubt someone like
>> you can be satisfied even with an expert's
>> answer.
>
>Well, here I am grasping at straws. I was thinking that if you have some strict
>meaning for general-purpose (or some other term that, in my mind, doesn't have very
>strict definition), it might explain something.
>
>By the way, I would think that expert's answer would be something along the lines
>"there is no strict definition" which would pretty much satisfy me.
>
>See how wrong you have read me? There's a saying in Finnish, called Wiio's law,
>coined by a professor specialized to communications. It says that communication
>between people always fails, except when it by accident succeeds. I have found this
>to be pretty accurate describtion also in online world.

Does his theory take proper account of people whose primary goal is to be misunderstood? :)

>>> As for the straws, I don't feel like I need
>>> to catch any, which just underlines the fact
>>> that we are talking about different things
>>> here. Or maybe in a different way, since my
>>> different things theory doesn't seem to be
>>> working either.
>>
>> Theory again? :) So where is a single
>> reference or citation?
>
>Sorry. Not theory, hypothesis. Do English speaking people actually differentiate between these in casual talk?

Unfortunately I'm not aware of practical examples and my own practice doesn't extend that far to know the answer which would be theoretically correct. But since you're the man of theory (apologies, fanboy of theory), you might be able to find it somewhere. :)

>> Even more so when all one sees from a
>> "theoretical fanboy" is many words, zero
>> references.
>
>Well, I thought you wanted me to dispute something like number of pipeline stages
>in EV5 or greatness of OoO in 603.

Err... I don't want anything from you, you're the one who joined the discussion (when it was actually over), so... Did you consider finding another finished thread here and jumping in there as well?

If I would have thought you wanted me to prove
>more transistors => more power or OoO => more transistors, this conversation would
>have been different, though not necessarily any better.

I think it's more up to you to decide what you really want, I don't really care much. :)

If anything, I'd like to get an idea of what kind of books or papers (the "Theory") gave you the understanding that you have demonstrated. That would be interesting.

>-JLarja