By: Linus Torvalds (torvalds.delete@this.linux-foundation.org), September 27, 2009 9:05 am
Room: Moderated Discussions
jeff (no@way.com) on 9/27/09 wrote:
>
>I don't. I understand its disruptiveness in the market, I
>was wondering more of the technical aspects of the A9.
Why do you glorify doing something new and stupid, when
doing good things well is what people really should
be admiring.
Now, I have to admit that I want to wait and see at how well
Cortex-A9 actually performs in real life, and maybe I'll
be disappointed. But the fact is, your kind of "why should
I be excited about the technology" is kind of stupid.
Doing something "new" is way overrated, while doing
something really well and making it available widely in
the market seems to be continually underrated.
Put another way: that "1% innovation" is overrated, while
the "99% perspiration" is sadly underrated. You really
should give them the same weight that Edison gave them:
anybody can have ideas, it's actually executing on them
that is the hard part.
The same way I'll take a basic boring x86 chip that just
does everything really well over some new "innovative"
architecture, I'll damn well take a well-executed Tomasulo
uarch over some fancy exciting new idea. The likelihood is
that the fancy new clever thing isn't actually that good
in practice.
So rather than be excited about things like transactional
processing etc, people should view them with a very healthy
dose of distrust. Hype is just that - hype. Why get so
excited about something that hasn't proven itself.
The fact that Tomasulo is 40 years old doesn't make it bad
or uninteresting. Quite the reverse. It makes it all the
more disappointing that you still have idiotic people out
there that try to do in-order crap, and even design their
whole architecture around in-orderness, when people have
known how to do better for so long.
Will Cortex-A9 beat the highest end? Obviously not. But I
suspect it will be reasonably competitive even in a very
small and low-power envelope, and be easy to design around
due to be compatible with older designs and likely not have
any horrible fragilities.
That should impress people. The people who create entirely
new architectures because they can't be bothered to do
things well - those people should be shunned.
Linus
>
>I don't. I understand its disruptiveness in the market, I
>was wondering more of the technical aspects of the A9.
Why do you glorify doing something new and stupid, when
doing good things well is what people really should
be admiring.
Now, I have to admit that I want to wait and see at how well
Cortex-A9 actually performs in real life, and maybe I'll
be disappointed. But the fact is, your kind of "why should
I be excited about the technology" is kind of stupid.
Doing something "new" is way overrated, while doing
something really well and making it available widely in
the market seems to be continually underrated.
Put another way: that "1% innovation" is overrated, while
the "99% perspiration" is sadly underrated. You really
should give them the same weight that Edison gave them:
anybody can have ideas, it's actually executing on them
that is the hard part.
The same way I'll take a basic boring x86 chip that just
does everything really well over some new "innovative"
architecture, I'll damn well take a well-executed Tomasulo
uarch over some fancy exciting new idea. The likelihood is
that the fancy new clever thing isn't actually that good
in practice.
So rather than be excited about things like transactional
processing etc, people should view them with a very healthy
dose of distrust. Hype is just that - hype. Why get so
excited about something that hasn't proven itself.
The fact that Tomasulo is 40 years old doesn't make it bad
or uninteresting. Quite the reverse. It makes it all the
more disappointing that you still have idiotic people out
there that try to do in-order crap, and even design their
whole architecture around in-orderness, when people have
known how to do better for so long.
Will Cortex-A9 beat the highest end? Obviously not. But I
suspect it will be reasonably competitive even in a very
small and low-power envelope, and be easy to design around
due to be compatible with older designs and likely not have
any horrible fragilities.
That should impress people. The people who create entirely
new architectures because they can't be bothered to do
things well - those people should be shunned.
Linus
Topic | Posted By | Date |
---|---|---|
Thoughts and questions on the Cortex A9 | Gabriele Svelto | 2009/09/26 02:46 AM |
Thoughts and questions on the Cortex A9 | none | 2009/09/26 03:27 AM |
Thoughts and questions on the Cortex A9 | jeff | 2009/09/27 05:06 AM |
Thoughts and questions on the Cortex A9 | Michael S | 2009/09/27 05:29 AM |
Thoughts and questions on the Cortex A9 | none | 2009/09/27 06:01 AM |
Thoughts and questions on the Cortex A9 | Howard Chu | 2009/09/27 10:39 AM |
Thoughts and questions on the Cortex A9 | Wilco | 2009/09/27 07:03 AM |
Thoughts and questions on the Cortex A9 | jeff | 2009/09/27 08:00 AM |
Thoughts and questions on the Cortex A9 | a reader | 2009/09/27 08:17 AM |
Thoughts and questions on the Cortex A9 | David Kanter | 2009/09/27 08:37 AM |
Thoughts and questions on the Cortex A9 | a reader | 2009/09/27 08:46 AM |
Thoughts and questions on the Cortex A9 | Mat | 2009/10/01 01:04 PM |
Thoughts and questions on the Cortex A9 | Wilco | 2009/10/01 06:09 PM |
Thoughts and questions on the Cortex A9 | anon | 2009/10/01 08:19 PM |
Thoughts and questions on the Cortex A9 | RagingDragon | 2009/09/28 05:11 PM |
Thoughts and questions on the Cortex A9 | Linus Torvalds | 2009/09/27 09:05 AM |
OOO hw vs SW&in-order hw | no thanks | 2009/09/27 04:47 PM |
OOO hw vs SW&in-order hw | Linus Torvalds | 2009/09/28 06:22 AM |
OOO hw vs SW&in-order hw | ? | 2009/09/28 11:37 AM |
OOO hw vs SW&in-order hw | RagingDragon | 2009/09/28 05:22 PM |
OOO hw vs SW&in-order hw | Megol | 2009/09/29 04:35 AM |
OOO hw vs SW&in-order hw | Anders Jensen | 2009/09/28 11:50 PM |
OOO hw vs SW&in-order hw | Linus Torvalds | 2009/09/29 07:44 AM |
OOO hw vs SW&in-order hw | Mark Roulo | 2009/09/29 09:58 AM |
OOO hw vs SW&in-order hw | Linus Torvalds | 2009/09/29 10:30 AM |
3- and 4-issue in-order CPUs | Mark Roulo | 2009/09/29 11:06 AM |
3- and 4-issue in-order CPUs | Linus Torvalds | 2009/09/29 11:29 AM |
3- and 4-issue in-order CPUs | Gian-Carlo Pascutto | 2009/09/30 12:35 AM |
3- and 4-issue in-order CPUs | Michael S | 2009/09/30 02:01 AM |
OOO hw vs SW&in-order hw | mpx | 2009/09/30 04:14 AM |
OOO hw vs SW&in-order hw | Pun Zu | 2009/10/02 02:44 AM |
OOO hw vs SW&in-order hw | none | 2009/10/02 05:22 AM |
OOO hw vs SW&in-order hw | Linus Torvalds | 2009/10/02 07:11 AM |
OOO hw vs SW&in-order hw | a reader | 2009/10/02 09:30 AM |
OOO hw vs SW&in-order hw | Linus Torvalds | 2009/10/02 09:59 AM |
Moorestown | David Kanter | 2009/10/02 10:59 AM |
What's the difference between Moorestown and Pine Trail cores? | anon | 2009/10/03 08:37 PM |
Moorestown | none | 2009/11/03 04:34 PM |
Moorestown | Anon | 2009/11/04 03:17 PM |
Moorestown | none | 2009/11/05 01:38 AM |
Moorestown | David Kanter | 2009/11/05 04:45 PM |
Moorestown | IntelUser2000 | 2009/11/06 04:17 AM |
Moorestown | Anon | 2009/11/06 01:51 PM |
Moorestown | none | 2009/11/07 07:07 AM |
OOO hw vs SW&in-order hw | Anon | 2009/10/02 07:55 PM |
Cluebat for graphics | David Kanter | 2009/10/02 09:19 PM |
Cluebat for graphics | Anon | 2009/10/03 05:45 PM |
Cluebat for graphics | David Kanter | 2009/10/04 01:57 AM |
Cluebat for graphics | Anon | 2009/10/04 08:15 PM |
Cluebat for graphics | David Kanter | 2009/10/05 03:09 AM |
Cluebat for graphics | Anon | 2009/10/05 03:36 PM |
Cluebat for graphics | David Kanter | 2009/10/05 09:54 PM |
Cluebat for graphics | Anon | 2009/10/06 05:58 PM |
OOO hw vs SW&in-order hw | Linus Torvalds | 2009/10/03 06:58 AM |
OOO hw vs SW&in-order hw | slacker | 2009/10/02 09:11 PM |
Linux graphics drivers | RagingDragon | 2009/10/03 08:27 PM |
Linux graphics drivers | anon | 2009/10/04 07:15 AM |
Linux graphics drivers | none | 2009/10/04 10:12 AM |
Thoughts and questions on the Cortex A9 | jeff | 2009/09/27 06:31 PM |
Thoughts and questions on the Cortex A9 | someone | 2009/09/27 09:30 AM |
Thoughts and questions on the Cortex A9 | none | 2009/09/27 10:09 AM |
Thoughts and questions on the Cortex A9 | Wilco | 2009/09/27 11:35 AM |
Thoughts and questions on the Cortex A9 | someone | 2009/09/27 11:55 AM |
Thoughts and questions on the Cortex A9 | Wilco | 2009/09/28 02:08 AM |
Thoughts and questions on the Cortex A9 | someone | 2009/09/28 05:58 AM |
Thoughts and questions on the Cortex A9 | none | 2009/09/28 06:18 AM |
Thoughts and questions on the Cortex A9 | someone | 2009/09/28 07:35 AM |
Thoughts and questions on the Cortex A9 | Wilco | 2009/09/28 08:25 AM |
Thoughts and questions on the Cortex A9 | Michael S | 2009/09/28 11:02 AM |
Thoughts and questions on the Cortex A9 | Wilco | 2009/09/29 01:35 AM |
Thoughts and questions on the Cortex A9 | Chuck | 2009/09/28 07:15 PM |
samples | AM | 2009/09/27 11:20 PM |
samples | Wilco | 2009/09/28 01:51 AM |
samples | AM | 2009/09/28 04:16 AM |
Shrinks and process tech | David Kanter | 2009/09/29 01:22 AM |
Thoughts and questions on the Cortex A9 | someone | 2009/09/27 11:42 AM |
Thoughts and questions on the Cortex A9 | none | 2009/09/27 12:52 PM |
Atom to stay in-oder or go OoO? | AM | 2009/09/27 11:09 PM |
Atom to stay in-oder or go OoO? | Ungo | 2009/09/28 05:34 AM |
Atom to stay in-oder or go OoO? | a reader | 2009/09/28 10:15 AM |
Atom to stay in-oder or go OoO? | anon | 2009/09/28 07:25 PM |
Atom to stay in-oder or go OoO? | AM | 2009/09/30 03:32 AM |
Atom to stay in-oder or go OoO? | baxeel | 2009/09/30 08:25 AM |
Atom to stay in-oder or go OoO? | AM | 2009/09/30 11:12 PM |
Atom to stay in-oder or go OoO? | Ungo | 2009/10/01 03:00 AM |
Atom to stay in-oder or go OoO? | AM | 2009/10/01 05:08 AM |
Atom to stay in-oder or go OoO? | anonymous | 2009/10/01 05:33 AM |
Atom to stay in-oder or go OoO? | AM | 2009/10/03 07:24 AM |
Atom to stay in-oder or go OoO? | Pun Zu | 2009/10/02 01:30 AM |
Atom to stay in-oder or go OoO? | Ungo | 2009/10/02 01:11 PM |
Atom to stay in-oder or go OoO? | AM | 2009/10/03 07:22 AM |
Atom to stay in-oder or go OoO? | Ungo | 2009/10/03 02:53 PM |
Atom to stay in-oder or go OoO? | AM | 2009/10/04 08:44 AM |
Atom to stay in-oder or go OoO? | David Kanter | 2009/10/04 11:02 PM |
Atom to stay in-oder or go OoO? | AM | 2009/10/05 07:18 AM |
Atom to stay in-oder or go OoO? | David Kanter | 2009/10/05 11:12 AM |
Atom to stay in-oder or go OoO? | AM | 2009/10/06 04:51 AM |
Atom to stay in-oder or go OoO? | anonymous | 2009/10/06 07:58 AM |
Do you have any proof? | David Kanter | 2009/10/06 09:58 AM |
Do you? | AM | 2009/10/06 11:30 PM |
Of course I do! | anonymous | 2009/10/07 05:58 AM |
Thanks :-) | AM | 2009/10/08 03:17 AM |
Thanks :-) | anonymous | 2009/10/08 05:52 AM |
Thanks :-) | AM | 2009/10/09 03:13 AM |
Thanks :-) | anonymous | 2009/10/09 06:03 AM |
Thanks :-) | Foo_ | 2009/10/09 06:47 AM |
Thanks :-) | AM | 2009/10/10 01:15 AM |
That's what I thought... | David Kanter | 2009/10/07 09:00 AM |
That's what I thought... | AM | 2009/10/08 03:26 AM |
That's what I thought... | anonymous | 2009/10/08 06:02 AM |
let's see... | AM | 2009/10/09 03:09 AM |
let's see... | anonymous | 2009/10/09 05:43 AM |
let's see... | AM | 2009/10/09 05:52 AM |
let's see... | anonymous | 2009/10/09 06:15 AM |
let's see... | AM | 2009/10/10 01:18 AM |
Atom to stay in-oder or go OoO? | someone | 2009/09/28 06:09 AM |
I call Troll | hobold | 2009/09/28 04:51 AM |
I call Troll | someone | 2009/09/28 06:15 AM |
OT: categories of motivation in a forum | hobold | 2009/09/29 06:01 AM |
Thoughts and questions on the Cortex A9 | Michael S | 2009/09/28 10:43 AM |
Thoughts and questions on the Cortex A9 | a reader | 2009/09/28 04:12 PM |
Thoughts and questions on the Cortex A9 | someone else | 2009/09/29 12:25 AM |
Why Cortex A9? | hobold | 2009/09/29 07:20 AM |
Why Cortex A9? | someone else | 2009/09/29 10:57 AM |
Why Cortex A9? | Richard Cownie | 2009/09/29 06:09 PM |
Why Cortex A9? | hobold | 2009/09/30 12:38 AM |
Why Cortex A9? | Richard Cownie | 2009/09/30 06:49 AM |
Why Cortex A9? | hobold | 2009/09/30 07:46 AM |
Why Cortex A9? | none | 2009/09/30 07:56 AM |
Marvell Sheeva and plug computing | Richard Cownie | 2009/09/30 09:03 AM |
Why Cortex A9? | Michael S | 2009/09/30 10:07 AM |
Why Cortex A9? | none | 2009/09/30 10:40 AM |
Why Cortex A9? | Gabriele Svelto | 2009/09/30 12:43 PM |
ARM architectural license | David Kanter | 2009/09/30 05:57 PM |
ARM architectural license | a reader | 2009/10/01 07:25 AM |
ARM architectural license | Richard Cownie | 2009/10/01 08:21 AM |
Why Cortex A9? | slacker | 2009/09/30 07:12 PM |
ARM architectural license | David Kanter | 2009/09/30 07:16 PM |
Why Cortex A9? | Michael S | 2009/10/01 07:45 AM |
Why Cortex A9? | slacker | 2009/10/02 02:41 AM |
Why Cortex A9? | Richard Cownie | 2009/10/02 10:28 AM |
Questions... | David Kanter | 2009/10/02 10:56 AM |
Questions... | Richard Cownie | 2009/10/02 11:29 AM |
Questions... | Wilco | 2009/10/02 01:05 PM |
Questions... | slacker | 2009/10/02 08:51 PM |
Why Cortex A9? | slacker | 2009/10/02 08:44 PM |
Why Cortex A9? | David W. Hess | 2009/09/30 08:42 AM |
Thoughts and questions on the Cortex A9 | Gabriele Svelto | 2009/09/28 01:28 AM |
Thoughts and questions on the Cortex A9 | Wilco | 2009/09/26 07:38 AM |
Thoughts and questions on the Cortex A9 | Gabriele Svelto | 2009/09/28 01:38 AM |
Thoughts and questions on the Cortex A9 | Costanza | 2009/10/01 03:45 PM |
Thoughts and questions on the Cortex A9 | sylt | 2009/09/28 05:54 AM |
Thoughts and questions on the Cortex A9 | Wilco | 2009/09/29 01:15 AM |