So a student is doing badly in a class, and really needs to Ace the final to get by.
The professor hands out the test, and the student is blown away. No possible way to pass this test, much less ace it. So the student thinks about it for a while, and eventually pulls out a $100 bill, attaches it to the test, and writes "$100 = 100 points = I pass" at the top of the paper. Having done so, the exam is turned in.
Next week, the exam is returned, and attached is a $50 bill and a note that says "$50 = 50 points = you fail".
Were you that confident that the libc/whatever other support libs were invoked were also bug free?
That the compiler didn't miscompile your code?
That the system calls invoked have no bugs?
Software bugs can crop up in lots of places other than the code you just wrote...and the farther you get from your own source, the more of a pain in the butt they are to find.
Was playing around with some of the apple ][ emulators floating about the other day, and I came across disk images of Strip Poker. As I recall, those types of games were very popular.
The resolution of a modern cellphone display must be on par with the old C=64 and Apple ][ stuff, why should pictures of naked people be any less popular in that medium?
Yes it is! In a modern office, computers end up taking a lot of power. Imagine your local server room. Don't you think they would like a 20% decrese in their power bill?
While I agree with the idea that saving power is a worth goal, asynchronous design isn't, by any stretch of the imagination, the only way to get there. There are plenty of ways to reduce power usage in synchronous designs; most of these methods are only just now being used as cooling becomes a bigger issue.
The StrongARM-1 ran at 160 Mhz and consumed under a half-watt of power in the worst case. That was about 5 years ago, in a.35u fabrication process. That's around a 90% decrease in power consumption from a 486dx4-100, which was actually slower at most computing tasks, and fabricated in the same process
The SA-1 aggressively used conditional clocking, ran at 1.6-2.3V, and generally made power consumption a high-priority design goal.
Asynchronous design may be a slight win in power consumption, but only because synchronous designs don't tend to worry about it, as a rule.
However, one thing that has struck me is that overclocking increases the prices of chips, on the whole. When someone buys a cheap chip, and then OC's it, they are not paying the huge
surcharge on the latest technology that everyone else has to, and so they are prolonging those inflated prices. Basic supply and demand, as outlined by Adam Smith, shows that this is irrefutably
the case.
Methinks perhaps you have an interesting view of supply and demand.
The huge surcharge on the latest technology is not some just some nice little premium the semiconductor manufacturer sticks on to recoup development costs. The way you talk about it, it sounds like you're saying the company adds this surcharge, and then, once they've gotten back their money for developing the high end part, they'll kindly drop the price, because they're nice folks, after all.
The price of the part reflects what the market will bear, assuming there is reasonable competition. In the x86 processor market, this wasn't the case on high end parts until fairly recently, and as a result, those high end Xeon parts and the like routinely sold at *massive* profit margins.
With reasonable competition, we fall back into the supply/demand relationship nicely. OCing only reinforces this natural market condition; by running a chip at a faster speed grade than it is rated" you lower the supply curve slightly and push down prices.
I think most OC'ing is pointless; when there's reasonable competition in the market, chips do tend to get sold at the grade at which they are reliable for normal use. There are exceptions, though, like Intel's Celeron 300A. The 300A was an attempt by Intel to artifically split the market to extract more profit from it; OC'ers actually help defeat this corporate market manipulation.
The question is one of morals. Myself, I have no particular problem with it. But many people may rightly regard overclocking as cheating, with some good reason. I am happy to admit that I am a
cheater - I don't give a shit.
I'm surprised this is even comes up as an issue. When you buy a chip, you buy a piece of hardware. You've no moral obligation back to the company to use it as they intended or desire. You bought it, you can use it as you like; if they're depending on you not OC'ing so they'll get more money, you're under no obligation, legal, moral, or otherwise, to support that.
Now resellers who OC systems without full disclosure to their customers are a whole different beast.
Not mega-steroidal processing power, not nonexistant processing power. Enough to saturate a 100Mbps link with some amount of processing work. (Say, URL scanning for web caching, firewall packet rewriting). We're talking 486 at 100Mhz; a StrongARM-1100 @ 233Mhz would be nice.
Low Power Consumption
I don't want to have to actively cool the unit; the unit should be whisper-quiet, and shouldn't feel like a miniature furnace.
1 100Mbps RJ45 out standard
This isn't a PDA, it's an appliance doing useful things on the network.
Additional 100Mbps RJ45 out optional
Gotta be able to firewall with it. And while we're at it, let's NAT/DHCP/DNS Cache and webcache. Be I need that second port!
Serial port
Need a console, don't want to put it on a monitor. 'nuff said.
Connector/mounting for 1 hard drive, preferable a portable
Disk space is cheap. If I'm doing any serious caching/serving, I'm going to want lots of it. So give me somewhere to put in my own hard drive. Flash is too expensive and not up to a server duty cycle. I'd prefer a laptop hard drive because
they tend to be quieter, less power hungry, and more shock resistant.
Space for lots of standard memory
I want 2 standard DIMM slots. I want to be able to blow $200 and put 512MB of ECC SDRAM in there, so that I can keep the aforementioned hard drive spun down 90% of the time.
Sub $200 price tag
For the bare unit. No Hard drive, no memory, I'll add those myself. It's not that much hardware, really.
That's my wish list for a linux appliance. There are similar products in existance (The Cobalt Qube springs to mind) but they are overpowered and overpriced for individual use. I've also looked through a fair bit of PC/104 stuff, and found one product that would have done all of the above...for $1200.
When someone does all of the above in a good, high-quality unit, I'll jump in line to buy one.
Similarly, at the University of Michigan, there is a senior level class where students implement a limited version of the Alpha 21264 in verilog. It is pipelined, and even superscalar.
I'm having a hard time believing this. Just seems a bit too...ambitious for a senior level class. Seniors can be smart cookies, but the 21264 (EV6/EV67/EV68) is is 4 way OOO/full register renaming implementation of the Alpha architecture. It was a several-year design cycle for a large number of very smart engineers.
Perhaps you meant they implement a limited version of the Alpha architecture in verilog?
Wow...PureFiction is doing a heck of a marketing job for this, if nothing else. I can remember at least 4 comments (threshold 2) from that account on the Gnutella will never scale discussion which promoted this system.
I don't know what the technical merits are, but the marketing is solid!:)
The mips tree has its own CVS repository which is the most current mips stuff (oss.sgi.com). That one is maintained primarily by Ralf Baechle.
Similarly, the sparc tree's most up-to-date stuff can be found at the repository David S Miller maintains on vger.
In addition to being the gatekeeper for the official tree, Linus is pulling double duty as the portmaster for the x86 port. Thus, the x86 stuff is always up-to-date in the official tree, and the other archs tend to have some lag time associated with them.
This is nothing new. It's just symptomatic of the hierarchical Linux development style. {Free|Net|Open}BSD don't tend to suffer from this due to their use of a central CVS repository with all portmasters having access to their relevant parts. Whether this is a better
system is left for others to flamewar about, but it does prevent the port floating the author is talking about.
But that would be contrary to the general/. methodology of Comment First, Read Second if you feel like it and have nothing better to do.:)
But seriously:
He explicitely said that these benchmarks are not scientific and are merely his opinion. Therefore, this error is permissible
May be permissible to you, but I find such basic errors disturbing. It makes me wonder what he considers to be "linux tuning", and how equivalent his setups really were.
Besides, this optimization depends highly on the compiler used. My
understanding is that gcc does not do as good a job optimizing as some other compilers
GCC gets this one right, regardless of architecture, unless you explicitly tell it to do no optimizations. (-O0). Check it out...compile that program with gcc and objdump -d it; you'll see no FP ops at all.
Furthermore, even if this stuff did get optimized out, I don't see how it would change the outcome of this
comparison.
It probably just skewed the results more towards process creation/reaping, which is appropriate, given this is supposed to be an OS benchmark. But the fact that he added such pointless code shows he doesn't have a really good grasp of what systems and areas he's stressing (or trying to).
All in all, this is the best comparison between FreeBSD and Linux I've seen yet. Everything else I've heard up until now was basically anecdotal evidence and hearsay. Yes, this is not a real
benchmark. But at least it shows some evidence.
It's better than a random "FreeBSD Rulez" comment on usenet or equivalent, but not by much, in my book.
The overall conclusions may be valid; I don't know. I've used NetBSD for a few things here and there, but don't have enough experience with FreeBSD to make any sort of judgement.
This, however, caught my eye:
int x;
long y;
y = 28.2839281;
x = 339829;
y = x / y; ...
Notice how I included some simple floating point arithmetic in the C
program to make things just a tad tougher.
He admits he's no benchmark specialist, but any compiler worth its salt (and many that aren't) will optimize the floating point operations away. Also, since the result of the divide is never used, that will be optimized out, too.
I don't know what the real story is, and I do know a lot of knowledgeable people split on the Linux vs. FreeBSD issue. However, such a blatant error in benchmarking methodology gives me large doubts about this guy's credibility as a competent judge.
Re:Stormix down, but not out - read the fine print
on
Stormix Bankruptcy
·
· Score: 1
What a sick notion. You make it sound like the creditors are evil, greedy attackers rather than the people and institutions that trusted Stormix and loaned them money. What would you think if you
lent someone money and they went to court to try to keep from repaying you? Would you view the court as a "protector"?
Here's to reviving debtor's prison! Let's shift more of the burder in loans to the lendee, further skewing the financial balance of powers!
Sarcasm aside, bankruptcy exists for a very good reason. Without it, freedom can be irrevocably linked to financial health. Any creditor enters into a financial agreement with trust, but also with the knowledge that *any* financial investment involves risk.
It's a choice to lend money. And sometimes you lose.
The bankruptcy system is sometimes abused (see home ownership rules in Florida, recently, and default rates on law student loans), but on the whole, I'm glad it exists...
That buying a Mac at CompUSA and Sears et al. is an exercise in futility and frustration. If you go to a car dealership, they don't steer you away from the model you ask for to show you another brand and try to bullshit you while they do it.
Is it just me, or is this a bad example? EVERY time I've been looking for a new car, I've had salesmen try to steer me to something more expensive/with better margins for them. It's a problem in more industries than computer sales.
I think Apple should sell on the Web and exclusively through its Mac retailers like MacZone, MacMall.
Why? Should Compaq/Dell only sell in a guaranteed monopoly environment, because someone coming in might be tempted to a machine from another vendor? Seems like if you're blaming the competition for taking away your customers, the solution isn't to take away the competition, it's to make your product demonstrably better and do a better job of telling the world why you are better
Unfortunatly, the lessons learned from recent market leaders seem to be summarized in "it's the marketing, stupid", so maybe you're right...
to put it simply, when an application is optimized for the G4, it wollops the Pentium III.
This is exactly the myth Apple is trying to force down your throat, and you're eagerly swallowing.
Altivec is a specialized instruction set that speeds up vector ops. To claim that a benchmark is not valid because it's not "altivec-enhanced" is roughly equivalent to saying "Oh, I didn't hand-optimize that application is assembly language for this platform, so it's not a valid benchmark"
One of the things spec tests is compiler performance, because compiler performance is an integral part of system performance. If you haven't got compilers that generate good, fast code for high level language constructs, you're dead in the water as a platform.
I'd bet there are some fp loops that altivec (or MMX on X86, MDMX on MIPS, MVI on Alpha, etc) could speed up significantly. Possibly an order of magnitude or more. I'll even give you that, of the so called "media instruction extensions" across the various architectures, altivec is the cleanest, most flexible, and most useful. But the fact of the matter is, those code paths are the rare exception, not the rule. System performance is not measured by hand-tuned inner assembly loops, except in ultra-specialized cases.
Even for people that work in photoshop 90% of the time, I would guess the benefits to be marginal at best. The things that Jobs loves to show off are the filters/wide pass tools which make use of the vectorized ops, but is it anyone's job (even an ultra-specialized graphic designer) to sit around all day applying filters to very large images?
Altivec is no magic bullet. It's a specialization, very useful for a narrow scope of probles, that speeds up some kinds of vectorized ops when the programmer devotes (usually a great deal of) extra time to optimize special cases. Don't fall into the trap of believing it's some revolutionary innovation by Motorola that's misunderstood/unused by most programmers coding for the platform.
These CD's are only covering 36^16 combinations (ok so that IS large), but all you need are a few good script kiddies that combine a large database of
these, grok what values are important and start to push out tons of valid CD numbers, most which have never been printed, but certainly a good number that have. And if
they start playing with that number, which happens to be the one on the back of your legitimate CD case, you are SOL
It's possible they're doing it that way, but that would be really, really dumb. Current windows keys work like this, but the checking for validity is done on your computer, at install time.
Doing it remotely means you don't have to have some special algorithm to check keys. In fact, you wouldn't want any such algorithm. With remote storage, you just have a gargantuan keyspace and then assign valid keys at random from that keyspace. Store the valid ones in a database, and there's no way to crack the algorithm of what's a "valid" key.
Of course, the database has to be secure, but that's a different issue....
OK, I hate it when people complain about moderation to their comments, but I have to say...
THIS IS NOT FLAMEBAIT!
It's a criticism of the deceptive practices Apple is currently using in marketing.
Is Apple joining Linux as one of the "untouchable negatives" on Slashdot? Say something negative about it, and you're on your way to moderation heck, 'cause you're obviously on the wrong side.:)
A worthy reply, unlike the general "But we SAW it's faster". I'm glad to see someone is actually thinking about the stats being thrown at them, instead of accepting what you want to hear blindly.
Let's take something from a IBM's site, just for fun:
SpecINT 95 is the geometric mean 10 different benchmarks, including a GCC compilation benchmark, a GO-playing program, a LISP interpreter, and database core, some PERL code...in other words, it's a pretty wide spectrum of performance, and is really *THE* number to talk about in terms of integer performance. It's not perfect, but it's a benchmark from an independent source, and measures a wide variety of metrics. SpecFP is a similarly structured suite stressing floating point performance.
The chart is a bit old, topping out with the 750 @ 500 Mhz. The numbers given are 23.8Int/14.5Fp.
From SPEC's own site, you can see a 550Mhz Athlon scores about 25Int/20.6FP
So the cores are roughly on par in a clock-for-clock basis. Since I can't find any numbers for the new 766Mhz PPC, I'll assume it scales linearly (This is generous), and say it goes to roughly 32Int/22Fp. a 1 Ghz Athlon (which isn't even the top of the line anymore) rates 42.9/29.4.
This isn't, of course, complete system performance. It says nothing about IO speed, hard storage speed, etc. but I'll be a monkey's uncle if Photoshop is anything other than CPU or memory bound.
Apple is being deceptive. Plain and simple. They've chosen a nonrepresentative benchmark to portray the PPC as faster despite the clock disadvantage, and that's just not true.
We use several "sawtooth" (AGP-based PowerMac G4 towers) at work for video editing. Couple DVCAM decks with IEEE1394 "firewire" cards for interface, using
FinalCutPro and MediaCleanerPro. For editing work and encoding (to Soresonson 2.0 and MPEG-1) our G4s keep up with a 1.something GHz Dell that was recently
purchased. Price to performance ratio looks fine to me. Especially since the G4s seem to be more stable than our Win2K boxes for video work. For $3000 a box can be
configured at store.apple.comm that includes the combo CDRW/DVD-R drive. That's a real DVD-R! Price to performance blows away pretty much everything else.
What's your point? Want a cheap, but still fast, box to run Linux? Want to run linux or windows period? Use x86, that's what it's made for.
I'm not a zealot that thinks Apple has no place anywhere...certainly, there are still niches where Apple products make sense. It sounds like your site is one of them.
BUT, if Apple wants to play in the personal computer market, they need to stop trying to BS people into believing Macs are faster than they are, and really compete in terms of price, performance, and features. They're last big financial coups have come as a result of slick industrial design, not core computing compentencies.
OS X looks like a step in the right direction, but they've got a long ways to go before they're competitive in most arenas.
If you watched the keynote, you saw a G4 733 MHz kick the butt of a Pentium 4 1500 MHz. You
can't explain that away with technobabble... why do you even try?
If you let me pick the benchmark, I'll happily show you a 333 Mhz PII beating up on a 733Mhz PPC. Have you ever looked at the fine print in PPC benchmarking from apple? 98% of the time it's photoshop. Why? Because there is a special instruction set (Altivec) that speeds up some operations significantly.
They don't quote industry-standard SPEC numbers. STREAM? Nope. No matter how much Apple benchmarks Photoshop, it's still not a significant metric of overall system performance.
As I see it, the only reason such benchmarking works is because people WANT to believe Apple is faster
Megahertz is certainly not everything. But it's also not nothing, especially when the microarchitectural aggressiveness of the cores being compared are similar, as they are in this case.
As someone in a position with some money to save, I've been investigating "Socially Responsible" funds, which screen holdings based on various criteria, depending on the fund. There are quite a few examples around these days, from the Citizens' Funds to the Domini Social Equity Index Fund.
The one thing they have in common is a heavy weighting towards technology stocks in their holdings. I can understand why; if you're looking for companies that don't pollute and have good labor relations, the technology sector has more options than most. But in terms of portfolio diversification, this tech-heavy bias is, at least for me, more risk than I'd like to have in those investments.
Is anyone else in the same boat? I'm wondering if there are socially responsible investment options out there that *don't* have that tech bias. I certainly haven't found them...
I know little about DVD movies, especially with respect to the menu systems and other interactive portions of a disc.
My impression was that this is a simple state-machinish description language that is used to determine when to play various portions of the disc. If this is the case, then this would only be possible if that language provided some way to query the region code of the player, and I don't think this is the case.
As I understand the system, it's the player's responsibility to check the region of the disc and disable playback if the region code doesn't match.
Even if there is some way for the disc to query the player, the simple workaround is to have a player report a single region at a time; don't say you're a regionless player, make like a region 1 player until the user tells you otherwise.
At worst, you've got a situation where you have to manually reset the region code for a movie.
However I must disagree with your assertion that a physical view of the world is as dogmatic as, say, a Christian one. My core assumption upon which I base *everything*
is what I just said: Only the physical exists. I make this assumption because according to *all available evidence*, ONLY THE PHYSICAL EXISTS. I have never
encountered any evidence that anything non-physical exists. Christians, on the other hand, start from the core assumption, "God exists." However there is no evidence for
this assumption and therefore I find it to be less likely to be correct than my assumption. I am always open to evidence, of course; if you have any I'd love to hear it.
An authoritative principle, belief, or statement of ideas or opinion, especially one considered to
be absolutely true.
That statement, the one core assumption upon which you base everything:
Only the physical exists
is just as dogmatic as
God Exists
Proving either one is a descent into metaphysical philosophy. Your evidence is, at the end, a circular argument. You've started with the assumption that only the physical exists, and then used it to build a system of beliefs which then supports the original statement.
For fun, from E. A. Poe:
Take this kiss upon the brow!
And, in parting from you now,
Thus much let me avow-
You are not wrong, who deem
That my days have been a dream;
Yet if hope has flown away
In a night, or in a day,
In a vision, or in none,
Is it therefore the less gone?
All that we see or seem
Is but a dream within a dream.
I stand amid the roar
Of a surf-tormented shore,
And I hold within my hand
Grains of the golden sand-
How few! yet how they creep
Through my fingers to the deep,
While I weep- while I weep!
O God! can I not grasp
Them with a tighter clasp?
O God! can I not save
One from the pitiless wave?
Is all that we see or seem
But a dream within a dream?
My point is, you are asking me to prove something outside your system of beliefs, using only those things within your beliefs. While this is sometimes possible, it can be quite impossible.
For example, if I were to state that I believed the whole world was but some sort of process analagous to a dream to some outside, greater universe, and that nothing in our world really existed, it would be easy, trivial, really, to build a self-consistent set of beliefs around that original dogma. But in the end, that original statement is still a dogma, a starting point which cannot truly be justified or proven in and of itself.
Go on, try to prove me wrong, using arguments and constructs from my system of beliefs. That's exactly what you are asking me to do for you, (and, to throw a self referential loop, what I'm trying to do.:)
On a slight tangent, have you ever read Sophie's World? Excellent read.
You seem to think that because I don't accept the existence of the supernatural, mystical, etc. that I don't use my imagination and intuition to solve problems when the need
arises. You could not be more wrong. I love my imagination; I write quite a lot, and very little of it is based in reality.
That's not what I mean at all, if anything, it's a question of scope. You're willing to accept intuition and imagination as a tool; but only insofar as you can understand how they work in concrete terms.
The important thing to me is recognition that this, as much as fundamental christianity, or new-agism, is a dogmatic foundation. You're starting from the point that the concrete and sensory is what's 'real'. By corollary, the intuitive, or mystic, is to be fiercly distrusted. And there is nothing wrong with that as a viewpoint. But it is, still, just a viewpoint, and to me, it's no more or less valid that many others.
Another question: "terribly limiting" in what way? Am I less capable of insight and thoughtfulness because I don't accept the existence of the mystical? Am I less intelligent
and creative because of it? How am I limited? Be specific. If you can't be specific to my case (as you don't know me very well, I would imagine:) then be specific about
what kinds of ways I might be "terribly limit[ed]" by such a philosophy.
I believe that this viewpoint is limiting in its very assertion that it is the only possible correct way to look at things. I'm a much, much more concrete person than, say, my wife. She does not insist on such a rigid, consistent interpretation of her experiences and life; yet she still surprises me, almost every single day, with insights that were beyond me. And, on the other side of the coin, I help to keep her grounded in the sensory, the 'real', as you put it.
It's that dogmatic belief in the tangible that I find limiting; if you don't flex your view of what's 'real' from time to time, there are insights and epiphanies that will stay out of your reach.
Slashdot Mirror
The professor hands out the test, and the student is blown away. No possible way to pass this test, much less ace it. So the student thinks about it for a while, and eventually pulls out a $100 bill, attaches it to the test, and writes "$100 = 100 points = I pass" at the top of the paper. Having done so, the exam is turned in.
Next week, the exam is returned, and attached is a $50 bill and a note that says "$50 = 50 points = you fail".
That the compiler didn't miscompile your code?
That the system calls invoked have no bugs?
Software bugs can crop up in lots of places other than the code you just wrote...and the farther you get from your own source, the more of a pain in the butt they are to find.
This message is NOT intended to ignite a rehash of the mozilla-feasibility flamewars.
The resolution of a modern cellphone display must be on par with the old C=64 and Apple ][ stuff, why should pictures of naked people be any less popular in that medium?
While I agree with the idea that saving power is a worth goal, asynchronous design isn't, by any stretch of the imagination, the only way to get there. There are plenty of ways to reduce power usage in synchronous designs; most of these methods are only just now being used as cooling becomes a bigger issue.
The StrongARM-1 ran at 160 Mhz and consumed under a half-watt of power in the worst case. That was about 5 years ago, in a .35u fabrication process. That's around a 90% decrease in power consumption from a 486dx4-100, which was actually slower at most computing tasks, and fabricated in the same process
The SA-1 aggressively used conditional clocking, ran at 1.6-2.3V, and generally made power consumption a high-priority design goal.
Asynchronous design may be a slight win in power consumption, but only because synchronous designs don't tend to worry about it, as a rule.
Methinks perhaps you have an interesting view of supply and demand.
The huge surcharge on the latest technology is not some just some nice little premium the semiconductor manufacturer sticks on to recoup development costs. The way you talk about it, it sounds like you're saying the company adds this surcharge, and then, once they've gotten back their money for developing the high end part, they'll kindly drop the price, because they're nice folks, after all.
The price of the part reflects what the market will bear, assuming there is reasonable competition. In the x86 processor market, this wasn't the case on high end parts until fairly recently, and as a result, those high end Xeon parts and the like routinely sold at *massive* profit margins.
With reasonable competition, we fall back into the supply/demand relationship nicely. OCing only reinforces this natural market condition; by running a chip at a faster speed grade than it is rated" you lower the supply curve slightly and push down prices.
I think most OC'ing is pointless; when there's reasonable competition in the market, chips do tend to get sold at the grade at which they are reliable for normal use. There are exceptions, though, like Intel's Celeron 300A. The 300A was an attempt by Intel to artifically split the market to extract more profit from it; OC'ers actually help defeat this corporate market manipulation.
The question is one of morals. Myself, I have no particular problem with it. But many people may rightly regard overclocking as cheating, with some good reason. I am happy to admit that I am a cheater - I don't give a shit.
I'm surprised this is even comes up as an issue. When you buy a chip, you buy a piece of hardware. You've no moral obligation back to the company to use it as they intended or desire. You bought it, you can use it as you like; if they're depending on you not OC'ing so they'll get more money, you're under no obligation, legal, moral, or otherwise, to support that.
Now resellers who OC systems without full disclosure to their customers are a whole different beast.
Space for lots of standard memory
Sub $200 price tag
That's my wish list for a linux appliance. There are similar products in existance (The Cobalt Qube springs to mind) but they are overpowered and overpriced for individual use. I've also looked through a fair bit of PC/104 stuff, and found one product that would have done all of the above...for $1200.
When someone does all of the above in a good, high-quality unit, I'll jump in line to buy one.
I'm having a hard time believing this. Just seems a bit too...ambitious for a senior level class. Seniors can be smart cookies, but the 21264 (EV6/EV67/EV68) is is 4 way OOO/full register renaming implementation of the Alpha architecture. It was a several-year design cycle for a large number of very smart engineers.
Perhaps you meant they implement a limited version of the Alpha architecture in verilog?
I don't know what the technical merits are, but the marketing is solid! :)
Similarly, the sparc tree's most up-to-date stuff can be found at the repository David S Miller maintains on vger.
In addition to being the gatekeeper for the official tree, Linus is pulling double duty as the portmaster for the x86 port. Thus, the x86 stuff is always up-to-date in the official tree, and the other archs tend to have some lag time associated with them.
This is nothing new. It's just symptomatic of the hierarchical Linux development style. {Free|Net|Open}BSD don't tend to suffer from this due to their use of a central CVS repository with all portmasters having access to their relevant parts. Whether this is a better system is left for others to flamewar about, but it does prevent the port floating the author is talking about.
But that would be contrary to the general /. methodology of Comment First, Read Second if you feel like it and have nothing better to do. :)
But seriously:
He explicitely said that these benchmarks are not scientific and are merely his opinion. Therefore, this error is permissible
May be permissible to you, but I find such basic errors disturbing. It makes me wonder what he considers to be "linux tuning", and how equivalent his setups really were.
Besides, this optimization depends highly on the compiler used. My understanding is that gcc does not do as good a job optimizing as some other compilers
GCC gets this one right, regardless of architecture, unless you explicitly tell it to do no optimizations. (-O0). Check it out...compile that program with gcc and objdump -d it; you'll see no FP ops at all.
Furthermore, even if this stuff did get optimized out, I don't see how it would change the outcome of this comparison.
It probably just skewed the results more towards process creation/reaping, which is appropriate, given this is supposed to be an OS benchmark. But the fact that he added such pointless code shows he doesn't have a really good grasp of what systems and areas he's stressing (or trying to).
All in all, this is the best comparison between FreeBSD and Linux I've seen yet. Everything else I've heard up until now was basically anecdotal evidence and hearsay. Yes, this is not a real benchmark. But at least it shows some evidence. It's better than a random "FreeBSD Rulez" comment on usenet or equivalent, but not by much, in my book.
This, however, caught my eye:
long y;
y = 28.2839281;
x = 339829;
y = x / y;
Notice how I included some simple floating point arithmetic in the C program to make things just a tad tougher.
He admits he's no benchmark specialist, but any compiler worth its salt (and many that aren't) will optimize the floating point operations away. Also, since the result of the divide is never used, that will be optimized out, too.
I don't know what the real story is, and I do know a lot of knowledgeable people split on the Linux vs. FreeBSD issue. However, such a blatant error in benchmarking methodology gives me large doubts about this guy's credibility as a competent judge.
Here's to reviving debtor's prison! Let's shift more of the burder in loans to the lendee, further skewing the financial balance of powers!
Sarcasm aside, bankruptcy exists for a very good reason. Without it, freedom can be irrevocably linked to financial health. Any creditor enters into a financial agreement with trust, but also with the knowledge that *any* financial investment involves risk.
It's a choice to lend money. And sometimes you lose.
The bankruptcy system is sometimes abused (see home ownership rules in Florida, recently, and default rates on law student loans), but on the whole, I'm glad it exists...
Sv hmh csy nywx qiww yt mr csyv vitiexmrk sj xli gctliv? Jiww yt! :)
Is it just me, or is this a bad example? EVERY time I've been looking for a new car, I've had salesmen try to steer me to something more expensive/with better margins for them. It's a problem in more industries than computer sales.
I think Apple should sell on the Web and exclusively through its Mac retailers like MacZone, MacMall.
Why? Should Compaq/Dell only sell in a guaranteed monopoly environment, because someone coming in might be tempted to a machine from another vendor? Seems like if you're blaming the competition for taking away your customers, the solution isn't to take away the competition, it's to make your product demonstrably better and do a better job of telling the world why you are better
Unfortunatly, the lessons learned from recent market leaders seem to be summarized in "it's the marketing, stupid", so maybe you're right...
This is exactly the myth Apple is trying to force down your throat, and you're eagerly swallowing.
Altivec is a specialized instruction set that speeds up vector ops. To claim that a benchmark is not valid because it's not "altivec-enhanced" is roughly equivalent to saying "Oh, I didn't hand-optimize that application is assembly language for this platform, so it's not a valid benchmark"
One of the things spec tests is compiler performance, because compiler performance is an integral part of system performance. If you haven't got compilers that generate good, fast code for high level language constructs, you're dead in the water as a platform.
I'd bet there are some fp loops that altivec (or MMX on X86, MDMX on MIPS, MVI on Alpha, etc) could speed up significantly. Possibly an order of magnitude or more. I'll even give you that, of the so called "media instruction extensions" across the various architectures, altivec is the cleanest, most flexible, and most useful. But the fact of the matter is, those code paths are the rare exception, not the rule. System performance is not measured by hand-tuned inner assembly loops, except in ultra-specialized cases.
Even for people that work in photoshop 90% of the time, I would guess the benefits to be marginal at best. The things that Jobs loves to show off are the filters/wide pass tools which make use of the vectorized ops, but is it anyone's job (even an ultra-specialized graphic designer) to sit around all day applying filters to very large images?
Altivec is no magic bullet. It's a specialization, very useful for a narrow scope of probles, that speeds up some kinds of vectorized ops when the programmer devotes (usually a great deal of) extra time to optimize special cases. Don't fall into the trap of believing it's some revolutionary innovation by Motorola that's misunderstood/unused by most programmers coding for the platform.
It's possible they're doing it that way, but that would be really, really dumb. Current windows keys work like this, but the checking for validity is done on your computer, at install time.
Doing it remotely means you don't have to have some special algorithm to check keys. In fact, you wouldn't want any such algorithm. With remote storage, you just have a gargantuan keyspace and then assign valid keys at random from that keyspace. Store the valid ones in a database, and there's no way to crack the algorithm of what's a "valid" key.
Of course, the database has to be secure, but that's a different issue....
THIS IS NOT FLAMEBAIT!
It's a criticism of the deceptive practices Apple is currently using in marketing.
Is Apple joining Linux as one of the "untouchable negatives" on Slashdot? Say something negative about it, and you're on your way to moderation heck, 'cause you're obviously on the wrong side. :)
Thank you.
Let's take something from a IBM's site, just for fun:
IBM PPC Benchmarks
SpecINT 95 is the geometric mean 10 different benchmarks, including a GCC compilation benchmark, a GO-playing program, a LISP interpreter, and database core, some PERL code...in other words, it's a pretty wide spectrum of performance, and is really *THE* number to talk about in terms of integer performance. It's not perfect, but it's a benchmark from an independent source, and measures a wide variety of metrics. SpecFP is a similarly structured suite stressing floating point performance.
The chart is a bit old, topping out with the 750 @ 500 Mhz. The numbers given are 23.8Int/14.5Fp.
From SPEC's own site, you can see a 550Mhz Athlon scores about 25Int/20.6FP
So the cores are roughly on par in a clock-for-clock basis. Since I can't find any numbers for the new 766Mhz PPC, I'll assume it scales linearly (This is generous), and say it goes to roughly 32Int/22Fp. a 1 Ghz Athlon (which isn't even the top of the line anymore) rates 42.9/29.4.
This isn't, of course, complete system performance. It says nothing about IO speed, hard storage speed, etc. but I'll be a monkey's uncle if Photoshop is anything other than CPU or memory bound.
Apple is being deceptive. Plain and simple. They've chosen a nonrepresentative benchmark to portray the PPC as faster despite the clock disadvantage, and that's just not true.
What's your point? Want a cheap, but still fast, box to run Linux? Want to run linux or windows period? Use x86, that's what it's made for.
I'm not a zealot that thinks Apple has no place anywhere...certainly, there are still niches where Apple products make sense. It sounds like your site is one of them.
BUT, if Apple wants to play in the personal computer market, they need to stop trying to BS people into believing Macs are faster than they are, and really compete in terms of price, performance, and features. They're last big financial coups have come as a result of slick industrial design, not core computing compentencies.
OS X looks like a step in the right direction, but they've got a long ways to go before they're competitive in most arenas.
If you let me pick the benchmark, I'll happily show you a 333 Mhz PII beating up on a 733Mhz PPC. Have you ever looked at the fine print in PPC benchmarking from apple? 98% of the time it's photoshop. Why? Because there is a special instruction set (Altivec) that speeds up some operations significantly.
They don't quote industry-standard SPEC numbers. STREAM? Nope. No matter how much Apple benchmarks Photoshop, it's still not a significant metric of overall system performance.
As I see it, the only reason such benchmarking works is because people WANT to believe Apple is faster
Megahertz is certainly not everything. But it's also not nothing, especially when the microarchitectural aggressiveness of the cores being compared are similar, as they are in this case.
The one thing they have in common is a heavy weighting towards technology stocks in their holdings. I can understand why; if you're looking for companies that don't pollute and have good labor relations, the technology sector has more options than most. But in terms of portfolio diversification, this tech-heavy bias is, at least for me, more risk than I'd like to have in those investments.
Is anyone else in the same boat? I'm wondering if there are socially responsible investment options out there that *don't* have that tech bias. I certainly haven't found them...
My impression was that this is a simple state-machinish description language that is used to determine when to play various portions of the disc. If this is the case, then this would only be possible if that language provided some way to query the region code of the player, and I don't think this is the case.
As I understand the system, it's the player's responsibility to check the region of the disc and disable playback if the region code doesn't match.
Even if there is some way for the disc to query the player, the simple workaround is to have a player report a single region at a time; don't say you're a regionless player, make like a region 1 player until the user tells you otherwise. At worst, you've got a situation where you have to manually reset the region code for a movie.
Unless I'm missing something big here?
You've made my point for me, here. From www.dictionary.com:
That statement, the one core assumption upon which you base everything:
Only the physical exists
is just as dogmatic as
God Exists
Proving either one is a descent into metaphysical philosophy. Your evidence is, at the end, a circular argument. You've started with the assumption that only the physical exists, and then used it to build a system of beliefs which then supports the original statement.
For fun, from E. A. Poe:
And, in parting from you now,
Thus much let me avow-
You are not wrong, who deem
That my days have been a dream;
Yet if hope has flown away
In a night, or in a day,
In a vision, or in none,
Is it therefore the less gone?
All that we see or seem
Is but a dream within a dream.
I stand amid the roar
Of a surf-tormented shore,
And I hold within my hand
Grains of the golden sand-
How few! yet how they creep
Through my fingers to the deep,
While I weep- while I weep!
O God! can I not grasp
Them with a tighter clasp?
O God! can I not save
One from the pitiless wave?
Is all that we see or seem
But a dream within a dream?
My point is, you are asking me to prove something outside your system of beliefs, using only those things within your beliefs. While this is sometimes possible, it can be quite impossible.
For example, if I were to state that I believed the whole world was but some sort of process analagous to a dream to some outside, greater universe, and that nothing in our world really existed, it would be easy, trivial, really, to build a self-consistent set of beliefs around that original dogma. But in the end, that original statement is still a dogma, a starting point which cannot truly be justified or proven in and of itself.
Go on, try to prove me wrong, using arguments and constructs from my system of beliefs. That's exactly what you are asking me to do for you, (and, to throw a self referential loop, what I'm trying to do. :)
On a slight tangent, have you ever read Sophie's World? Excellent read.
That's not what I mean at all, if anything, it's a question of scope. You're willing to accept intuition and imagination as a tool; but only insofar as you can understand how they work in concrete terms.
The important thing to me is recognition that this, as much as fundamental christianity, or new-agism, is a dogmatic foundation. You're starting from the point that the concrete and sensory is what's 'real'. By corollary, the intuitive, or mystic, is to be fiercly distrusted. And there is nothing wrong with that as a viewpoint. But it is, still, just a viewpoint, and to me, it's no more or less valid that many others.
Another question: "terribly limiting" in what way? Am I less capable of insight and thoughtfulness because I don't accept the existence of the mystical? Am I less intelligent and creative because of it? How am I limited? Be specific. If you can't be specific to my case (as you don't know me very well, I would imagine :) then be specific about
what kinds of ways I might be "terribly limit[ed]" by such a philosophy.
I believe that this viewpoint is limiting in its very assertion that it is the only possible correct way to look at things. I'm a much, much more concrete person than, say, my wife. She does not insist on such a rigid, consistent interpretation of her experiences and life; yet she still surprises me, almost every single day, with insights that were beyond me. And, on the other side of the coin, I help to keep her grounded in the sensory, the 'real', as you put it.
It's that dogmatic belief in the tangible that I find limiting; if you don't flex your view of what's 'real' from time to time, there are insights and epiphanies that will stay out of your reach.