From the article: and an all new system architecture that delivers up to five times the performance of the PowerBook(R) G4 I wish Apple would stop doing this - this creates a misleading impression about the macbooks, it would be better phrased as: Up to 5 times faster for many operations, but substatially slower for legacy software & software that relies on altivec.
Actually you're wrong. We've compared 250MB Powerpoint presentations on a Core Duo 1.8 iMac vs. a 1.5 GHz Powerbook G4. The presentation author, who produces presentations of this size all the time and who is quite tuned to the performance of his notebook told us that Office 2004 running in emulation on the core Duo was faster than his powerbook. His powerbook had more memory than our iMac also. Granted the subsytems of the iMac are also much faster but Powerpoint loads the presentation into memory (in fact, the iMac almost ran out, PowerPoint was using nearly 700MB running a file that large in emulation).
The point is, YES, the CoreDuo is several times faster than the decrepit G4s. Even in applications that are not pervasively multi-threaded many PPC applications (All I've seen) run as fast or faster in emulation on a Core Duo then they do on a modern G4 machine. The worst case is, you old applications either won't run (no classic, and a subset of Carbon apps don't run).. or it will run as fast or slightly faster in emulation that it did natively on your old notebook. Boo-friggin-hoo. How terrible, you app may run just as well in emulation and it will run 5x faster when you get your next 3rd party software upgrade.. not to mention that it's dual core and faster so it will always be more responsive in general.
The one thing Apple's failing at is not stressing enough that emulated apps require about twice the memory as native apps did on your old machine.
P.S. I should also mention that Rosetta dynamically recompiles binaries and caches that recompile so they get faster as you use your PPC apps. First launch is much slower than 2nd or 3rd launch for example.
No, it's been held back by a lack of Windows, which prevents people from running applications.
I think you meant to say that the lack of Windows prevents people from running CERTAIN applications since I've never been want for any particular class of application on the Mac. There are, however, many niche industries (CAD for example) where Windows software options provide a massive platform advantage to PCs. Saying people can't run applications.. or even implying that people can't run useful applications would have been a moronic statement on your part. In fact, I find that the Mac versions or alternatives to most common applications (video editing, photo management... ) are much nicer on the Mac.
It's not called fear, although a characterization like that is typical of an Apple fanboy.
Here, I'll fully disagree. People ARE affraid of the unknown. It's what makes so many of us sniveling little monkies. I've seen WAY too many people who have never ever used a Mac bash the Macintosh platform and Apple with lame and false 'evidence' to come to any conclusion other than they are affraid to leave their PC comefort zone.
See, people buy computers (whether they be PCs or game consoles or handhelds) because those platforms have the software they want. That's #1. #2 is style, and that can be enough to sell computers when people don't know what software they want to run - hence the people who buy Macs because "they're easier to use" or other bullshit excuses.
Again.. you're kind of full of crap. I agree that many people buy PCs because of specific software. Gaming is a huge example. Some niche industries (CAD comes to mind, the Law profession is very Windows and so are their software tools...). Yet another would be large corporations with custom Windows code that is required in the normal functioning. That's not why average everyday people buy PCs though. They buy them because they are cheaper, because it's what they know (they've never seen a Mac) and because that's what they are told to buy when they talk to the 'computer person' in the family. Honestly, if you take someone without Mac-PC biases.. somone who isn't a techie.. they want the computer that they can use. They don't give two craps if it runs windows or if it runs OS X. My Mom is getting into her computer more. I was going to get her a Mini and when I showed her one she REALLY wanted it. She had her first PC at the time. I chose, for purely financial reasons, to augment my spare parts and build her a cheap P4 2.66GHz box. She's having a terrible time with it. I should have just spent more money and got her a mini. I have to show her the same things over and over. I update her spyware every time I'm there. Yesterday Virex found two portscanners only after the files were touched by Ad-Aware (and yes, Virex was updating every night even when she picked up the portscanners). Not to mention her biege box is 10x the size, uses 5x the power and it's easily 3x as loud. Bottom line is, she wanted a Mac and she'd have been much happier with a Mac. She's a very average user. It has nothing to do with software because there are mac versions or counterparts for everything she uses (internet apps/web browsers, email, Skype, photo management (crappy software from her Camera, Office suite, iTunes.. )
Apple's biggest problem remains price. It's difficult to justify that extra few hundred dollars even though it costs you endless hours of hassle down the line. Time always seems cheaper than real paper dollars. For most people it's not software. Average people actually do want pretty things that just work.. they just tend to look at the bottom line first. Apple needs to get the price down more and hammer in the ease of use message and we'll see Macs with 10% of the market sooner than you can call them 'Beleagured'.
I heard some retard said Jordan Hubbard left FreeBSD development. I wonder where Jordan went.. bet he's writing Windows software now..
Oh wait, Jordan Hubbard is leading work at Apple on that OS X thingy they have.. you know, the FreeBSD based OS with that Darwin open source core thingy.
Poor beleagured Apple, they'll be out of business any day now. Oops. Scratch that. Poor beleagured FreeBSD, they'll be out of buisness any day now. Oops. Scratch that. Poor beleagured OpenBSD. They'll..... got to blow my nose, excuse me.
This is interesting and all, but this isn't exactly a ground-breaking news item. PGP lets you do this on various platforms. As a matter of fact, this is how I manage personal info on my OS X Macintosh. I create an strong-encrypted virtual disk image with banking, internet login, software key, and (un)related information. When I need something I mount it and when I'm done I umount it and it's nice and safe (as long as I never tell Keychain to remember the password). You can do this on a vanilla OS X install with Disk Utility.
No, it's not Napster complaining that they can't load napster music onto iPods. Napster music is even more restrictive than Apple's Fairplay. It is only active as long as your napster subscription is valid because you don't own Napster songs. If Napster could load their music onto iPods, you could fill a 60GB iPod and cancel your napster service. That would violate your Naptster license unless the songs just exploded at some point and ceased to work. The reality is that Napster has painted themselves into an even worse corner with their DRM.
If anything, Napster is probably pissed that they can't load iTunes songs onto their devices.. the devices they have to sell because they are intertwined with their music licensing scheme.
More specifically, OS X is mostly 32bit. 64bit libraries are available. You can run native 64bit integer math with the accelerate framework so you can do your fast, high-precision work on a G5. The big problem is, the GUI parts of the OS (most notably) are still 32bit. GUI apps must be 32 bit. Apps like Mathematica run kind of like X-Window.. they have a GUI and a mathematical engine running in the background. It's kind of client/server. Wolfram has a 64bit engine, but not a 64bit GUI but you don't need the GUI to be 64bit native.
The problem is, other apps aren't logically de-coupled like this so it's difficult write these 32bit/64bit applications. The big issue, as I understand it, is that there needs to be a distinct separation of 32bit native and 64bit native code.. not just in spawned threads but in actual binaries that are compiled. In Mathematica, the front end is a separately compiled binary from the computation engine.
It was widely reported at the Mactel announcment that Apple currently pays less to IBM for G4 and G5 chips than it will cost them to buy Intel chips.
I think "widely reported" is a mischaracterization. First off, I'm a technical IT manager.. that is I'm an IT Manager who doesn't expect his staff to do anything I can't or won't do myself. I've NOT heard this characterisation and I follow the trends, especially this one, closely. I am, after all, a mac-god.;-P (seriously though, it is my primary specialty.. Mac/OS X related support)
Second, Apple isn't just buying chips, they are buying a full Intel package. Apple's getting the same deal that Dell will be getting. They will get volume discounts for FULL systems. Apple's repackaging Intel chipset/processor systems and I strongly suspect we'll see Apple systems which are reference systems. Apple's R&D for new Intel systems will be nearly nothing compared to what they currently spend on PowerPC systems. Apple designs the system controller themselves after all. Not only will they get bulk system pricing, but they'll qualify for the same exclusivity pricing that Dell leverages. Could there be any other reason why Apple didn't simultaneously announce future server support for Opteron? Aside from just being a better architecture, Opteron's support for large SMP configurations is at least a generation ahead of Xeon and XeonMP.
I'm not sure I follow your argument on capacity either. I don't disagree that Apple's G5 production is 5% of plant capacity for IBM. This doesn't mean that IBM can easily ramp up production for PPC970s though. The 5% staticstic doesn't address the overall capacity of the plant at all. Are they running at 50%? 80%? 100%.. are they pulling overtime and trying to push 105% capacity? Additionally, running a product line in a fab that takes up 5% of that fab's capacity will NOT make it cheaper. There are certainly the penalties of lack of economies of scale at play in this case.
My basic points.. I don't believe for one second that Apple will pay more for Intel chips than PowerPC chips. That MAY be the case with outdated G4 processors but I simply can't believe that's the case for PPC 970s. Even if I were completely wrong on this, you have to acknowledge that Apple will end up paying significantly less to build Intel systems v. G5 systems since they will have SIGNIFICANTLY less R&D with Intel's complete solutions.
You know Apple hardware is way less expensive because???
I'm sure it's because you've combed their financials and you've figured out what per unit profit is after removing cost/profit associated with R&D, retail, distribution, software sales....
Or did you simply decide this because you did the most obvious thing, you compared them to Dell? You figured a G5 is pretty much the same thing as a P4 even though Apple has to buy a relatively low volume processor from a different company, and they have to design and contractract the fab of their own system controller and motherboards, and they have smaller economies of scale, and they make a nicer box (there's about 10lb of Aluminum in just the G5 tower shell)... That's how you know that Apple charges way too much, right?
Of course every kid knows this.. that's why the average ACT score is like 13.
I have directly compared prices recently. A Quad core G5 tower (dual-dual) with the same drives and same video as a quad Core Opteron (dual-dual) is actually about a grand cheaper . Yes, the Dual Processor Dual Core Opteron from Boxx (a very nice computer btw) was significantly cheaper.
I'll send you a 6pack of good beer if you can find me a quad core Opteron from a (as in one) reputable company (that won't go out of business).. with a warranty and one number to call in case of problems.. and a supported OS installed for anything near a new Quad Core G5 price.
And now you're "who"ing a fucking film!??! You fucking moron, a film is a fucking object. It is a THAT. Who is for people! Though you probably "that" people and "who" objects. Get off my fucking planet you retarded fuck. English is the only fucking language I know, and it's fucked up enough without you assholes destroying what little sensible structure it fucking has left.
Hey jerkwad...
"that" and "who" are not verbs.
For someone who is so annoyingly critical of web-post english grammer you 'sho don speek engrish no good'.
Here's a free tip. Maybe if you weren't such an annoying asshole, you might get a date.
Well, this isn't my area of expertise, but I do have the advantage of being somewhat intelligent and rational.
As I stated, I can not cite a reputable scientist in the field as I don't know any though I'm sure you could find plenty of relevent info if you google. I can, however, see obvious trends in cells and related biological formations TODAY that seem to me as relevant examples of a feasible evolutionary map in general cellular design.
The argument you make is, I presume, based not only the complexity of a complex animal cell but also on those parts of the cell that are also claimed to be irreducably complex like the mitochondria. Let's look at a generic animal cell first. Cell wall, mitochondria, nucleus.. lots of complex stuff in there and it's all wrapped up in a package that just works. Intelligent no? What about a red blood cell. It's not actually a cell, it's really called a leukocyte as I recall because though it is a self-contained 'cell' it lacks a proper nucleus. In the human body, there are various 'cells' of varying function and complexity.. some more complex than other. Now pull back to all 'life'. There are other examples of much more remedial cells. Anaerobic organisms often don't posses typical mitochondria since mitochondria are required for aerobic energy metabolism. I'll have to once again admit my lack of expertise here so I'll skip other intermediaries. We can, however, just jump to the end example. What about viruses? They are self-contained 'living' units much like cells. Some of them are quite structurally complex with dna injecting mechanisms. They are, however, so simple compared to the typical animal cell that some biologists hesitate to call them alive.
The point I'm trying to make is that, although individual cells are too fragile to be found intact in the fossil record, there are presently a wide variety of cells and sub-cell organisms/organic packages that span a great range of complexity. Personally, though I don't have a full record of every cellular missing link, I don't find it shocking to believe that simple organisms evolved into more complex organism even if those organisms began as little more than replicating and self-replicating protiens (aka. Viruses)
There is a difference between this line of reasoning and faith however. This process, cellular evolution, is currently being studied and it has been for some time. We know WAY more about how cells have or may have evolved into complex structures now than we ever have before. In by know I mean we can build models on our theories and prove them by direct inspection or experimentation. Some models will undoubtedly fail. This is SCIENCE. You throw things at the wall and keep what sticks. Faith is COMPLETELY DIFFERENT though. You don't throw hypothesis and theory at the wall to see what sticks with Faith. Faith REQUIRES you to belive you know what will stick to the wall without doing the throwing. By definition, if you test the targets of your faith and prove those positions, it is no longer faith it's hard fact. Faith is no longer required at that point. My dictionary tells me that Faith is based on spiritual apprehension RATHER THAN PROOF. That's just the way Faith works.
Science isn't faith. There is hypothesis and theory but there are really very few laws in science. Science is based on the ability to test and disprove any aspect of science. This is just the way science works. If you can't potentially disprove it, it isn't science it's faith. This is especially true of hypothesis and theory. Attempts to disprove hypothesis and theory are every bit as important or even more so than attempts to 'prove' hypothesis and theory. It is, after all, much easier to find one case where a theory fails than to find every possible case (possibly an infinite number) where a theory succeeds. Modern theorys, as a result, become more accurate as exceptions are found an theories evolve to account for those exceptions.
This one's totally off topic but I'm still in shock. We were driving in Chicago today and I've added a new one to the "now I've seen it all" list.
The guy in front of us remained stopped at a green light, it was clear that he was trying to talk out the passenger window of his car to a woman in another car. She was trying to exit a parking lot of a strip mall, he was blocking her by stopping his car in the middle of a traffic lane right in front of the parking lot exit. This was on a very busy street in Chicago.
My wife was driving and she hit the horn when it was clear he wasn't planning on actually driving through the green light any time soon. He responed by putting the car in park in the lane and getting out to talk to the woman. He looked into our car and I gave him the hands up and I clearly mouthed WTF.. he didn't seem to notice and we went to talk to the girl. My wife pulled into the other lane to get around him before it escellated any more.
THAT is the quality of drivers in the United States.
No dumbass, But I'd have appreciated it if the GERMAN PEOPLE had stood up to him BEFORE he launched a war that resulted in the persecution of tens of millions, nearly 100 million people (if you look at the massive Russian losses and the exterminated Jews)
Your ignorance is profoundly frightening. I can't figure out how you equated "Give me Liberty or give me death" to capitulation to Hitlers attrocities. Perhaps you can enlighten us.
Birge, You seem to know a lot of.. well, ficticious information about the Mac OS. It's really fascinating.
Where exactly is Apple still using Resource forks? I haven't run into a Resource fork issue in Years and I run a Mac and Windows support office.
As for Applescript.. I do hate the loose syntax but you probably don't realize that Applescript is actually a real object oriented language. It does have one important advantage over other scripting languages too, intra-application message passing via. AppleEvents. The funny thing about your Apple script comments though are the fact that you're comparing Applescript (which is what, 2 decades old? nearly there? to the shell of a beta OS that's due, maybe, in a year. I think that counts as a win when Applescript (the baby of Apple Employee #5) deserves any comparision to something the Mighty MS hasn't even finished yet.
For Spotlight.. well, it's not a DB filesystem like WinFS.. but since Vista won't ship with WinFS and it's officially vaporware, what's the point? Spotlight doesn't wake up before me and scramble my eggs and put the coffee on either but it's still a sweet tool. It's got a well documented API. It's got command line tools so I can script it. It's got an Apache Module on OS X Server so I can use it for web searches. It can create search databases on volumes other than my boot drive. It's extensible via plugins. And.. it's fast. It does everything I need a search engine to do. My only complaint is that there should be an easy way to do more complex searches based on various unrelated criteria.. but If I weren't so lazy I'd write that interface.. because I can with the Spotlight framework. Now back to it not actually being the filesystem. Well, it is tied into the filesystem at a low level (file opens and writes trigger indexing) but more importantly.. It's not the filesystem. WooHoo! Databases make for great archives of information that are easy to access quickly.. but they can be a real bitch when they get corrupted. Thank god my filesystem is just some lame old tree based file system. They are soo much easier to repair than corrupted Databases.
I actually thought that the story would be about something like my desk setup. I'm in a 1950's research building.. 15" concrete floors, concrete block walls and imbedded shelf rails in the walls. The obvious solution for me was to take a long shelf rail, have the guy in the ship saw a sliver out of it, bend it, and weld it back together so it angles down as it leaves the wall. I finished a small plank of 1/4" plywood (fancy stuff). I ran it through the router table to put a nice edge on it and clear sealed it.
My Powerbook actually floats over the desk, straight from the wall. It's about Starbucks tall off the desk in the front so I can easily slide my keyboard and mouse under when another notebook comes in for repair. It's too high by design for normal typing but the angle makes it easy to use in a pinch. It also allows me to use my mouse/keyboard when working on a desktop on my desk.. but I can still use the laptop when needed.
To be honest, I don't think I'm that clever that I should have a better solution than something in a slashdot article. Pretty sad.;-P
I don't look it, but I've got an enormous head. I can't wear anything but an xL hat. I know this is utter crap because I know many quite intelligent people, people I consider as intelligent as I, and the running joke is about my big ass head. I've gone one buddy.. if I toss my hat on him, he looks like he's got progeria because he's got a pea-head and he's skinny. Think bony adult looking child with grown-ups hat on.
This is like someone saying the sky is green. You can present your research, but I've actually looked up a few times.
I wish I could edit posts.:-) I went and reread my initial post and I can easily see that it appears to let NSIT off the hook too easily. I really intended to say 'don't crucify them yet till you know the facts' but it did come out as 'this isn't their fault, it's the outside web developers'. I didn't intend that, I only ment to demonstrate that it's more complex an issue that nsit must have done it.
Thanks for the additional info. I'll admit, I've still not read the other links like that Maroon article.
In the post you replied to, I was specifically trying to quell the absoluteism and idiocity that is too common on slashdot (by a minority I like to believe). I primarily wanted to point out that lower pay than the industry doesn't mean there aren't talented people. There are a lot of talented people in NSIT, on campus, and at most if not all Universities. I know plenty of people who are more than smart enough and talented enough to do the grind in corporate. I know some who have and immediately returned.
I'm not excusing NSIT or WHO EVER was directly or passively involved. It's a major F up. I can't imagine why a file with SS#s were up there in the first place (though I've done this work long enough to know that just because I can't see the purpose outside the black box, that doesn't mean it isn't there).
This is true, and it's absolutely wrong at the same time. How cool is that.
I'll make the disclaimer now, I'm a UC employee and I work in IT. I'm not affiliated with NSIT, the group under who's watch this problem occurred.
First off. There are plenty of very smart people working at UC. The quality and the size of the central IT staff is superior, imho, to that of my previous employeer.. a State University that was actually larger (plenty of friends on staff at that State university and they are good smart people too.. I'm making broad comparisons here).
As for pay, Yes, Education pays less than the 'real world'. It often pays MUCH less. UChicago doesn't pay salaries comparible to the corporate world but it's far from the worst EDU in the area in this regard. EDU staff doesn't tend to come or stay for the pay, however. Working in EDU is often a geeks dream. Flexible hours.. I know plenty who work 2pm till they are done because they prefer that. Getting good work accomplished is more important than being at your desk at 9am. Great benefits... 4 weeks vacation/personal holiday to start at UC.. more at the last Univeristy I worked for staff with a degree. Overall, a fun, relaxed, and often challanging work place where you help interesting people day in and day out. Bottom line, there are plenty of very very intelligent, very very talented people doing IT at Universities and they are retained because they'd rather be happy with a smaller pay check than misserable and rolling in dough. (I wish it could be both as much as anyone).
All that said, yes, there are plenty of dumb asses doing IT at Universities. One pattern I see is that smaller groups like Departments and Institutes will hire Techs for internal use but they won't seek out the right people on campus to properly evaluate the applicants. It's awfully easy to BS your way through an interview when the interviewers are the ones desperate for tech support because they are clueless about computers. Me, I interviewed with a one of those people.. and an associate Director of NSIT. That and an internal reference are the reasons my position isn't currently filled by a real dumbass (just my minor and occasional dumbass-id-ness).
OK, nuff of my rant. Read the story again. It appears that NSIT may not have even been responsible for creating this problem. The University of Chicago has a professional web developer on retainer. From the story, it looks like a web developer put a sensitive file in the wrong location with the wrong permissions. I'm not sure if this was even caused by an NSIT staff member.
BTW: The current security group on campus is actually quite good. I'm very happy with them, they are really on top of their stuff as far as network monitoring, reporting, and control. Machine get's violated and bam, the port is off with a quickness. The report shows up in our email immediately and we can go off to see who brought an unsecured box into our division with out even bothering to ask us about whether it's buttoned down.:-)
It's my understanding that HT is an attempt to use sections of the CPU that are idle at any given moment. HT isn't a solution that was designed to address problems related to the long pipe like blown branch predictions.
No matter how you look at it though, the G5 is a good candidate for symetric multi-threading (aka. intel-style Hyperthreading). It has a lot of execution units... there plenty of sections of the cpu that are idle in any given clock cycle. The pipeline of the G5 (PPC 970) is also as long as the pipeline in a Northwood P4. To be specific.. different sections of the PPC 970 have different length pipes.. the interger units have shorter pipes than the altivec units.. but they average nearly the same lengh as the Northwood.. up to 23 stages I believe.
So, the 970 is good for HT based on how you believe HT to work, and it's good for how I believe HT to work.
Oh, and IBM has released SMT enabled cpus too.. the Power5 has symetric multi-threading and its a lot more efficient than Intel's HT. So.. there is no particular reason why a wide power[pc] processor can't have SMT. It's just not a feature in in IBM's commodity chips yet.
The suit is a red herring of some sort. Apple can't possibly hope to prevail because the suit claims damage, or at least potential damage if these rumors allow a competetor to duplicate their efforts and come to market first. Think about this for a second. If TS posts two weeks out of Mac World that Apple is releasing a product.. something that took months and months to design and months to ramp up production.. how is a competetor going to use this information to duplicate their product? This is likely about the imac. There are already cheap headless computers out there and it takes more than weeks to design and market a new system. It could also be about the rumored iWorks software suite (though not sure if TS commented on this) or leaked Panther stuff, or the rumored firewire direct in box. Do you suspect that releasing this info a couple weeks early would allow a competetor to beat them to market?
First off.. I don't believe you (about the 70's chip being faster than today's fastest Pentium[IV]s)
Second.. what you are describing is NOT general purpose CPU but a very very simple DSP or ASIC. DSPs are designed for specific purposes and they run those calculations much faster than a general purpose cpu.. but the problem is, they aren't well suited to doing things like Compression, 3D, and running your word processor at the same time. You want to simplify the instruction set so much that it isn't useful as a computer processor?
third.. REAL RISC, like Alphas, have been proven to be very very fast, but wait till you need to do a calculation that you don't have an instruction for.. then you're building an algorithm with various instructions and that is NOT faster. Back when Alpha's were destroying x86 and PPC boxes in a lot of benchmarks, the Pentium and the PPC even more so would absolutely obliterate it in RC5 bruteforce cracking. Why? Because the Alpha didn't have a required instruction in hardware so they had to work around that in code.
fourth.. I've absolutely no idea what you are talking about with 8GHz performance? What's super cooling? Liquid Nitrogen? Strained Silicon won't get anything like a modern processor to run at 1 volt on.13 micron unless you clock it really low. And why is a 4 stage pipe going to magically make this chip as fast as an 8 GHz PentiumX? Because it has a lower latency of ticks? If the latency is 4 ticks but it runs at 200MHz.. and your P4 has a latency of 25 ticks and it runs at 3000 MHz.. the P4 has a lower latency BY TIME! That's what really matters isn't it?
finally.. I wish you luck with your 8000+ (pr rating of course) box that can only add and subtract. I'm sure it will be a fabulous computing experience.
Maybe the poster should have actually read the wiki article they posted to... Algorithms are named after Abu Abdullah Muhammad bin Musa al-Khwarizmi. He's the father of the algorithm.
I think the confusing here is that the author thinks that algorithm is a concept that only applies to computers.
It was my impression that the original G5 cluster was up and performing real work for several months before it was tore down for a rebuild. It's probably safe to assume that the work in setting up the first cluster decreased the deployment of the second cluster so it wasn't just an exercise is vanity even if it didn't perform a lot of production calculations.
Slashdot Mirror
From the article:
and an all new system architecture that delivers up to five
times the performance of the PowerBook(R) G4
I wish Apple would stop doing this - this creates a misleading impression about the macbooks, it would be better phrased as:
Up to 5 times faster for many operations, but substatially slower for legacy software & software that relies on altivec.
Actually you're wrong.
We've compared 250MB Powerpoint presentations on a Core Duo 1.8 iMac vs. a 1.5 GHz Powerbook G4. The presentation author, who produces presentations of this size all the time and who is quite tuned to the performance of his notebook told us that Office 2004 running in emulation on the core Duo was faster than his powerbook. His powerbook had more memory than our iMac also. Granted the subsytems of the iMac are also much faster but Powerpoint loads the presentation into memory (in fact, the iMac almost ran out, PowerPoint was using nearly 700MB running a file that large in emulation).
The point is, YES, the CoreDuo is several times faster than the decrepit G4s. Even in applications that are not pervasively multi-threaded many PPC applications (All I've seen) run as fast or faster in emulation on a Core Duo then they do on a modern G4 machine. The worst case is, you old applications either won't run (no classic, and a subset of Carbon apps don't run).. or it will run as fast or slightly faster in emulation that it did natively on your old notebook.
Boo-friggin-hoo. How terrible, you app may run just as well in emulation and it will run 5x faster when you get your next 3rd party software upgrade.. not to mention that it's dual core and faster so it will always be more responsive in general.
The one thing Apple's failing at is not stressing enough that emulated apps require about twice the memory as native apps did on your old machine.
P.S. I should also mention that Rosetta dynamically recompiles binaries and caches that recompile so they get faster as you use your PPC apps. First launch is much slower than 2nd or 3rd launch for example.
ffakr.
I think you meant to say that the lack of Windows prevents people from running CERTAIN applications since I've never been want for any particular class of application on the Mac. There are, however, many niche industries (CAD for example) where Windows software options provide a massive platform advantage to PCs.
Saying people can't run applications.. or even implying that people can't run useful applications would have been a moronic statement on your part. In fact, I find that the Mac versions or alternatives to most common applications (video editing, photo management... ) are much nicer on the Mac.
Here, I'll fully disagree. People ARE affraid of the unknown. It's what makes so many of us sniveling little monkies. I've seen WAY too many people who have never ever used a Mac bash the Macintosh platform and Apple with lame and false 'evidence' to come to any conclusion other than they are affraid to leave their PC comefort zone.
Again.. you're kind of full of crap. I agree that many people buy PCs because of specific software. Gaming is a huge example. Some niche industries (CAD comes to mind, the Law profession is very Windows and so are their software tools...). Yet another would be large corporations with custom Windows code that is required in the normal functioning.
That's not why average everyday people buy PCs though. They buy them because they are cheaper, because it's what they know (they've never seen a Mac) and because that's what they are told to buy when they talk to the 'computer person' in the family. Honestly, if you take someone without Mac-PC biases.. somone who isn't a techie.. they want the computer that they can use. They don't give two craps if it runs windows or if it runs OS X.
My Mom is getting into her computer more. I was going to get her a Mini and when I showed her one she REALLY wanted it. She had her first PC at the time. I chose, for purely financial reasons, to augment my spare parts and build her a cheap P4 2.66GHz box. She's having a terrible time with it.
I should have just spent more money and got her a mini. I have to show her the same things over and over. I update her spyware every time I'm there. Yesterday Virex found two portscanners only after the files were touched by Ad-Aware (and yes, Virex was updating every night even when she picked up the portscanners). Not to mention her biege box is 10x the size, uses 5x the power and it's easily 3x as loud.
Bottom line is, she wanted a Mac and she'd have been much happier with a Mac. She's a very average user. It has nothing to do with software because there are mac versions or counterparts for everything she uses (internet apps/web browsers, email, Skype, photo management (crappy software from her Camera, Office suite, iTunes.. )
Apple's biggest problem remains price. It's difficult to justify that extra few hundred dollars even though it costs you endless hours of hassle down the line. Time always seems cheaper than real paper dollars.
For most people it's not software. Average people actually do want pretty things that just work.. they just tend to look at the bottom line first. Apple needs to get the price down more and hammer in the ease of use message and we'll see Macs with 10% of the market sooner than you can call them 'Beleagured'.
ffakr.
I heard some retard said Jordan Hubbard left FreeBSD development.
I wonder where Jordan went.. bet he's writing Windows software now..
Oh wait, Jordan Hubbard is leading work at Apple on that OS X thingy they have.. you know, the FreeBSD based OS with that Darwin open source core thingy.
Poor beleagured Apple, they'll be out of business any day now.
Oops. Scratch that.
Poor beleagured FreeBSD, they'll be out of buisness any day now.
Oops. Scratch that.
Poor beleagured OpenBSD. They'll..... got to blow my nose, excuse me.
This is interesting and all, but this isn't exactly a ground-breaking news item.
PGP lets you do this on various platforms.
As a matter of fact, this is how I manage personal info on my OS X Macintosh. I create an strong-encrypted virtual disk image with banking, internet login, software key, and (un)related information. When I need something I mount it and when I'm done I umount it and it's nice and safe (as long as I never tell Keychain to remember the password).
You can do this on a vanilla OS X install with Disk Utility.
ffakr
No, it's not Napster complaining that they can't load napster music onto iPods. Napster music is even more restrictive than Apple's Fairplay. It is only active as long as your napster subscription is valid because you don't own Napster songs. If Napster could load their music onto iPods, you could fill a 60GB iPod and cancel your napster service. That would violate your Naptster license unless the songs just exploded at some point and ceased to work.
The reality is that Napster has painted themselves into an even worse corner with their DRM.
If anything, Napster is probably pissed that they can't load iTunes songs onto their devices.. the devices they have to sell because they are intertwined with their music licensing scheme.
More specifically,
OS X is mostly 32bit. 64bit libraries are available. You can run native 64bit integer math with the accelerate framework so you can do your fast, high-precision work on a G5.
The big problem is, the GUI parts of the OS (most notably) are still 32bit. GUI apps must be 32 bit. Apps like Mathematica run kind of like X-Window.. they have a GUI and a mathematical engine running in the background. It's kind of client/server. Wolfram has a 64bit engine, but not a 64bit GUI but you don't need the GUI to be 64bit native.
The problem is, other apps aren't logically de-coupled like this so it's difficult write these 32bit/64bit applications. The big issue, as I understand it, is that there needs to be a distinct separation of 32bit native and 64bit native code.. not just in spawned threads but in actual binaries that are compiled. In Mathematica, the front end is a separately compiled binary from the computation engine.
ffakr.
I think "widely reported" is a mischaracterization. First off, I'm a technical IT manager.. that is I'm an IT Manager who doesn't expect his staff to do anything I can't or won't do myself. I've NOT heard this characterisation and I follow the trends, especially this one, closely. I am, after all, a mac-god.
Second, Apple isn't just buying chips, they are buying a full Intel package. Apple's getting the same deal that Dell will be getting. They will get volume discounts for FULL systems. Apple's repackaging Intel chipset/processor systems and I strongly suspect we'll see Apple systems which are reference systems. Apple's R&D for new Intel systems will be nearly nothing compared to what they currently spend on PowerPC systems. Apple designs the system controller themselves after all.
Not only will they get bulk system pricing, but they'll qualify for the same exclusivity pricing that Dell leverages. Could there be any other reason why Apple didn't simultaneously announce future server support for Opteron? Aside from just being a better architecture, Opteron's support for large SMP configurations is at least a generation ahead of Xeon and XeonMP.
I'm not sure I follow your argument on capacity either.
I don't disagree that Apple's G5 production is 5% of plant capacity for IBM.
This doesn't mean that IBM can easily ramp up production for PPC970s though. The 5% staticstic doesn't address the overall capacity of the plant at all. Are they running at 50%? 80%? 100%.. are they pulling overtime and trying to push 105% capacity?
Additionally, running a product line in a fab that takes up 5% of that fab's capacity will NOT make it cheaper. There are certainly the penalties of lack of economies of scale at play in this case.
My basic points.. I don't believe for one second that Apple will pay more for Intel chips than PowerPC chips. That MAY be the case with outdated G4 processors but I simply can't believe that's the case for PPC 970s. Even if I were completely wrong on this, you have to acknowledge that Apple will end up paying significantly less to build Intel systems v. G5 systems since they will have SIGNIFICANTLY less R&D with Intel's complete solutions.
ffakr.
You know Apple hardware is way less expensive because???
I'm sure it's because you've combed their financials and you've figured out what per unit profit is after removing cost/profit associated with R&D, retail, distribution, software sales....
Or did you simply decide this because you did the most obvious thing, you compared them to Dell? You figured a G5 is pretty much the same thing as a P4 even though Apple has to buy a relatively low volume processor from a different company, and they have to design and contractract the fab of their own system controller and motherboards, and they have smaller economies of scale, and they make a nicer box (there's about 10lb of Aluminum in just the G5 tower shell)...
That's how you know that Apple charges way too much, right?
Of course every kid knows this.. that's why the average ACT score is like 13.
Stupid Ffakr.
Second part of that was supposed to be..
The Quad Core Opteron was significantly MORE EXPENSIVE (THAN THE MAC)
I have directly compared prices recently. A Quad core G5 tower (dual-dual) with the same drives and same video as a quad Core Opteron (dual-dual) is actually about a grand cheaper .
.. with a warranty and one number to call in case of problems.. and a supported OS installed for anything near a new Quad Core G5 price.
Yes, the Dual Processor Dual Core Opteron from Boxx (a very nice computer btw) was significantly cheaper.
I'll send you a 6pack of good beer if you can find me a quad core Opteron from a (as in one) reputable company (that won't go out of business)
Well, this isn't my area of expertise, but I do have the advantage of being somewhat intelligent and rational.
As I stated, I can not cite a reputable scientist in the field as I don't know any though I'm sure you could find plenty of relevent info if you google.
I can, however, see obvious trends in cells and related biological formations TODAY that seem to me as relevant examples of a feasible evolutionary map in general cellular design.
The argument you make is, I presume, based not only the complexity of a complex animal cell but also on those parts of the cell that are also claimed to be irreducably complex like the mitochondria. Let's look at a generic animal cell first. Cell wall, mitochondria, nucleus.. lots of complex stuff in there and it's all wrapped up in a package that just works. Intelligent no?
What about a red blood cell. It's not actually a cell, it's really called a leukocyte as I recall because though it is a self-contained 'cell' it lacks a proper nucleus. In the human body, there are various 'cells' of varying function and complexity.. some more complex than other.
Now pull back to all 'life'. There are other examples of much more remedial cells. Anaerobic organisms often don't posses typical mitochondria since mitochondria are required for aerobic energy metabolism.
I'll have to once again admit my lack of expertise here so I'll skip other intermediaries. We can, however, just jump to the end example.
What about viruses? They are self-contained 'living' units much like cells. Some of them are quite structurally complex with dna injecting mechanisms. They are, however, so simple compared to the typical animal cell that some biologists hesitate to call them alive.
The point I'm trying to make is that, although individual cells are too fragile to be found intact in the fossil record, there are presently a wide variety of cells and sub-cell organisms/organic packages that span a great range of complexity. Personally, though I don't have a full record of every cellular missing link, I don't find it shocking to believe that simple organisms evolved into more complex organism even if those organisms began as little more than replicating and self-replicating protiens (aka. Viruses)
There is a difference between this line of reasoning and faith however. This process, cellular evolution, is currently being studied and it has been for some time. We know WAY more about how cells have or may have evolved into complex structures now than we ever have before. In by know I mean we can build models on our theories and prove them by direct inspection or experimentation. Some models will undoubtedly fail. This is SCIENCE. You throw things at the wall and keep what sticks.
Faith is COMPLETELY DIFFERENT though. You don't throw hypothesis and theory at the wall to see what sticks with Faith. Faith REQUIRES you to belive you know what will stick to the wall without doing the throwing. By definition, if you test the targets of your faith and prove those positions, it is no longer faith it's hard fact. Faith is no longer required at that point. My dictionary tells me that Faith is based on spiritual apprehension RATHER THAN PROOF. That's just the way Faith works.
Science isn't faith. There is hypothesis and theory but there are really very few laws in science. Science is based on the ability to test and disprove any aspect of science. This is just the way science works. If you can't potentially disprove it, it isn't science it's faith. This is especially true of hypothesis and theory. Attempts to disprove hypothesis and theory are every bit as important or even more so than attempts to 'prove' hypothesis and theory. It is, after all, much easier to find one case where a theory fails than to find every possible case (possibly an infinite number) where a theory succeeds. Modern theorys, as a result, become more accurate as exceptions are found an theories evolve to account for those exceptions.
ffakr.
This one's totally off topic but I'm still in shock.
We were driving in Chicago today and I've added a new one to the "now I've seen it all" list.
The guy in front of us remained stopped at a green light, it was clear that he was trying to talk out the passenger window of his car to a woman in another car. She was trying to exit a parking lot of a strip mall, he was blocking her by stopping his car in the middle of a traffic lane right in front of the parking lot exit. This was on a very busy street in Chicago.
My wife was driving and she hit the horn when it was clear he wasn't planning on actually driving through the green light any time soon. He responed by putting the car in park in the lane and getting out to talk to the woman. He looked into our car and I gave him the hands up and I clearly mouthed WTF.. he didn't seem to notice and we went to talk to the girl.
My wife pulled into the other lane to get around him before it escellated any more.
THAT is the quality of drivers in the United States.
No dumbass, But I'd have appreciated it if the GERMAN PEOPLE had stood up to him BEFORE he launched a war that resulted in the persecution of tens of millions, nearly 100 million people (if you look at the massive Russian losses and the exterminated Jews)
Your ignorance is profoundly frightening. I can't figure out how you equated "Give me Liberty or give me death" to capitulation to Hitlers attrocities. Perhaps you can enlighten us.
Birge,
You seem to know a lot of.. well, ficticious information about the Mac OS. It's really fascinating.
Where exactly is Apple still using Resource forks? I haven't run into a Resource fork issue in Years and I run a Mac and Windows support office.
As for Applescript.. I do hate the loose syntax but you probably don't realize that Applescript is actually a real object oriented language. It does have one important advantage over other scripting languages too, intra-application message passing via. AppleEvents. The funny thing about your Apple script comments though are the fact that you're comparing Applescript (which is what, 2 decades old? nearly there? to the shell of a beta OS that's due, maybe, in a year. I think that counts as a win when Applescript (the baby of Apple Employee #5) deserves any comparision to something the Mighty MS hasn't even finished yet.
For Spotlight.. well, it's not a DB filesystem like WinFS.. but since Vista won't ship with WinFS and it's officially vaporware, what's the point? Spotlight doesn't wake up before me and scramble my eggs and put the coffee on either but it's still a sweet tool.
It's got a well documented API. It's got command line tools so I can script it. It's got an Apache Module on OS X Server so I can use it for web searches. It can create search databases on volumes other than my boot drive. It's extensible via plugins. And.. it's fast. It does everything I need a search engine to do. My only complaint is that there should be an easy way to do more complex searches based on various unrelated criteria.. but If I weren't so lazy I'd write that interface.. because I can with the Spotlight framework.
Now back to it not actually being the filesystem. Well, it is tied into the filesystem at a low level (file opens and writes trigger indexing) but more importantly.. It's not the filesystem. WooHoo! Databases make for great archives of information that are easy to access quickly.. but they can be a real bitch when they get corrupted. Thank god my filesystem is just some lame old tree based file system. They are soo much easier to repair than corrupted Databases.
I hate to brag.. no, I don't really.
;-P
I actually thought that the story would be about something like my desk setup.
I'm in a 1950's research building.. 15" concrete floors, concrete block walls and imbedded shelf rails in the walls.
The obvious solution for me was to take a long shelf rail, have the guy in the ship saw a sliver out of it, bend it, and weld it back together so it angles down as it leaves the wall.
I finished a small plank of 1/4" plywood (fancy stuff). I ran it through the router table to put a nice edge on it and clear sealed it.
My Powerbook actually floats over the desk, straight from the wall. It's about Starbucks tall off the desk in the front so I can easily slide my keyboard and mouse under when another notebook comes in for repair.
It's too high by design for normal typing but the angle makes it easy to use in a pinch. It also allows me to use my mouse/keyboard when working on a desktop on my desk.. but I can still use the laptop when needed.
To be honest, I don't think I'm that clever that I should have a better solution than something in a slashdot article. Pretty sad.
I can't believe this stuff passes for science.
I don't look it, but I've got an enormous head. I can't wear anything but an xL hat. I know this is utter crap because I know many quite intelligent people, people I consider as intelligent as I, and the running joke is about my big ass head. I've gone one buddy.. if I toss my hat on him, he looks like he's got progeria because he's got a pea-head and he's skinny. Think bony adult looking child with grown-ups hat on.
This is like someone saying the sky is green. You can present your research, but I've actually looked up a few times.
I wish I could edit posts. :-)
I went and reread my initial post and I can easily see that it appears to let NSIT off the hook too easily. I really intended to say 'don't crucify them yet till you know the facts' but it did come out as 'this isn't their fault, it's the outside web developers'.
I didn't intend that, I only ment to demonstrate that it's more complex an issue that nsit must have done it.
That's what I get for posting just before bed.
Thanks for the additional info. I'll admit, I've still not read the other links like that Maroon article.
In the post you replied to, I was specifically trying to quell the absoluteism and idiocity that is too common on slashdot (by a minority I like to believe).
I primarily wanted to point out that lower pay than the industry doesn't mean there aren't talented people. There are a lot of talented people in NSIT, on campus, and at most if not all Universities. I know plenty of people who are more than smart enough and talented enough to do the grind in corporate. I know some who have and immediately returned.
I'm not excusing NSIT or WHO EVER was directly or passively involved. It's a major F up. I can't imagine why a file with SS#s were up there in the first place (though I've done this work long enough to know that just because I can't see the purpose outside the black box, that doesn't mean it isn't there).
thanks,
ffakr
This is true, and it's absolutely wrong at the same time. How cool is that.
:-)
I'll make the disclaimer now, I'm a UC employee and I work in IT. I'm not affiliated with NSIT, the group under who's watch this problem occurred.
First off. There are plenty of very smart people working at UC. The quality and the size of the central IT staff is superior, imho, to that of my previous employeer.. a State University that was actually larger (plenty of friends on staff at that State university and they are good smart people too.. I'm making broad comparisons here).
As for pay, Yes, Education pays less than the 'real world'. It often pays MUCH less. UChicago doesn't pay salaries comparible to the corporate world but it's far from the worst EDU in the area in this regard. EDU staff doesn't tend to come or stay for the pay, however. Working in EDU is often a geeks dream. Flexible hours.. I know plenty who work 2pm till they are done because they prefer that. Getting good work accomplished is more important than being at your desk at 9am. Great benefits... 4 weeks vacation/personal holiday to start at UC.. more at the last Univeristy I worked for staff with a degree. Overall, a fun, relaxed, and often challanging work place where you help interesting people day in and day out.
Bottom line, there are plenty of very very intelligent, very very talented people doing IT at Universities and they are retained because they'd rather be happy with a smaller pay check than misserable and rolling in dough. (I wish it could be both as much as anyone).
All that said, yes, there are plenty of dumb asses doing IT at Universities. One pattern I see is that smaller groups like Departments and Institutes will hire Techs for internal use but they won't seek out the right people on campus to properly evaluate the applicants. It's awfully easy to BS your way through an interview when the interviewers are the ones desperate for tech support because they are clueless about computers.
Me, I interviewed with a one of those people.. and an associate Director of NSIT. That and an internal reference are the reasons my position isn't currently filled by a real dumbass (just my minor and occasional dumbass-id-ness).
OK, nuff of my rant.
Read the story again. It appears that NSIT may not have even been responsible for creating this problem. The University of Chicago has a professional web developer on retainer. From the story, it looks like a web developer put a sensitive file in the wrong location with the wrong permissions. I'm not sure if this was even caused by an NSIT staff member.
BTW: The current security group on campus is actually quite good. I'm very happy with them, they are really on top of their stuff as far as network monitoring, reporting, and control. Machine get's violated and bam, the port is off with a quickness. The report shows up in our email immediately and we can go off to see who brought an unsecured box into our division with out even bothering to ask us about whether it's buttoned down.
It's my understanding that HT is an attempt to use sections of the CPU that are idle at any given moment. HT isn't a solution that was designed to address problems related to the long pipe like blown branch predictions.
No matter how you look at it though, the G5 is a good candidate for symetric multi-threading (aka. intel-style Hyperthreading). It has a lot of execution units... there plenty of sections of the cpu that are idle in any given clock cycle. The pipeline of the G5 (PPC 970) is also as long as the pipeline in a Northwood P4. To be specific.. different sections of the PPC 970 have different length pipes.. the interger units have shorter pipes than the altivec units.. but they average nearly the same lengh as the Northwood.. up to 23 stages I believe.
So, the 970 is good for HT based on how you believe HT to work, and it's good for how I believe HT to work.
Oh, and IBM has released SMT enabled cpus too.. the Power5 has symetric multi-threading and its a lot more efficient than Intel's HT. So.. there is no particular reason why a wide power[pc] processor can't have SMT. It's just not a feature in in IBM's commodity chips yet.
The suit is a red herring of some sort. Apple can't possibly hope to prevail because the suit claims damage, or at least potential damage if these rumors allow a competetor to duplicate their efforts and come to market first. Think about this for a second.
If TS posts two weeks out of Mac World that Apple is releasing a product.. something that took months and months to design and months to ramp up production.. how is a competetor going to use this information to duplicate their product?
This is likely about the imac. There are already cheap headless computers out there and it takes more than weeks to design and market a new system. It could also be about the rumored iWorks software suite (though not sure if TS commented on this) or leaked Panther stuff, or the rumored firewire direct in box. Do you suspect that releasing this info a couple weeks early would allow a competetor to beat them to market?
First off.. I don't believe you (about the 70's chip being faster than today's fastest Pentium[IV]s)
.13 micron unless you clock it really low. And why is a 4 stage pipe going to magically make this chip as fast as an 8 GHz PentiumX? Because it has a lower latency of ticks? If the latency is 4 ticks but it runs at 200MHz.. and your P4 has a latency of 25 ticks and it runs at 3000 MHz.. the P4 has a lower latency BY TIME! That's what really matters isn't it?
Second.. what you are describing is NOT general purpose CPU but a very very simple DSP or ASIC. DSPs are designed for specific purposes and they run those calculations much faster than a general purpose cpu.. but the problem is, they aren't well suited to doing things like Compression, 3D, and running your word processor at the same time. You want to simplify the instruction set so much that it isn't useful as a computer processor?
third.. REAL RISC, like Alphas, have been proven to be very very fast, but wait till you need to do a calculation that you don't have an instruction for.. then you're building an algorithm with various instructions and that is NOT faster. Back when Alpha's were destroying x86 and PPC boxes in a lot of benchmarks, the Pentium and the PPC even more so would absolutely obliterate it in RC5 bruteforce cracking. Why? Because the Alpha didn't have a required instruction in hardware so they had to work around that in code.
fourth.. I've absolutely no idea what you are talking about with 8GHz performance? What's super cooling? Liquid Nitrogen? Strained Silicon won't get anything like a modern processor to run at 1 volt on
finally.. I wish you luck with your 8000+ (pr rating of course) box that can only add and subtract. I'm sure it will be a fabulous computing experience.
Maybe the poster should have actually read the wiki article they posted to...
Algorithms are named after Abu Abdullah Muhammad bin Musa al-Khwarizmi. He's the father of the algorithm.
I think the confusing here is that the author thinks that algorithm is a concept that only applies to computers.
It was my impression that the original G5 cluster was up and performing real work for several months before it was tore down for a rebuild.
It's probably safe to assume that the work in setting up the first cluster decreased the deployment of the second cluster so it wasn't just an exercise is vanity even if it didn't perform a lot of production calculations.