I will increase my wealth/happiness by $10000 with computer A. I will increase me wealth/happiness by $9000 with computer B. If Computer A and B both cost $3000, I buy A. If B only costs $2000, I am indifferent. If B only costs $1000, I buy B. I determine what gives me the most value.
However, in this case we are comparing two clusters, one of x86 machines running Linux with one of Apple PPC machines running OS X. In either case I am buying many computers.
I need to do X operations per second. How much x86 hardware would this take? How much would it cost? How much Apple PPC hardware would this take? How much would it cost?
You are right that MOST computer buyers look at the price and not the benefit. Almost ANY productivity increase from the Apple makes it a good choice, even if it costs an extra $1000-$2000 for the machine.
However, in this particular case, we are discussing clusters. We are buying a certain amount of computer power. We should compare the variable costs of power ($X/gigaflop, or whatever unit you want to use), plus the fixed costs of setup time, and compare.
If you have a scientific cluster, you don't want to be swapping things out. You don't want to take nodes offline because a video card fried. You want a system that is going to work.
I just priced out some Compaq Workstations yesterday and compared them to Apple Powermacs (Apple's workstations) for doing some OpenGL game development.
Apple Powermac with dual monitors and the upgrades we'd want... $5k. Compaq Workstations... $5k.
In the price-conscious area, Apple's iMacs/iBooks offer a good solution at a reasonable price. You can't compare Apple's workstation line with your "look ma, I built it myself" machine.
Apple does QC. You don't. You and your screw driver does not equal scientific requirements for reliable and predictable. If a node fries, you likely need to start over again. You can't just try to fix the damage.
Linux is great, OS X is great. They are very different UNIXes in different markets.
Alex
That's why I mentioned photoshop
on
Macintosh Clustering
·
· Score: 3, Insightful
Photoshop is optimized for the G4. That was my point. We're not clustering Quake here. We're talking about special purpose applications that do scientific calculations.
If you application does better on the Intel, you are likely better off considering a Linux cluster. However, if it isn't much better, you might be better off with the Mac cluster by adding a few more machines to compensate... depends on the costs of time.
If you are running an application that, LIKE Photoshop, does better on the G4, you will see the price performance favor the Mac line. That's my point.
If this market was a decent size, I bet Apple could get some really competitive cluster systems. It would be nice to see an Apple dual or quad G4-1 GHz, with a CD-ROM, ATI Rage 128, and Gigabit Ethernet for the scientific community.
They could make the machine without PCI slots and fit in a 1U case for OS X processing goodness.
However, the reality is that the extras (better video card, Superdrive, etc.) don't add much to the Apple's price. However, the right form factor could make them tremendous cluster machines.
Cost of 10 good Intel machines to install Linux on... trivial (pobably about $15,000)...
Cost of 10 good Highend Macs, (about $30,000)...
Both are in the trivial range compared to the costs of time, energy, etc.
There is a more important question, which machine gives you the most bang for your buck?
We know that Photoshop runs better on the G4, what about your operation?
If the Mac gets a 2:1 performance advantage, then the costs are equal. If the Mac out-performs it regardless, you get an advantage.
For the moment, let's assume that you are getting real machines that are tested, not parts off of a sketchy vendor from pricewatch.com. If you are really trying to build a parallel computer, you want real systems, not junk that may or may not work.
This also rules out eMachines, or home computers. You are basically in the Compaq Workstation, Dell Workstation, HP Workstation, or IBM Workstation area. You aren't setting up a bunch of Presarios.
Cause we've got the bomb... or more directly, because the might of the United States military protects US citizens from these dangers.
Any country is welcome to try to hold the US President criminally responsible for violating their laws. My money is on the US Military keeping Bush our of their court room.
If you are a third world dictator and want to keep out of the US Courts, build a military that can compete.
Notions of International Justice and human rights are great ideas to either mentally masterbate over or use as a motivational tool for actions.
If you were to try to reverse the situation you wouldn't here about sovereignty, you'd hear about beating the bad guys.
Look, the free world exists because of the US military's actions from 1917-1991. Until you accept that, don't expect American's to care about your perceived injustices.
Yes, Americans think they run the world.
History of the previous century has given Americans no indication that they are wrong.
Actions of last fall attempted to put the US in its place... things are looking better right now.
Alex
Lose the 90% bull and you're fine
on
LinuxWorld Preview
·
· Score: 3, Interesting
Yes, Office users probably range from 5%-20% of Office's functionality. Unfortunately, outside of the 5% common stuff, the other portions that people use are different.
If you are throwing together high school papers, you don't need word. However, for a corporate environment, they are powerful.
Look, if you don't need Office, you don't buy Office. MS Works exists, and it is similar in capability to Apple's AppleWorks and most of the "Free" office suites.
The problem is the people that push these apps. Take Excel. If you just did some graphing in high school chem, you may be satisfied with Gnumetric. Have you ever seen a real finance guru with Excel? They churn out really complex items. What does this mean for the rest of the corporation? If I want to view their spreadsheets, I need Excel. Therefore, anyone in the Enterprise that needs to work with these spreadsheets needs excel.
Now it doesn't matter that 90% of the Enterprise doesn't need Excel's features for their work, another 20% may work with the finance people that are cranking out excel.
Now we have Access. Access is a silly app, but it is damned convenient for basic databases. Sure, I can crank out an impressive web GUI and build a real database on PostgreSQL, but if a small department wants something without clearing it by IT, Access gets them up and running quickly.
Powerpoint?
Powerpoint sucks, everyone who uses it knows that it sucks.
Name a competitor that sucks less. (Please, I've been looking for a better product, I can't find one). However, if you need a quick presentation, it gets the job done. It's easy to use, and everyone either HAS Powerpoint (from Office) or can get the free viewer (which you can send them). It is a quick way to send ideas including graphical explanations.
Word is extremely flexible. Most people in the organization don't need it. However, a handful likely push Word to its limits. They build the templates and otherwise utilize its features. Now, if the rest of the company is using Word, they can leverage these creations. No Word? No luck.
Sure, VBA isn't useful for most users. If your IT department found a use for it, then they'll crank stuff out. I've been at clients that really use VBA, and many that never use it.
If you guys spent 10% of the time in various IT rooms at real companies instead of listening to other Linux-heads on Slashdot you'd understand Linux's deficiencies on the desktop.
For a home computer, Linux is adequate. For a corporate environment, most need more.
Ya know, for all Outlook's security problems, the group scheduling and other features when combined with Exchange are REALLY powerful.
The Radeon has been smoking the equivalent (pricewise) Nvidia card on the Mac line. Nvidia's name sells however, not ATi. PC Converts that know something about hardware respect Nvidia in a way they don't respect ATi. As a result, Nvidia is default, ATI is the built to order option.
The Mac benchmarks were showing the Radeon line in a very good light. Also, the Geforce4 MX supports dual monitors (VERY important if you do Photoshop editting) on a single card, and other niceties.
It's a good default graphics option. If you are a gamer, save the $100 and get the Radeon card, then upgrade to a faster gaming card in the future.
We're talking about real development environments. Most development (including open source work) is done at work. People work on projects that their company needs. Sure, there are lots of hackers that bang out unfinished code on source forge at home, but the real work on the Linux kernel, Apache, PHP, etc., is done as part of commercial endeavors. The interesting thing is that instead of building internal tools and keeping them to yourself, you share with others and work together to build the glue you need.
If the new development machine will save 80 hours over 2 years, you're a fool if you don't drop $3000 on a new computer.
Most people purchasing computers aren't in the "no money," give me "free or death" crowd that frequents Slashdot. All of us were hobbyists that had a $400/year computer budget. If you start making a living with your computer, you spend the money on the tools you need to make a living and be competitive.
Spending 8 hours playing with sound drivers on a Linux system (which I've done) is suicide in a work environment. I already ate any savings that we got by going with Linux as opposed to OS X. That's my point.
Once your time has value, Linux on unsupported hardware isn't a bargain anymore.
If computer costs really factor into your decision, you don't make a living coding.
My PC environment was easily over 5K when including a laptop (bought 3 months used to save $800 off the costs), a docking station at home and the office, Win2K, Off2K Pro, Visio Enterprise Edition, Text Editors that don't suck, etc.
We may switch the office over to Macs. The OS X experiment was promissing, but the platform isn't there yet. 6 Months? Hell yeah. We do PHP/Java development. The cost of the Windows machines don't even account for the hardware to have a Unix development environment to actually work in.
I know that you read/. from time to time, as I got an e-mail from you in response to a posting. Perhaps you can enlighten us here, because I'm really confused.
In discussing the LGPL vs GPL for libraries, you mention the idea that if the ability doesn't exist outside of the library (ie readline) you should GPL it. Then, if someone wants to use your library, they need to GPL it, and this advanced free software.
However, if you are reimplementing a standard (i.e. glibc) then you should use the LGPL so that others can build on your work.
So, assuming we shared your goals of using licensing to advance free software, I still don't see how this hurts.
Right now, in the pragmatic marketplace, the Unix vendors are retreating up the ladder. Linux and GNU based systems are replacing the low-end UNIX system. Proprietary UNIX is slowly being confined to areas where Free Unix-like OSes can't perform. I think that worrying about liberating Unix users is quite silly. At this point, any markets that Unix competes in will belong to GNU when it matures to that level. UNIX isn't the enemy, its the advanced team. Crippling the commercial UNIXes in a Unix vs. MS fight really hurts free software, as we have a Free Unix, but not a free Windows. The Free Unix will displace the non-Free Unixes, but if the service runs on Windows, you won't liberate those users.
From this view point, I fail to see how this licensing change hurts thing? These classes are duplicates of the Microsoft classes. As they are based upon compatibility, you can't really do much with them directly. I don't see the leverage that even GPL'd versions give you.
If your goal is to prevent Sun from using this work to sell Solaris in this market, I think you are missing the situation here. The first choice that is made is Unix vs. WinNT. If WinNT wins, then your free tools are ignored. If Unix wins, then GNU systems get the job if they can handle it, otherwise a Unix is chosen. When the server is replaced in 2-3 years, it will likely be replaced by a GNU system.
We can't offer things that Sun and HP can. If they do the job, GNU systems kick in when they can handle it. If Win32 gets the job, you are unlikely to liberate them.
Please, explain how crippling the development efforts advanced free software?
GNUstep could have done wonders had the project been nearly completed 3-4 years ago. It is just coming to maturity now, and will likely me 2 years from true usefulness.
This industry moves quickly, and GNU is making it move faster. Any space gets eaten by Free Software within 5 years of existance now, with good prototypes in 2-3 years. Isn't it simply enough to speed up the Free Software Goliath? Why attack the Unix vendors, they're adopting the GNU way slowly as they can.
We were trying to add old BBS doors support to a friends BBS years ago, and we could either get a stack of 286s or a reasonable machine with Desqview 386. Well, Desqview wasn't on the market anymore, so we tried OS/2, Win95, and later WinNT, none would handle our doors. We tried to warez it but failled. We later tried a stack of 286s, but the systems weren't playing nicely with our NT Server (didn't have the expertise or budget for an admin for a Novell server).
A few years ago it would have been great for me. Maybe I'll drop the cash and try the system now...
Alex
So WINElib = Carbon?
on
Wired Talks Wine
·
· Score: 3, Interesting
Intersting, however OS 9 apps run in a box of sorts. They can take over the screen, but it is clear when an app is running in classic mode. It is much the same as where VMWare could take you.
WINElib has the potential to do for Linux what Carbon did for OS X.
Carbon is an OS X native API that is based on the classic API. Additionally, Carbon was ported to OS 8 and OS 9. This meant that you could have been developing for Carbon the past few years and having OS X native applications that ran under Mac OS. For extra fun, they could have FAT binaries (I think that I'm using the term right, they had something like that) where they could include a Classic PPC binary, Classic 68K binary, and Carbon OS X binary all as one application.
WINElib is interesting, you can build against WINElib and compile for Windows and Linux, supporting both platforms with native applications. The trick is a strategy that lets you target both OSes for now, it lets you keep your Windows market and expand into the Linux market as it matures.
Personally, I think that Apple should work on getting WINElib to be Aquafied. Then you could build targetting WINElib for Windows/OS X, and Linux or other UNIXes. Obviously you'd hate to make Win32 the standard API, but Apple dropped it when they dropped OpenSTEP for Win32, so oh well.
Wow, everyone that I saw posting apparently believes that their home computer is the end all and be all of computing. You're an idiot.
Look, we only use a handful of Linux machines, so we aren't likely to use this. However, if I was rolling out 1000 workstations in my enterprise, and we were tweaking/tuning the OS before rolling it out, recompiling with this would work.
Assuming Red Hat makes compiling under the Intel compiler a requirement for inclusion in their distribution, they're in a great situation.
Why not compile everything with an optimized compiler? You still have the freely redistributable GCC for compiling open source code, but for stuff that is being downloaded in binary format, wouldn't you want it to run faster?
Does it compile quicker? Who cares. When you are doing software development, you want something that compiles quickly. When you are rolling out a production environment, free speed is good.
Look, your precious GCC is terrific, it is a flexible, cross-platform compiler. It's always been week on the performance. The GCC team has always made it clear that the biggest problem ISN'T processor-specific tweaks, its general compiler improvements that are patented.
GCC is a baseline, things should compile with it. Things should also compile to the POSIX standard. That doesn't mean you don't add tweaks on the platforms that you support and set it up so that./configure figures out which to use.
Give me a break. I realize that many of you just use Linux to configure and tweak Linux to the point that you can post on Slashdot about how you can do anything with Linux. However, those of us that have included it as one of our tools to solve problems can use ANY tools that are made available to us.
If I can get a 47% performance improvement by recompiling some of my applications, terrific. Replacing the server may be cheap in terms of hardware (a few grand for a new server every 6-12 months isn't bad, its one of the few reasons to use x86 servers), but it takes time. Building and testing new hardware is easily 2-3 man-weeks before TESTING (expensive, look at your salaries and double it to estimate costs to the company), recompiling on your test machine and testing is just the testing time.
You could do the same abuse with less elegant solutions than NAT. Simply running a simple Proxy server for your neighbors would provide them access. Only 1 machine is on the Internet, the rest aren't. Hell, if you are running MS's busted proxy, the rest don't even need TCP/IP, they could run IPX/SPX. (Lousy program, NEAT configuration options, I never want to go near it again...)...
Myself, I have a $90/month DSL connection. Why? If I need to get a VNC connection through the VPN to a work machine, I want the 384K uplink.
We have a NAT box with wireless, and technically, 4 computers there. I live with my fiancee. She web browses from her iBook, and I work from home on the weekends. We barely use the bandwidth.
However, I pay the premium so it is there when I need it.
Ban NAT and I lose Wireless. If that is the case, I drop DSL. I can't run Wires all over my apartment, so I use Wireless to send the signals around.
Find the abusers, by all means. However, leave those of us that don't abuse it alone.
Relax, take a deep breath. Stop calling people stupid or morons. This isn't personal. I understand that you're stuck not being in America, but that is no reason to be angry (that was a joke, BTW, bring on the flames).
This Slashdot dream of taking over people's computer desktops everywhere is a silly one. I'm suggesting that there are people that will have a need for a KDE Desktop, and it doesn't matter that it isn't your mother.
You, however, choose to go ballistic. Relax, it isn't that important.
The desktop PC wars are over. They were fought between Microsoft and Apple, Apple lost.
We now have lots of processing power and lots to do with the machines. It doesn't all need to be general purpose. There is a role for KDE to play, even in a MS desktop dominated future.
OTOH, keeping MS to 95% of the market (or even, joy, rolling them back to 90%) would be huge. Don't let them own the web browser market and you are okay. Keep open protocols. If we are constantly reverse engineering their stuff, we lose,.NET wins.
MS owns the desktop PC market. Who cares, we're all moving on.
The notion of a personal computer is from the past. People use their network interface device. Right now it is a PC, in the future, who knows.
Microsoft Homestation is their answer for consumer access. AOL will likely have their answer. Apple seems to be ignoring that market and focusing on people that have money and want digital toys. The new iMac with DVD burning, iDVD, iMovie, and iPhoto aims for this market.
The work environment? You need a system that supports something like Outlook and MS Office. Microsoft owns this market and will for the forseeable future.
What markets are in flux?
Web programmers. As we move for web servers, we need systems for people that program them..NET programmers will need Win32 desktops, MS owns that.
However, Java programmers, PHP programmers, etc., will likely want to consider Unix desktops. Linux can fill a niche here. They still need e-mail, word processing, and printing.
Tablet PCs, open market. Linux based solutions can compete with Windows based solutions.
Television computing... who wants to fight the Homestation? Tivo, you going to step up with Linux? AOL, what are you going to use? These are the markets to fight in.
Sure, the KDE/GNOME desktop may not make it there... Microsoft's Explorer (the Win95 and up GUI) won't either. However, if it is Windows based, COM/DCOM/ActiveX/OLE will be used. If it is Linux/KDE/Qt based, then Kparts will be used.
Developers need a desktop to develop for the target system. If you are doing a Kparts/Qt/KDE/Linux set top box, what makes more sense, a KDE Desktop, an OS X Desktop, a GNOME Desktop, or a Windows desktop?
Welcome to the networked world. We can all pick our platform. The Microsoft monopoly will die... long live the Microsoft monopoly. Alternatives to Windows for the non-PC market is important to stopping the market.
Sure Microsoft will be a player, but they don't need to be the only one.
It sounds far-fetched, and yet there are so many very intelligent students who graduated at the same time as me whose only goal is to get out of the western market lifestyle at any cost because they feel that the nature of ideas has been fundamentally changed, from a kind of forum for the enrichment of man to a tightly-controlled, tightly-protected profit-making establishment serving only this new caste system.
And some of us didn't go to school in all of California, let alone Berkeley.
I don't know that I buy the arguement that more advertisement = less purchasing. I would assume by an increased attempt to advertise these items that people are registering them.
The engineer in me HATES x86. I understand why it was a good solution in the late 70s and early-mid 80s. It became an ugly hack in the early 90s, and is now a legacy solution.
Including support for "legacy" assembly code in Microsoft's first generation video game console is stupid.
The Gamecube has horrible software for the crucial 16-24 segment, but find outside of that. At 22, I'm unusual in LIKING Nintendo's offerings. We'll see what happens, because my peers were kids for the NES days, in college during the N64 days, and now out starting to earn a living. We'll see what happens, but I think that Gamecube has a shot.
I REALLY like Luigi's mansion. Pikmin is awesome. I don't like DOA3 or Halo.
I have Rogue Leader II, it's fun to show off my Gamecube in surround sound and 480p. However, I don't really care for the game.
I'm a 22 year old college graduate, and I like the game cube.
This Linuxhead you slammed is an MCSE. I kicked Linux out of my server room a while ago. We run OpenBSD for almost all our Unix needs (I run Linux database servers far from the Internet because of OpenBSD performance problems).
Look at the Xbox hardware. It is a Celeron processor with OTS components that are found in PCs. The Gamecube uses a custom processor that was based on IBM's desktop CPU, but with a custom GPU and unusual memory configurations.
The system, componentwise, is a PC, with a custom OS (using kernel code from Win2K, but a completely custom Ring-0 OS). The kicker to me is the inclusion of DirectX. DirectX means that applications for WinXbox are source compatible with Win32.
All the yelling in the world that this is a custom system built for gaming doesn't make it so. It is a PC with a good gaming sound card AND a special chip that converts PC resolutions -> HDTV resolutions (taking up to 1024x768 resolutions).
It burns cash for MS because the component costs are expensive. However, their R&D costs are really low. It's a tradeoff.
However, if you rev the system regularly, there is less time to recoup R&D costs. The Sony, Sega (Saturn and later), and MS strategy is to burn money early on and make money late in the console's life. A R&D light approach let's MS put everyone out of business.
I disagree with you on an assessment of computer hardware so I'm a moron?
I'm not a huge Halo fan. It looks good, but I don't like FPS games. However, I really enjoy Luigi's Mansion. It's a fun game, not particularly deep, but fun and silly.
However, I think that picking two arbitrary games (one system's flagship with 5 years of development, the other a technical demo that morphed into a short game) and comparing the systems is a little silly.
The hardware is different.
The GPUs are really hard to tell. I THINK that the Xbox GPU is stronger, but I don't know enough about the Gamecube's to know for sure.
The CPUs are night and day. The Xbox has all the legacy garbage of the Intel Celeron, compared to the sleak Gamecube CPU. The Gamecube CPU is a more intelligent design, better at floating point, and overall should kick its butt.
Xbox has more memory, point Xbox. Given the use of standard computer RAM, it should have 128MB, not 64MB. Stupid decision. As a result, this advantage become more minor than it should.
GCN has faster RAM. There is RAM everywhere, with highspeed interconnects.
This lends itself towards not having slowdowns. This should allow the GCN to sit closer to 100% utilization all the time.
The game development environment is more condusive to building games with depth? What the fuck are you talking about. The Nintendo style of not having FMV lends itself to games that you PLAY more.
The Xbox is a crappy PC thrown in a black box on the theory that games will be better because the hardware is fixed. However, if you lock all the hardware behind an abstraction layer, you don't really have much room to optimize.
The advantage to the Xbox is that you can release games NOW that require its hardware, while PC games normally have to target PCs 2-3 years old. Given that everything save the GPU is 2-3 years old in the Xbox, this isn't a REAL advantage.
Xbox has the hard drive and larger memory areas, this lends itself towards these deeper games that you discuss. However, having a GPU that shares the memory really knocks down that memory advantage.
The Xbox is a sloppy system, justified entirely on hype based upon numbers that aren't that good when you analyze them.
Two years and the Xbox is toast.
Microsoft's strategy may be based upon the fact that they can crank out PC-based systems with near zero R&D. This means that they can rev the hardware every 3 years. Shrink the lifespan of the consoles, and you cramp Sony and Nintendo's abilities to do R&D. Take all the money out of the industry and Sony pulls out, and Nintendo gets hurt. Then you can leave crap out there and sell games.
With Windows, MS needs to force you to WANT to upgrade. With an X-box monopoly, they just need you to WANT a new game, no need to put out new hardware or anything.
I love my GBA. It isn't a substitute. It's awesome, has a monopoly on the handheld market, but it isn't in the same competition as the others. You don't see people deciding between a GBA and GCN, or a GBA and PS2. If you want a handheld, you get a GBA.
The games rock, its a money-maker for Nintendo, etc. However, it isn't competition for home systems.
I won't grant that the Gamecube controller is a rip-off of the Playstation's controller.
To me it is a natural evolution.
The Playstation's controller was a rip-off of the SNES controller that got extra shoulder pads and LATER analog sticks (after they ripped the idea off Nintendo).
The Gamecube controller is very clearly derivative of the N64's controller. The left hand options before (D-pad + L, or Analog + Z) have been merged into (D-pad OR Analog) + L with the added bonus of the D-pad and Analog both being usable with an easy switch.
The button layout on the right is the SNES layout reoriented around the reality of a primary button (A), secondary button (B) and optional extra buttons (X, Y).
The SNES had four equal buttons, but they were rarely uses as such.
The Z-button? It's a hack tacked on at the last minute for people worried about losing a button. There it is, developers, please don't use it much.
Analog shoulder pads, brilliant new invention (like N64's analog stick, SNES's shoulder pads, and NES's D-pad) that everyone will copy.
C-stick, it's neat. An adaption of the C-buttons into a stick. The C-buttons had the advantage of letting the N64 ACT like a 6-button controller (for things like Street Fighter).
Nintendo's N64 controller was large and unwieldy but REALLY flexible.
Games didn't use the flexibilty.
Gamecube keeps the controls and options and tweaks the layout to be more useful.
No it isn't 3-controllers in one (theoretically, N64 = Dpad + buttons, Dpad + analog, Analog + buttons), but 1 awesome controller with everything in a clsoe distance.
Yes the Playstation dual-shock is a nice controller (once you get past the shock of not grabbing the left side and having the primary controller there... drove me crazy on the N64 and hit me again now), but the GCN isn't a ripoff of Sony.
X-box has a custom SIMD by IBM. Motorola wrote Altivec for the G4. IBM wrote a custom SIMD for the G3. Same platform (PPC), similar idea (SIMD), years newer (IBM's implementation).
I used Altivec as an example. Gamecube has its own SIMD implementation taht was designed with the particular instructions Nintendo asked to be implemented for gameplay. It is probably better than Altivec is because it is newer and custom.
I'm anti-bullshit, and the Xbox is full of it.
The greatest GPU in the world? Who cares? You're playing at 640x480x60 at best, I don't need a Nvidia Geforce-12000. I need a gaming system that can do innovative a new things. The Gamecube GPU is pretty good, maybe even comparable to the Nvidia one in the Xbox.
We're all ASSUMING that the Xbox has the best GPU, as we are more familiar with Nvidia than the company that ATI bought. However, the Gamecube stuff apparently blows away ATI's in-house tech, and the ATI Radeon 8500 competes pretty well with the GeForce 3, so who the hell knows.
Gamecube's hardware may be much better than Xbox and we won't know.
However, it should be obvious to anyone taht has ANY background in microprocessor design (no, reading Tom's Hardware Guide or Anandtech is not an education, it's a benchmark shootout) that the Gamecube CPU will beat MS's clunky system.
You're only as fast as your weakest link, so all MS's other advantages are relatively trivial.
Slashdot Mirror
You are attempted to determine the value of X.
I will increase my wealth/happiness by $10000 with computer A. I will increase me wealth/happiness by $9000 with computer B. If Computer A and B both cost $3000, I buy A. If B only costs $2000, I am indifferent. If B only costs $1000, I buy B. I determine what gives me the most value.
However, in this case we are comparing two clusters, one of x86 machines running Linux with one of Apple PPC machines running OS X. In either case I am buying many computers.
I need to do X operations per second. How much x86 hardware would this take? How much would it cost? How much Apple PPC hardware would this take? How much would it cost?
You are right that MOST computer buyers look at the price and not the benefit. Almost ANY productivity increase from the Apple makes it a good choice, even if it costs an extra $1000-$2000 for the machine.
However, in this particular case, we are discussing clusters. We are buying a certain amount of computer power. We should compare the variable costs of power ($X/gigaflop, or whatever unit you want to use), plus the fixed costs of setup time, and compare.
Alex
If you have a scientific cluster, you don't want to be swapping things out. You don't want to take nodes offline because a video card fried. You want a system that is going to work.
I just priced out some Compaq Workstations yesterday and compared them to Apple Powermacs (Apple's workstations) for doing some OpenGL game development.
Apple Powermac with dual monitors and the upgrades we'd want... $5k. Compaq Workstations... $5k.
In the price-conscious area, Apple's iMacs/iBooks offer a good solution at a reasonable price. You can't compare Apple's workstation line with your "look ma, I built it myself" machine.
Apple does QC. You don't. You and your screw driver does not equal scientific requirements for reliable and predictable. If a node fries, you likely need to start over again. You can't just try to fix the damage.
Linux is great, OS X is great. They are very different UNIXes in different markets.
Alex
Photoshop is optimized for the G4. That was my point. We're not clustering Quake here. We're talking about special purpose applications that do scientific calculations.
If you application does better on the Intel, you are likely better off considering a Linux cluster. However, if it isn't much better, you might be better off with the Mac cluster by adding a few more machines to compensate... depends on the costs of time.
If you are running an application that, LIKE Photoshop, does better on the G4, you will see the price performance favor the Mac line. That's my point.
If this market was a decent size, I bet Apple could get some really competitive cluster systems. It would be nice to see an Apple dual or quad G4-1 GHz, with a CD-ROM, ATI Rage 128, and Gigabit Ethernet for the scientific community.
They could make the machine without PCI slots and fit in a 1U case for OS X processing goodness.
However, the reality is that the extras (better video card, Superdrive, etc.) don't add much to the Apple's price. However, the right form factor could make them tremendous cluster machines.
Alex
Cost of 10 good Intel machines to install Linux on... trivial (pobably about $15,000)...
Cost of 10 good Highend Macs, (about $30,000)...
Both are in the trivial range compared to the costs of time, energy, etc.
There is a more important question, which machine gives you the most bang for your buck?
We know that Photoshop runs better on the G4, what about your operation?
If the Mac gets a 2:1 performance advantage, then the costs are equal. If the Mac out-performs it regardless, you get an advantage.
For the moment, let's assume that you are getting real machines that are tested, not parts off of a sketchy vendor from pricewatch.com. If you are really trying to build a parallel computer, you want real systems, not junk that may or may not work.
This also rules out eMachines, or home computers. You are basically in the Compaq Workstation, Dell Workstation, HP Workstation, or IBM Workstation area. You aren't setting up a bunch of Presarios.
Cause we've got the bomb... or more directly, because the might of the United States military protects US citizens from these dangers.
Any country is welcome to try to hold the US President criminally responsible for violating their laws. My money is on the US Military keeping Bush our of their court room.
If you are a third world dictator and want to keep out of the US Courts, build a military that can compete.
Notions of International Justice and human rights are great ideas to either mentally masterbate over or use as a motivational tool for actions.
If you were to try to reverse the situation you wouldn't here about sovereignty, you'd hear about beating the bad guys.
Look, the free world exists because of the US military's actions from 1917-1991. Until you accept that, don't expect American's to care about your perceived injustices.
Yes, Americans think they run the world.
History of the previous century has given Americans no indication that they are wrong.
Actions of last fall attempted to put the US in its place... things are looking better right now.
Alex
Yes, Office users probably range from 5%-20% of Office's functionality. Unfortunately, outside of the 5% common stuff, the other portions that people use are different.
If you are throwing together high school papers, you don't need word. However, for a corporate environment, they are powerful.
Look, if you don't need Office, you don't buy Office. MS Works exists, and it is similar in capability to Apple's AppleWorks and most of the "Free" office suites.
The problem is the people that push these apps. Take Excel. If you just did some graphing in high school chem, you may be satisfied with Gnumetric. Have you ever seen a real finance guru with Excel? They churn out really complex items. What does this mean for the rest of the corporation? If I want to view their spreadsheets, I need Excel. Therefore, anyone in the Enterprise that needs to work with these spreadsheets needs excel.
Now it doesn't matter that 90% of the Enterprise doesn't need Excel's features for their work, another 20% may work with the finance people that are cranking out excel.
Now we have Access. Access is a silly app, but it is damned convenient for basic databases. Sure, I can crank out an impressive web GUI and build a real database on PostgreSQL, but if a small department wants something without clearing it by IT, Access gets them up and running quickly.
Powerpoint?
Powerpoint sucks, everyone who uses it knows that it sucks.
Name a competitor that sucks less. (Please, I've been looking for a better product, I can't find one). However, if you need a quick presentation, it gets the job done. It's easy to use, and everyone either HAS Powerpoint (from Office) or can get the free viewer (which you can send them). It is a quick way to send ideas including graphical explanations.
Word is extremely flexible. Most people in the organization don't need it. However, a handful likely push Word to its limits. They build the templates and otherwise utilize its features. Now, if the rest of the company is using Word, they can leverage these creations. No Word? No luck.
Sure, VBA isn't useful for most users. If your IT department found a use for it, then they'll crank stuff out. I've been at clients that really use VBA, and many that never use it.
If you guys spent 10% of the time in various IT rooms at real companies instead of listening to other Linux-heads on Slashdot you'd understand Linux's deficiencies on the desktop.
For a home computer, Linux is adequate. For a corporate environment, most need more.
Ya know, for all Outlook's security problems, the group scheduling and other features when combined with Exchange are REALLY powerful.
Alex
The Radeon has been smoking the equivalent (pricewise) Nvidia card on the Mac line. Nvidia's name sells however, not ATi. PC Converts that know something about hardware respect Nvidia in a way they don't respect ATi. As a result, Nvidia is default, ATI is the built to order option.
The Mac benchmarks were showing the Radeon line in a very good light. Also, the Geforce4 MX supports dual monitors (VERY important if you do Photoshop editting) on a single card, and other niceties.
It's a good default graphics option. If you are a gamer, save the $100 and get the Radeon card, then upgrade to a faster gaming card in the future.
For professionals, this is a great video card.
We're talking about real development environments. Most development (including open source work) is done at work. People work on projects that their company needs. Sure, there are lots of hackers that bang out unfinished code on source forge at home, but the real work on the Linux kernel, Apache, PHP, etc., is done as part of commercial endeavors. The interesting thing is that instead of building internal tools and keeping them to yourself, you share with others and work together to build the glue you need.
If the new development machine will save 80 hours over 2 years, you're a fool if you don't drop $3000 on a new computer.
Most people purchasing computers aren't in the "no money," give me "free or death" crowd that frequents Slashdot. All of us were hobbyists that had a $400/year computer budget. If you start making a living with your computer, you spend the money on the tools you need to make a living and be competitive.
Spending 8 hours playing with sound drivers on a Linux system (which I've done) is suicide in a work environment. I already ate any savings that we got by going with Linux as opposed to OS X. That's my point.
Once your time has value, Linux on unsupported hardware isn't a bargain anymore.
Alex
If computer costs really factor into your decision, you don't make a living coding.
My PC environment was easily over 5K when including a laptop (bought 3 months used to save $800 off the costs), a docking station at home and the office, Win2K, Off2K Pro, Visio Enterprise Edition, Text Editors that don't suck, etc.
We may switch the office over to Macs. The OS X experiment was promissing, but the platform isn't there yet. 6 Months? Hell yeah. We do PHP/Java development. The cost of the Windows machines don't even account for the hardware to have a Unix development environment to actually work in.
Alex
I know that you read /. from time to time, as I got an e-mail from you in response to a posting. Perhaps you can enlighten us here, because I'm really confused.
In discussing the LGPL vs GPL for libraries, you mention the idea that if the ability doesn't exist outside of the library (ie readline) you should GPL it. Then, if someone wants to use your library, they need to GPL it, and this advanced free software.
However, if you are reimplementing a standard (i.e. glibc) then you should use the LGPL so that others can build on your work.
So, assuming we shared your goals of using licensing to advance free software, I still don't see how this hurts.
Right now, in the pragmatic marketplace, the Unix vendors are retreating up the ladder. Linux and GNU based systems are replacing the low-end UNIX system. Proprietary UNIX is slowly being confined to areas where Free Unix-like OSes can't perform. I think that worrying about liberating Unix users is quite silly. At this point, any markets that Unix competes in will belong to GNU when it matures to that level. UNIX isn't the enemy, its the advanced team. Crippling the commercial UNIXes in a Unix vs. MS fight really hurts free software, as we have a Free Unix, but not a free Windows. The Free Unix will displace the non-Free Unixes, but if the service runs on Windows, you won't liberate those users.
From this view point, I fail to see how this licensing change hurts thing? These classes are duplicates of the Microsoft classes. As they are based upon compatibility, you can't really do much with them directly. I don't see the leverage that even GPL'd versions give you.
If your goal is to prevent Sun from using this work to sell Solaris in this market, I think you are missing the situation here. The first choice that is made is Unix vs. WinNT. If WinNT wins, then your free tools are ignored. If Unix wins, then GNU systems get the job if they can handle it, otherwise a Unix is chosen. When the server is replaced in 2-3 years, it will likely be replaced by a GNU system.
We can't offer things that Sun and HP can. If they do the job, GNU systems kick in when they can handle it. If Win32 gets the job, you are unlikely to liberate them.
Please, explain how crippling the development efforts advanced free software?
GNUstep could have done wonders had the project been nearly completed 3-4 years ago. It is just coming to maturity now, and will likely me 2 years from true usefulness.
This industry moves quickly, and GNU is making it move faster. Any space gets eaten by Free Software within 5 years of existance now, with good prototypes in 2-3 years. Isn't it simply enough to speed up the Free Software Goliath? Why attack the Unix vendors, they're adopting the GNU way slowly as they can.
Alex
We were trying to add old BBS doors support to a friends BBS years ago, and we could either get a stack of 286s or a reasonable machine with Desqview 386. Well, Desqview wasn't on the market anymore, so we tried OS/2, Win95, and later WinNT, none would handle our doors. We tried to warez it but failled. We later tried a stack of 286s, but the systems weren't playing nicely with our NT Server (didn't have the expertise or budget for an admin for a Novell server).
A few years ago it would have been great for me. Maybe I'll drop the cash and try the system now...
Alex
Intersting, however OS 9 apps run in a box of sorts. They can take over the screen, but it is clear when an app is running in classic mode. It is much the same as where VMWare could take you.
WINElib has the potential to do for Linux what Carbon did for OS X.
Carbon is an OS X native API that is based on the classic API. Additionally, Carbon was ported to OS 8 and OS 9. This meant that you could have been developing for Carbon the past few years and having OS X native applications that ran under Mac OS. For extra fun, they could have FAT binaries (I think that I'm using the term right, they had something like that) where they could include a Classic PPC binary, Classic 68K binary, and Carbon OS X binary all as one application.
WINElib is interesting, you can build against WINElib and compile for Windows and Linux, supporting both platforms with native applications. The trick is a strategy that lets you target both OSes for now, it lets you keep your Windows market and expand into the Linux market as it matures.
Personally, I think that Apple should work on getting WINElib to be Aquafied. Then you could build targetting WINElib for Windows/OS X, and Linux or other UNIXes. Obviously you'd hate to make Win32 the standard API, but Apple dropped it when they dropped OpenSTEP for Win32, so oh well.
Alex
Wow, everyone that I saw posting apparently believes that their home computer is the end all and be all of computing. You're an idiot.
./configure figures out which to use.
Look, we only use a handful of Linux machines, so we aren't likely to use this. However, if I was rolling out 1000 workstations in my enterprise, and we were tweaking/tuning the OS before rolling it out, recompiling with this would work.
Assuming Red Hat makes compiling under the Intel compiler a requirement for inclusion in their distribution, they're in a great situation.
Why not compile everything with an optimized compiler? You still have the freely redistributable GCC for compiling open source code, but for stuff that is being downloaded in binary format, wouldn't you want it to run faster?
Does it compile quicker? Who cares. When you are doing software development, you want something that compiles quickly. When you are rolling out a production environment, free speed is good.
Look, your precious GCC is terrific, it is a flexible, cross-platform compiler. It's always been week on the performance. The GCC team has always made it clear that the biggest problem ISN'T processor-specific tweaks, its general compiler improvements that are patented.
GCC is a baseline, things should compile with it. Things should also compile to the POSIX standard. That doesn't mean you don't add tweaks on the platforms that you support and set it up so that
Give me a break. I realize that many of you just use Linux to configure and tweak Linux to the point that you can post on Slashdot about how you can do anything with Linux. However, those of us that have included it as one of our tools to solve problems can use ANY tools that are made available to us.
If I can get a 47% performance improvement by recompiling some of my applications, terrific. Replacing the server may be cheap in terms of hardware (a few grand for a new server every 6-12 months isn't bad, its one of the few reasons to use x86 servers), but it takes time. Building and testing new hardware is easily 2-3 man-weeks before TESTING (expensive, look at your salaries and double it to estimate costs to the company), recompiling on your test machine and testing is just the testing time.
Alex
My SNES was across the house from the computer... the Genesis was right next to the computer. Take a turn, play some Powerball! :)
You could do the same abuse with less elegant solutions than NAT. Simply running a simple Proxy server for your neighbors would provide them access. Only 1 machine is on the Internet, the rest aren't. Hell, if you are running MS's busted proxy, the rest don't even need TCP/IP, they could run IPX/SPX. (Lousy program, NEAT configuration options, I never want to go near it again...)...
Myself, I have a $90/month DSL connection. Why? If I need to get a VNC connection through the VPN to a work machine, I want the 384K uplink.
We have a NAT box with wireless, and technically, 4 computers there. I live with my fiancee. She web browses from her iBook, and I work from home on the weekends. We barely use the bandwidth.
However, I pay the premium so it is there when I need it.
Ban NAT and I lose Wireless. If that is the case, I drop DSL. I can't run Wires all over my apartment, so I use Wireless to send the signals around.
Find the abusers, by all means. However, leave those of us that don't abuse it alone.
Alex
Relax, take a deep breath. Stop calling people stupid or morons. This isn't personal. I understand that you're stuck not being in America, but that is no reason to be angry (that was a joke, BTW, bring on the flames).
.NET wins.
This Slashdot dream of taking over people's computer desktops everywhere is a silly one. I'm suggesting that there are people that will have a need for a KDE Desktop, and it doesn't matter that it isn't your mother.
You, however, choose to go ballistic. Relax, it isn't that important.
The desktop PC wars are over. They were fought between Microsoft and Apple, Apple lost.
We now have lots of processing power and lots to do with the machines. It doesn't all need to be general purpose. There is a role for KDE to play, even in a MS desktop dominated future.
OTOH, keeping MS to 95% of the market (or even, joy, rolling them back to 90%) would be huge. Don't let them own the web browser market and you are okay. Keep open protocols. If we are constantly reverse engineering their stuff, we lose,
Alex
MS owns the desktop PC market. Who cares, we're all moving on.
.NET programmers will need Win32 desktops, MS owns that.
The notion of a personal computer is from the past. People use their network interface device. Right now it is a PC, in the future, who knows.
Microsoft Homestation is their answer for consumer access. AOL will likely have their answer. Apple seems to be ignoring that market and focusing on people that have money and want digital toys. The new iMac with DVD burning, iDVD, iMovie, and iPhoto aims for this market.
The work environment? You need a system that supports something like Outlook and MS Office. Microsoft owns this market and will for the forseeable future.
What markets are in flux?
Web programmers. As we move for web servers, we need systems for people that program them.
However, Java programmers, PHP programmers, etc., will likely want to consider Unix desktops. Linux can fill a niche here. They still need e-mail, word processing, and printing.
Tablet PCs, open market. Linux based solutions can compete with Windows based solutions.
Television computing... who wants to fight the Homestation? Tivo, you going to step up with Linux? AOL, what are you going to use? These are the markets to fight in.
Sure, the KDE/GNOME desktop may not make it there... Microsoft's Explorer (the Win95 and up GUI) won't either. However, if it is Windows based, COM/DCOM/ActiveX/OLE will be used. If it is Linux/KDE/Qt based, then Kparts will be used.
Developers need a desktop to develop for the target system. If you are doing a Kparts/Qt/KDE/Linux set top box, what makes more sense, a KDE Desktop, an OS X Desktop, a GNOME Desktop, or a Windows desktop?
Welcome to the networked world. We can all pick our platform. The Microsoft monopoly will die... long live the Microsoft monopoly. Alternatives to Windows for the non-PC market is important to stopping the market.
Sure Microsoft will be a player, but they don't need to be the only one.
Alex
It sounds far-fetched, and yet there are so many very intelligent students who graduated at the same time as me whose only goal is to get out of the western market lifestyle at any cost because they feel that the nature of ideas has been fundamentally changed, from a kind of forum for the enrichment of man to a tightly-controlled, tightly-protected profit-making establishment serving only this new caste system.
And some of us didn't go to school in all of California, let alone Berkeley.
I don't know that I buy the arguement that more advertisement = less purchasing. I would assume by an increased attempt to advertise these items that people are registering them.
Alex
The engineer in me HATES x86. I understand why it was a good solution in the late 70s and early-mid 80s. It became an ugly hack in the early 90s, and is now a legacy solution.
Including support for "legacy" assembly code in Microsoft's first generation video game console is stupid.
The Gamecube has horrible software for the crucial 16-24 segment, but find outside of that. At 22, I'm unusual in LIKING Nintendo's offerings. We'll see what happens, because my peers were kids for the NES days, in college during the N64 days, and now out starting to earn a living. We'll see what happens, but I think that Gamecube has a shot.
I REALLY like Luigi's mansion. Pikmin is awesome. I don't like DOA3 or Halo.
I have Rogue Leader II, it's fun to show off my Gamecube in surround sound and 480p. However, I don't really care for the game.
I'm a 22 year old college graduate, and I like the game cube.
Alex
This Linuxhead you slammed is an MCSE. I kicked Linux out of my server room a while ago. We run OpenBSD for almost all our Unix needs (I run Linux database servers far from the Internet because of OpenBSD performance problems).
Look at the Xbox hardware. It is a Celeron processor with OTS components that are found in PCs. The Gamecube uses a custom processor that was based on IBM's desktop CPU, but with a custom GPU and unusual memory configurations.
The system, componentwise, is a PC, with a custom OS (using kernel code from Win2K, but a completely custom Ring-0 OS). The kicker to me is the inclusion of DirectX. DirectX means that applications for WinXbox are source compatible with Win32.
All the yelling in the world that this is a custom system built for gaming doesn't make it so. It is a PC with a good gaming sound card AND a special chip that converts PC resolutions -> HDTV resolutions (taking up to 1024x768 resolutions).
It burns cash for MS because the component costs are expensive. However, their R&D costs are really low. It's a tradeoff.
However, if you rev the system regularly, there is less time to recoup R&D costs. The Sony, Sega (Saturn and later), and MS strategy is to burn money early on and make money late in the console's life. A R&D light approach let's MS put everyone out of business.
I disagree with you on an assessment of computer hardware so I'm a moron?
I'm not a huge Halo fan. It looks good, but I don't like FPS games. However, I really enjoy Luigi's Mansion. It's a fun game, not particularly deep, but fun and silly.
However, I think that picking two arbitrary games (one system's flagship with 5 years of development, the other a technical demo that morphed into a short game) and comparing the systems is a little silly.
The hardware is different.
The GPUs are really hard to tell. I THINK that the Xbox GPU is stronger, but I don't know enough about the Gamecube's to know for sure.
The CPUs are night and day. The Xbox has all the legacy garbage of the Intel Celeron, compared to the sleak Gamecube CPU. The Gamecube CPU is a more intelligent design, better at floating point, and overall should kick its butt.
Xbox has more memory, point Xbox. Given the use of standard computer RAM, it should have 128MB, not 64MB. Stupid decision. As a result, this advantage become more minor than it should.
GCN has faster RAM. There is RAM everywhere, with highspeed interconnects.
This lends itself towards not having slowdowns. This should allow the GCN to sit closer to 100% utilization all the time.
The game development environment is more condusive to building games with depth? What the fuck are you talking about. The Nintendo style of not having FMV lends itself to games that you PLAY more.
The Xbox is a crappy PC thrown in a black box on the theory that games will be better because the hardware is fixed. However, if you lock all the hardware behind an abstraction layer, you don't really have much room to optimize.
The advantage to the Xbox is that you can release games NOW that require its hardware, while PC games normally have to target PCs 2-3 years old. Given that everything save the GPU is 2-3 years old in the Xbox, this isn't a REAL advantage.
Xbox has the hard drive and larger memory areas, this lends itself towards these deeper games that you discuss. However, having a GPU that shares the memory really knocks down that memory advantage.
The Xbox is a sloppy system, justified entirely on hype based upon numbers that aren't that good when you analyze them.
Two years and the Xbox is toast.
Microsoft's strategy may be based upon the fact that they can crank out PC-based systems with near zero R&D. This means that they can rev the hardware every 3 years. Shrink the lifespan of the consoles, and you cramp Sony and Nintendo's abilities to do R&D. Take all the money out of the industry and Sony pulls out, and Nintendo gets hurt. Then you can leave crap out there and sell games.
With Windows, MS needs to force you to WANT to upgrade. With an X-box monopoly, they just need you to WANT a new game, no need to put out new hardware or anything.
What a sad, sad, sad possible future.
The DC controller was horrible. I blocked it from my memory...
Let's START with the fact that the cord comes out of the controller in the wrong part... With a notch to run it back through?
This thins is huge and hideous... I feel like hitting the shoulder pads makes me lose balance.
How many fucking adapters does a controller need spaces for? How many adapaters does each adapater need?
The N64 was bad enough when the rumble pack came out, the dreamcast controller sucks ass, competitive with the Xbox controller.
Alex
I love my GBA. It isn't a substitute. It's awesome, has a monopoly on the handheld market, but it isn't in the same competition as the others. You don't see people deciding between a GBA and GCN, or a GBA and PS2. If you want a handheld, you get a GBA.
The games rock, its a money-maker for Nintendo, etc. However, it isn't competition for home systems.
I won't grant that the Gamecube controller is a rip-off of the Playstation's controller.
To me it is a natural evolution.
The Playstation's controller was a rip-off of the SNES controller that got extra shoulder pads and LATER analog sticks (after they ripped the idea off Nintendo).
The Gamecube controller is very clearly derivative of the N64's controller. The left hand options before (D-pad + L, or Analog + Z) have been merged into (D-pad OR Analog) + L with the added bonus of the D-pad and Analog both being usable with an easy switch.
The button layout on the right is the SNES layout reoriented around the reality of a primary button (A), secondary button (B) and optional extra buttons (X, Y).
The SNES had four equal buttons, but they were rarely uses as such.
The Z-button? It's a hack tacked on at the last minute for people worried about losing a button. There it is, developers, please don't use it much.
Analog shoulder pads, brilliant new invention (like N64's analog stick, SNES's shoulder pads, and NES's D-pad) that everyone will copy.
C-stick, it's neat. An adaption of the C-buttons into a stick. The C-buttons had the advantage of letting the N64 ACT like a 6-button controller (for things like Street Fighter).
Nintendo's N64 controller was large and unwieldy but REALLY flexible.
Games didn't use the flexibilty.
Gamecube keeps the controls and options and tweaks the layout to be more useful.
No it isn't 3-controllers in one (theoretically, N64 = Dpad + buttons, Dpad + analog, Analog + buttons), but 1 awesome controller with everything in a clsoe distance.
Yes the Playstation dual-shock is a nice controller (once you get past the shock of not grabbing the left side and having the primary controller there... drove me crazy on the N64 and hit me again now), but the GCN isn't a ripoff of Sony.
X-box has a custom SIMD by IBM. Motorola wrote Altivec for the G4. IBM wrote a custom SIMD for the G3. Same platform (PPC), similar idea (SIMD), years newer (IBM's implementation).
I used Altivec as an example. Gamecube has its own SIMD implementation taht was designed with the particular instructions Nintendo asked to be implemented for gameplay. It is probably better than Altivec is because it is newer and custom.
I'm anti-bullshit, and the Xbox is full of it.
The greatest GPU in the world? Who cares? You're playing at 640x480x60 at best, I don't need a Nvidia Geforce-12000. I need a gaming system that can do innovative a new things. The Gamecube GPU is pretty good, maybe even comparable to the Nvidia one in the Xbox.
We're all ASSUMING that the Xbox has the best GPU, as we are more familiar with Nvidia than the company that ATI bought. However, the Gamecube stuff apparently blows away ATI's in-house tech, and the ATI Radeon 8500 competes pretty well with the GeForce 3, so who the hell knows.
Gamecube's hardware may be much better than Xbox and we won't know.
However, it should be obvious to anyone taht has ANY background in microprocessor design (no, reading Tom's Hardware Guide or Anandtech is not an education, it's a benchmark shootout) that the Gamecube CPU will beat MS's clunky system.
You're only as fast as your weakest link, so all MS's other advantages are relatively trivial.
Alex