Do they get more detail at same/higher FPS? Probably yes.
Is this the result of some new groundbreaking design? No.
Is this EVEN LESS power efficient then the last series of non-power efficient cores? Yes.
You're trying to go along the lines of "this is the way it is and that's all there is to it." and I'm trying to say it isn't impressive, worthy enough to spend money on. What is their incentive to come up with a new GPU design if people keep buying the power hungry current designs?
I personally can't stand blogs simply because the vast majority of people lead superficial and annoyingly shallow lives.
"like totally tina said 'pshaw' and I was like no way and she was like 'uh yeah!' and I was totally like 'talk to the hand' and she was like about to burst a tear it was HI-lar-rious!"
I find some developer blogs interesting but that's only because I want to see what PRODUCTIVE shit they're up to [and occasionally there are tidbits of funny shit].
I'm not trying to say people shouldn't put their diaries online [er.... blog]. I'm saying big fucking deal.
Where's the newstory about 1990s geocities webpages? I had one! Am I famous? Do you like me yet? Same shit, different name, different decade.
Rather, people like you are. This mentality that UNIX/Linux is like some kind of priesthood that requires a decade of hard work and dedication before you are allowed to use it. This idea, that you either have to be a master, or you can just get the fuck out.
I never said that. I've been learning development skills since I was 12... but that was rather unstructured learning. I'm confident that with college and few projects to anyones name they'd be worthy colleague.
But that said... Why should it be "so easy"? Linux [and the GNU userland tools] are so powerful because they aren't "pushy clicky". They're not all hard to use [most are common things we don't think of like ls or cd] but can be used to do powerful things [e.g. filters for instance].
Anything worth accomplishing shouldn't be trivial.
Think about it. If being a developer or admin was something just anyone could do why would you be employed? And yes, it may be cool to have everyone be able to develop software they want but the REALITY of the situation is that developing software takes skill because there is quite a bit to manage correctly.
Computers are just tools, albeit powerful ones. Not everyone needs to be a grand-fucking-master with them to get use out of them.
I never had a problem with this either. What I have a problem with is the misleading subject lines. If it's an introductory book put the word intro in it. That way when you walk into work and you see "starter verilog for novices" on the shelf instead of "verilog in 21 steps" or whatever you'd question to yourself "is this the sort of employee we want as an EE?"
The dummies series at least are a bit more forward. But there is an entire world of what ammount to introductory texts out there in the guise of reference manuals and advanced aids. Finally the title is just stupid. "UNIX in 24 hours"... isn't "learning UNIX" just as easy and more informative about what the subject and theme of the book is?
And finally if the big put off for people to move to Linux is that they may have to learn something what does that say of our society as a whole? "Oh I can't do that, I might have to think!" yeah heaven forbid that. It's bad enough we have smoking, caffeine addicted yuppy cellphone owners behind the wheel of gas guzzling cars beeming down the interstate at 90mph. We'd hate to interrupt their oh so busy fucking lives with a bit of knowledge.
Knowledge is empowerment. If you can make the most out of your computer [e.g. no how to use OSS] you're less likely to shell out money for tools you don't need and only would buy because you don't know better [I've for instance, never once bought a C compiler or office suite...]. And if the price of not having to deal with the yuppy license schemes of piss-poor software from redmond is I have to learn how to "emerge openoffice-bin" then so be it.
But this is the price we pay as a society. We blindly walk into Wal-Marts because the price is "so low". We don't think about HOW or WHY the prices are the way they are. I mean afterall we're just too busy and important for that. We blindly consume a lot of shit food [in high quantities] because it's simple, affordable and available. We blindly buy stupid cars that are less efficient than 1970s muscle cars. We blindly lock ourselves into monopoly owned telecommunication equipment when more logical pricing plans are possible, etc.
We as a whole do a lot of things out of complete fucking ignorance and indifference because the fast lane is the right lane. It comes to a point where quality and efficiency suffer so greatly you run out of the ability to perform productive work. At the point where we have to dominate complete continents [re: europe and asia] to make sure we have enough oil to put in our SUVs... you'd think people may say "hey wait a minute, this isn't scalable!". You don't see 5 billion dollar air craft carriers defending your right to a fre
Pumping a lot of electricity into a circuit is ***NOT*** "cutting edge".
I'm sure my 3.2Ghz P4 could run at 4.2Ghz if I nitrogen cooled it and pumped 16 times the current into it.
Big deal. Show me where they get this performance at an EQUAL or LOWER amount of waste?
The SR-71 is not exactly a good example. They sacrifice payload and a lot of fuel to fly that high and fast. Compare that to a 757. Last time I did the rough calculation per person transported a 757 wasn't that much less efficient than your average SUV [all while travelling much faster and further]. That's progress. That's also 20 years ago.
Show me the SR-71 version 2.0 which uses less fuel or has a higher payload... etc...
A lot of "really cool things" are possible if we just decrease the efficiency of a device or process. You can make cars quicker if you accept a higher defect rate. You can get places quicker if you speed [and lower your mpg]. You can read books faster if you drop every other sentence, etc.
On a similar but slightly different note games like F.E.A.R. aren't really that much better because of the graphics. Just because it *can* use those cards doesn't mean you need to. And this gets back to the whole point, why are people impressed with tech demos, er, um, ah, games that require such heavy lifting just to look half decent?
Would you be equally impressed with my sloppy game engine? Oh but it's good I mean it requires a 2.8Ghz AMD64 to run!!! That must mean it's good!!!
I just don't find it very compelling that just because something requires a lot of power to run it must be somehow better than before. I've played far cry in both high and medium settings [my PCI-E 6600 can handle both just fine at 1024x768] and frankly I really didn't notice an improvement that really stood out. I mean the draw distance seemed maybe further but that's about it.
You're ability to base camp, spawn camp, stalk and other such griefing, er, winning game plans aren't affected by rendering detail. That's a load of shit. You'd do just as good with solid non-textured objects. All you need to do is see the shape and point. The fact that the texture is 512x512 instead of 64x64 won't help you do that.
I mean how much detail do you take away anyways at 80fps in a fast paced battle? When I play UT2k4 [with a shitload of bots] all I see are red and blue and I shoot. They could have "fuck you trebek" written on their fronts and I probably wouldn't notice.
People want the cards for the single fact that they have to have the latest. It's bragging rights. And it's something to brag about because people are ignorant of engineering principles and think all the wrong things are impressive. Also a lot of these kids don't pay the hydro bill [or for their computers] and it's nothing to waste 100s of Watts on a overblown GPU to "frag some chump".
Flash news: Current AMD Turions are, in fact, 64 bits (which *may* be why their name really is AMD Turion(TM) 64)
The 32-bit comment was about PPC not Turion....But yes I did word that wrong and it was misleading. Sorry.
The biggest flaw I see with most incarnations of PPC [like the G4] is they shoot themselves in the foot with bandwidth. 133Mhz FSB? What's this the mid 90s? Put a good ol' dual-channel DDR400 [or DDR-II-533] on the front of it and be done with.
I'm sure with a slightly longer pipeline [iirc the ALU in the G4 has 7 stages] and an integrated dual-channel memory controller you could kick out 32-bit PPCs that would give the rest a run for their money. I'm not suggesting this is trivial or cheap. Just suggesting it's a worthy project.
Keep in mind the PPC has some clear advantages over x86. The instruction set is RISC so there is no need for advanced decoders [e.g. to make macro-ops and the like], there are 32 GPRs and FPRs available making code run off registers a lot easier, etc.
Crank a G4 like processor up to 2Ghz or so and you'd see cool shit happen.:-)
Intel and Apple are like those kids in school who always got to go on ski trips with their year passes and high end skis, etc. They think they're all important because things are good. Then when they get into the real world they get eating alive.
Apple as far as I can see has a bit of a disconnect with "reality". They're not customer-oriented [e.g. shitty dead pixel policies, really costly vendor-locked in gear, etc]. It was always "oh that's apple gear only because the quality is higher" yet all the folk I talk to say "just get a wintel laptop and be done with". And it isn't like the PPC isn't a nice processor [the G4 actually looks fairly sweet]. It's just they're a bunch of pricks.
Apple went with Intel for appearance sake only. A Turion-based laptop would have done them just fine [or hell just invest and extend the PPC line]. I mean the G5 was a bit extreme, no need for that in a laptop [though it is cool]. A 32-bit laptop is mighty fine given that you don't normally run multi-GB database engines or whatever on it.
Is you confused or something? This is what theys call in America "modern capitalisms".
And besides you're supposed to use retarded circular and self-defeating logic like
"AMD doesn't have the big customers because they can't afford to upgrade their supply chain to have the big people as customers." -- Average retarded suit.
If you're calling me a degree-snob you're way off base there. First off I went to COMMUNITY COLLEGE. Second, I'm largely self-taught in all of my prefessional endeavours [though I did take a 3 yr comp.sci program to ensure I have some formalism behind what I'm doing and I have had experience with subjects I may not have taken a liking too on my own].
My point about bringing college into this is sometimes you have to be told what you don't know you don't know so you can later know what you don't know and then learn what you know you don't know.:-) I mean I know everything I think there is to know about space propulsion. Sounds impressive? Shouldn't. I don't know squat about the subject so what I think there is to know is pretty small.
Being thrown into lab after lab writing servers, embedded applications, compilers, etc, forces you to confront head on what you don't know [e.g. how the fuck do I solve this problem]. If you limit your home-study to things you already know about you'll go quite slow because you're not likely to bump into subjects you didn't know about.
The books aren't bad they're just misleading and the titles are annoying. Pocket references and the like are perfect for people who already have mastered computer science and need to pick up the quarks of a given tool. Not for people who don't get what a for loop is or how modular programming is supposed to look like...
At least then it's not some bullshit ploy to get you to buy it and display it proudly on your bookshelf.
I have no problem with beginner books [or self-study courses]. I'm just saying the whole notion that these books produce professionals is just insulting.
Because you know for a FACT that people reads these books and boom apply for every job in sight under the guise of experience and talent. And sure they may get their kick in the arse but not until a year or two down the road after you have been passed up for the same job or worse, bought a product they produced!!!
It's called a discussion. I'm raising the question of whether we should encourage this trend. Suggesting perhaps that a better course of action would be looking at new sources of performance.
I'm not saying I'm right or everyone should bow to my wisdom. I'm just trying to act anti-sheep and suggest we as customers ought to demand more than power-hungry graphics cards [or processors for that matter].
I upgraded to the 6600 because my AMDx2 motherboard is PCI-E. So is my 915G [and 945] I use for my Intel gear.
Trust me, I wasn't happy to have to buy the new card. My choice was an 6200, 6600 or not buy the dual-cores. The 6200 is just a piece of shit. I mean I'm not into excess but I DO like some ability to render a 3d scene;-)
I find all these "$THING for $PEOPLE" and "$THING in $QUANTITY of time" books insulting.
Sure you can learn how to type "cd/usr/bin ; sudo rm -rf *" in 20 minutes. Can you learn to develop and debug shell scripts in 24 hours? I think not.
Nor do I think people can learn C or C++ or Java in 24 hours. It's just insulting. Now I know they don't literally mean one day, but even college classes run longer than 24 hours. In college you'll have a 50-60 hour class on "intro to C" followed by FIVE MORE SEMESTERS of classes that build on it.
I hate these books because they're retarded. I learned C primarily from "type and learn C" [I think by Sams] when I was 12. Then I proceeded to actually write programs [lots of them, 1000s of them]. I learned by doing and it took a long time. I wasn't half-way decent at "coding" until I was 19 and I'm just getting solid at proper development [well I'd say the last year has been really smooth].
For all of us who do take it serious and have been through a lot of training I find these books insulting. And no, it isn't because I sunk a boatload of cash into the courses like a MCSE. I think people are quite capable of teaching themselves how to use UNIX shells or C programming. I just don't think it's the sort of thing you can do over a weekend or two.
So fuck off already with the books that serve no purpose but to flood the market with a lot of "smart" people who turn out to be useless as the day is long.
Actually I have my own "nifty" computing devices [among other things AMDx2 and Pentium D desktops]. But I buy them because I actually do productive work on them [ok the pentium was solely for benchmarking...]. When building a project that has 100K lines in it [from scratch when doing customer drop tests] it's nice to know I can do 30 builds an hour instead of 5. [to fine tune scripts and such].
Things like these cards make no sense. My FX5200 was capable of playing games like UT2k4 just fine. My 6600 is more than enough for things like Far Cry. I'm not saying people shouldn't be allowed to get faster cards. I'm asking why are you impressed? Engineering is about making the most from the least.
Getting 2x the FPS at a cost of 3x the power isn't really progress now is it?
People clamour over the Intel EE processors [apparently given the reviews] because of the super sized cache and "biggie sized" power requirements all while a lower powered AMD64 [or heck even a P4] is often just as good or if not sufficiently good.
What does it matter once your card can get over 60fps at 1024x768? You're not going to see any more detail just because your HUD is smaller and above 60fps and you're a laughing joke anyways [hint: not a lot of monitors can do 1600x1200 at 100Hz...] so really a lot of the frames go to waste and the detail is irrelevent.
It'd be like having a dual-core processor where the 2nd core mirrors the first core only [e.g. it doesn't do anything other useful]. Yeah it's totally cool you have two cores on die but you're not extracting any useful benefit from it.
As for your car reference, the same line applies there. How many people do drive SUVs and minivans when a 5L/100Km car would serve 99% of their tasks properly? It's cheaper for most people to just rent a van when you need it and save the rest of the year on gas.
Sure you have a right to be wasteful but is that really something to be proud of?
Compare a 2MB L2 cache on a P4 to a box with 1MB of L2 or 512K... the cache ends up contributing less and less to the overall performance.
What is also important is associativity. If you have a low-assoc cache, meaning a given address has few places in the cache it could reside you end up wasting more space. That's why [iirc] the AMD processors have high associvity L2 caches. They make good use of the 512K available.
At my previous job we built Gentoo distros on 128 and 256K semprons and the time to build wasn't really that different even though we were building 100s of packages.
So you could have a relatively high performance web server or file store [or whatever] without the 2MB of cache stuck on the back of the thing.
One phrase: 8 cores + with 16/8 K cache per core and **one** L2 cache.
I'm sorry but I'd like to think that part of the SMP boost AMDx2 delivers so smoothly is that each core has a fair chunk of cache [64/64 + 512 or 1024] with a dedicated HT bus.
So is this an 8-port L2 ? What is the latency on it when all 8 cores are busy? etc... I think we'll find this core will suffer greatly from this point.
I mean look at the Pentium D. It's only saving grace is the 1MB of L2 each core has. The interlink bus is just the FSB and it ain't none too fast.
Sure why not. Someone has to keep these mindless idiots in check.
The problem is thinking like yours means you'll never have "the right card". Once you save up your salary from Walmart to buy this 7800 the "7800 X-treme e-gamer E-edition" will be out and you'll be like "shit, damn and I just bought this 350W graphics card for nothing!"
And frankly I don't know why you guys are impressed. My graphics card spends 99.9999% of the time in 2D mode [rendering my glorious X11/Gnome desktop]. I do play games but really not a big part of my daily life.
Why would I spend the money on the card and power just so once in a blue moon when I play games the crapply modelled M16 my character is holding will look "ultra detailed and super rendered". All I care about FPS type games [where most of this bullshit technology goes] is a decent framerate and the ability to totally own a bot or two.
If I have to use "medium" detail to get that then so be it. If the game visually is crap I'll not buy the sequel, etc...
Dude, a PPC 440 is 6mm^2 and consumes 700mW of power at 667Mhz. You could easily fit a dozen on a die the size of a P4 and still take FAR less than this Niagra core. At a rated 1200MIPS per core a die with eight of them would net you close to 9600MIPS max. Of course you'd need some form of L2 cache and high speed internal bus.
Similar the new ARM cores Cortex it takes roughly the same power at 1Ghz which gives it apparently 2000MIPS. The area is about the same as PPC 440. So in theory you could hook 4-8 of these up as well and get a killer chip too..
Point is Suns quotes of being "2 possibly 3 generations ahead" is totally bullshit. They're at most one generation ahead. It takes one multi-core ARM or PPC to totally destroy this.
For the same reason I can't find my ARM or PPC powered laptop you won't see things like HD really become that important.
Though generally from what I've been reading here [and seen firsthand] HD varies in quality between vendors and for the most part you won't notice anyways. A gaping plot hole in a B-rated movie is a gaping plot hole in a B-rated movie regardless of the number of pixels on the screen.
You mean I can save money on my electricty bill AND my hardware bill by just buying an incredibly out dated machine???
How is it progress to spend more energy then you have to, to accomplish a goal?
I mean, what next? Are 3mpg cars better for commuting than 40mpg cars? [assuming relatively equal levels of safety].
You seem to think everyone plays games, does software builds and transcodes movies. Would you be surprised to learn that most desktops are being used for data entry? Email, web development, word processing, marketting, etc. Fuck, even some software development [e.g. when you're code works against hardware not inside your box] doesn't even need a lot of power.
So what you should say is
"What? I've been mislead into thinking I need a power-sucking processor to type text?"
You're just buying into the hype and hysteria of "phong-shaded goodyness" that is todays gpu market. If you think PS2 capability to display graphics takes away from the ability for a game to be fun... you live a sad life. You'll ALWAYS be disappointed. Though I don't even look at it as a negative. PS2 graphics are realistic enough for me to get into the game. It's the gameplay that makes me enjoy the actual game.
And for the record, PPC 440 processors are newer than you think. [so are some of the cool ARM processors out there].
You think because something takes less power [re: gas] it's inferior. That's because you're an idiot and you don't know what you are talking about.
I'm sitting here at a 400Mhz P4 processor typing this text. what does that tell you? [I'm running gaim, xmms playing an mp3 and Gnome at the same time]. My 3200Mhz P4 actually spends most of it's time [I use the computer about 8 hours a day] at less than a Ghz. Even when I'm busy debugging something or coding it ramps up to build and down again to idle after.
If I can type this message in slashdot [well in firefox] and play an mp3 at the same time and my cpu is running AS SLOW AS HUMANLY POSSIBLE [and it's a P4 which isn't exactly an efficient core] do you really think you need faster processors?
For that matter, even with the P4 at 400Mhz the PPC 440 running at 667Mhz would take less power and still have more MIPS to deliver.
Best place to get a PPC would be in an embedded device... unfotunately not a lot of people [re: any] make desktop computers based on the PPC.
It's one of those chicken an egg problems. It would likely cost a good $100,000 or so [if you paid yourself nothing] to develop a desktop kit based on the PPC on your own. Who has that money? And since the mass public are so blinded by mindless advertisement they're not seeking alternatives.
"will this PC run my win3.1 programs?" etc...
I got my PPC from http://www.projectblackdog.com/ but it's not a desktop. It's just a USB powered device you can ssh into. I use it to test and benchmark my crypto libraries [which are getting faster by the day:-)].
I think this sort of thing would just need a petition of sorts. E.g. get a good thousands or so people to petition IBM to design a desktop mobo with a PPC 440 in it.
I'd say reasonable specs would be
- 667 Mhz PPC 440 - 256MB of ram [if at all possible make it use DIMM or SODIMM DDR memory] - PCI bus with a couple slots - onboard IDE [ATA133], Ethernet [10/100], Sound [AC'97 compliant], VGA [SiS?], PS/2 and USB 2.0 ports - mini-atx form factor
I'd suspect such a motherboard would competently sell for around $200 USD. Add a $50 case, $80 HD, $30 DVD/CDRW and you have a desktop box for $360. It'd be dead silent [except for the HD], really small and take little power.
lack of customer desire? I know a lot of gamers, I've yet to meet one with an HD setup [nor one bragging about their pending move to HD]
If adding HD support adds 30$ to the console and maybe 1% of the customers want it... guess what. It also makes the games cost more because you have to test multiple resolutions [though I'm sure they'll keep that in mind...hehehehe].
But you'd be pretty foolish to believe that the best they can do with their massive budget is to buy a bunch of white boxes.
Why? What proof do you have that they actually have custom hardware?
I have no doubt about them owning a good cluster or two. I just don't think it's anything that special. Their linear increases in computing power mean nothing against adding a bit or two to a key.
Most likely the non-trivial leaks they get are from the human factor and not them cracking 128-bit keys. I have no doubt though they get lucky often. Just because you want to blow up a building doesn't mean you know not to use "password" as your password:-)
I disagree. The Athlon design [overall] has been fairly consistent. It's implementation has varied greatly since the first 500Mhz Slot A Athlons were introduced.
What is the Athlon? It's a 3-pipe ALU/AGU with the multiplier on pipe 0, 64KB of L1 Code, 64KB of L1 Data and a L2 cache, 3 pipes for FPU [add,mult,load/store]. Engine has directpath/vectorpath instruction sets where the cores use macro-rom for vector ops. They can decode upto three opcodes at once to feed down one of the 9 pipes. The engine is out of order and it can speculatively execute instructions [as well as register rename].
This hasn't changed. EVEN IN THE NEWER AMD64 CORES!!!
What has changed
1. Size of the L2 cache [implementation detail] 2. Length of the ALU and FPU pipelines [longer in later cores] 3. Instructions [Added SSE, SSE2 then SSE3, x86_64] 4. Introduction of DDR controller on board [implementation detail] 5. Newer transistors, process and package [implementation detail]
Athlon is much as a "cpu design" as it is a brand name. Just like the K6 describes a particular core [so does P6 on the intel side].
Slashdot Mirror
Improvements are not always obvious.
Are the new GPUs faster? Yes.
Do they get more detail at same/higher FPS? Probably yes.
Is this the result of some new groundbreaking design? No.
Is this EVEN LESS power efficient then the last series of non-power efficient cores? Yes.
You're trying to go along the lines of "this is the way it is and that's all there is to it." and I'm trying to say it isn't impressive, worthy enough to spend money on. What is their incentive to come up with a new GPU design if people keep buying the power hungry current designs?
Hint: Intel anyone?
Tom
Why not... do this interesting thing called VISIT YOUR FAMILY.
Want to know what your family members are like? Get in a car, plane, boat, whatever and spend the week.
You don't need yuppy-thoughts on the web to find out what your family is like.
Tom
Blog == diary put on the web.
I personally can't stand blogs simply because the vast majority of people lead superficial and annoyingly shallow lives.
"like totally tina said 'pshaw' and I was like no way and she was like 'uh yeah!' and I was totally like 'talk to the hand' and she was like about to burst a tear it was HI-lar-rious!"
I find some developer blogs interesting but that's only because I want to see what PRODUCTIVE shit they're up to [and occasionally there are tidbits of funny shit].
I'm not trying to say people shouldn't put their diaries online [er.... blog]. I'm saying big fucking deal.
Where's the newstory about 1990s geocities webpages? I had one! Am I famous? Do you like me yet? Same shit, different name, different decade.
Didn't care about people then, still don't now.
Tom
Rather, people like you are. This mentality that UNIX/Linux is like some kind of priesthood that requires a decade of hard work and dedication before you are allowed to use it. This idea, that you either have to be a master, or you can just get the fuck out.
... but that was rather unstructured learning. I'm confident that with college and few projects to anyones name they'd be worthy colleague.
... isn't "learning UNIX" just as easy and more informative about what the subject and theme of the book is?
... you'd think people may say "hey wait a minute, this isn't scalable!". You don't see 5 billion dollar air craft carriers defending your right to a fre
I never said that. I've been learning development skills since I was 12
But that said... Why should it be "so easy"? Linux [and the GNU userland tools] are so powerful because they aren't "pushy clicky". They're not all hard to use [most are common things we don't think of like ls or cd] but can be used to do powerful things [e.g. filters for instance].
Anything worth accomplishing shouldn't be trivial.
Think about it. If being a developer or admin was something just anyone could do why would you be employed? And yes, it may be cool to have everyone be able to develop software they want but the REALITY of the situation is that developing software takes skill because there is quite a bit to manage correctly.
Computers are just tools, albeit powerful ones. Not everyone needs to be a grand-fucking-master with them to get use out of them.
I never had a problem with this either. What I have a problem with is the misleading subject lines. If it's an introductory book put the word intro in it. That way when you walk into work and you see "starter verilog for novices" on the shelf instead of "verilog in 21 steps" or whatever you'd question to yourself "is this the sort of employee we want as an EE?"
The dummies series at least are a bit more forward. But there is an entire world of what ammount to introductory texts out there in the guise of reference manuals and advanced aids. Finally the title is just stupid. "UNIX in 24 hours"
And finally if the big put off for people to move to Linux is that they may have to learn something what does that say of our society as a whole? "Oh I can't do that, I might have to think!" yeah heaven forbid that. It's bad enough we have smoking, caffeine addicted yuppy cellphone owners behind the wheel of gas guzzling cars beeming down the interstate at 90mph. We'd hate to interrupt their oh so busy fucking lives with a bit of knowledge.
Knowledge is empowerment. If you can make the most out of your computer [e.g. no how to use OSS] you're less likely to shell out money for tools you don't need and only would buy because you don't know better [I've for instance, never once bought a C compiler or office suite...]. And if the price of not having to deal with the yuppy license schemes of piss-poor software from redmond is I have to learn how to "emerge openoffice-bin" then so be it.
But this is the price we pay as a society. We blindly walk into Wal-Marts because the price is "so low". We don't think about HOW or WHY the prices are the way they are. I mean afterall we're just too busy and important for that. We blindly consume a lot of shit food [in high quantities] because it's simple, affordable and available. We blindly buy stupid cars that are less efficient than 1970s muscle cars. We blindly lock ourselves into monopoly owned telecommunication equipment when more logical pricing plans are possible, etc.
We as a whole do a lot of things out of complete fucking ignorance and indifference because the fast lane is the right lane. It comes to a point where quality and efficiency suffer so greatly you run out of the ability to perform productive work. At the point where we have to dominate complete continents [re: europe and asia] to make sure we have enough oil to put in our SUVs
Pumping a lot of electricity into a circuit is ***NOT*** "cutting edge".
... etc...
I'm sure my 3.2Ghz P4 could run at 4.2Ghz if I nitrogen cooled it and pumped 16 times the current into it.
Big deal. Show me where they get this performance at an EQUAL or LOWER amount of waste?
The SR-71 is not exactly a good example. They sacrifice payload and a lot of fuel to fly that high and fast. Compare that to a 757. Last time I did the rough calculation per person transported a 757 wasn't that much less efficient than your average SUV [all while travelling much faster and further]. That's progress. That's also 20 years ago.
Show me the SR-71 version 2.0 which uses less fuel or has a higher payload
A lot of "really cool things" are possible if we just decrease the efficiency of a device or process. You can make cars quicker if you accept a higher defect rate. You can get places quicker if you speed [and lower your mpg]. You can read books faster if you drop every other sentence, etc.
On a similar but slightly different note games like F.E.A.R. aren't really that much better because of the graphics. Just because it *can* use those cards doesn't mean you need to. And this gets back to the whole point, why are people impressed with tech demos, er, um, ah, games that require such heavy lifting just to look half decent?
Would you be equally impressed with my sloppy game engine? Oh but it's good I mean it requires a 2.8Ghz AMD64 to run!!! That must mean it's good!!!
I just don't find it very compelling that just because something requires a lot of power to run it must be somehow better than before. I've played far cry in both high and medium settings [my PCI-E 6600 can handle both just fine at 1024x768] and frankly I really didn't notice an improvement that really stood out. I mean the draw distance seemed maybe further but that's about it.
You're ability to base camp, spawn camp, stalk and other such griefing, er, winning game plans aren't affected by rendering detail. That's a load of shit. You'd do just as good with solid non-textured objects. All you need to do is see the shape and point. The fact that the texture is 512x512 instead of 64x64 won't help you do that.
I mean how much detail do you take away anyways at 80fps in a fast paced battle? When I play UT2k4 [with a shitload of bots] all I see are red and blue and I shoot. They could have "fuck you trebek" written on their fronts and I probably wouldn't notice.
People want the cards for the single fact that they have to have the latest. It's bragging rights. And it's something to brag about because people are ignorant of engineering principles and think all the wrong things are impressive. Also a lot of these kids don't pay the hydro bill [or for their computers] and it's nothing to waste 100s of Watts on a overblown GPU to "frag some chump".
Tom
Flash news: Current AMD Turions are, in fact, 64 bits (which *may* be why their name really is AMD Turion(TM) 64)
:-)
The 32-bit comment was about PPC not Turion....But yes I did word that wrong and it was misleading. Sorry.
The biggest flaw I see with most incarnations of PPC [like the G4] is they shoot themselves in the foot with bandwidth. 133Mhz FSB? What's this the mid 90s? Put a good ol' dual-channel DDR400 [or DDR-II-533] on the front of it and be done with.
I'm sure with a slightly longer pipeline [iirc the ALU in the G4 has 7 stages] and an integrated dual-channel memory controller you could kick out 32-bit PPCs that would give the rest a run for their money. I'm not suggesting this is trivial or cheap. Just suggesting it's a worthy project.
Keep in mind the PPC has some clear advantages over x86. The instruction set is RISC so there is no need for advanced decoders [e.g. to make macro-ops and the like], there are 32 GPRs and FPRs available making code run off registers a lot easier, etc.
Crank a G4 like processor up to 2Ghz or so and you'd see cool shit happen.
Tom
Intel and Apple are like those kids in school who always got to go on ski trips with their year passes and high end skis, etc. They think they're all important because things are good. Then when they get into the real world they get eating alive.
Apple as far as I can see has a bit of a disconnect with "reality". They're not customer-oriented [e.g. shitty dead pixel policies, really costly vendor-locked in gear, etc]. It was always "oh that's apple gear only because the quality is higher" yet all the folk I talk to say "just get a wintel laptop and be done with". And it isn't like the PPC isn't a nice processor [the G4 actually looks fairly sweet]. It's just they're a bunch of pricks.
Apple went with Intel for appearance sake only. A Turion-based laptop would have done them just fine [or hell just invest and extend the PPC line]. I mean the G5 was a bit extreme, no need for that in a laptop [though it is cool]. A 32-bit laptop is mighty fine given that you don't normally run multi-GB database engines or whatever on it.
Tom
Is you confused or something? This is what theys call in America "modern capitalisms".
And besides you're supposed to use retarded circular and self-defeating logic like
"AMD doesn't have the big customers because they can't afford to upgrade their supply chain to have the big people as customers." -- Average retarded suit.
Tom
If you're calling me a degree-snob you're way off base there. First off I went to COMMUNITY COLLEGE. Second, I'm largely self-taught in all of my prefessional endeavours [though I did take a 3 yr comp.sci program to ensure I have some formalism behind what I'm doing and I have had experience with subjects I may not have taken a liking too on my own].
:-) I mean I know everything I think there is to know about space propulsion. Sounds impressive? Shouldn't. I don't know squat about the subject so what I think there is to know is pretty small.
My point about bringing college into this is sometimes you have to be told what you don't know you don't know so you can later know what you don't know and then learn what you know you don't know.
Being thrown into lab after lab writing servers, embedded applications, compilers, etc, forces you to confront head on what you don't know [e.g. how the fuck do I solve this problem]. If you limit your home-study to things you already know about you'll go quite slow because you're not likely to bump into subjects you didn't know about.
The books aren't bad they're just misleading and the titles are annoying. Pocket references and the like are perfect for people who already have mastered computer science and need to pick up the quarks of a given tool. Not for people who don't get what a for loop is or how modular programming is supposed to look like...
Tom
So why don't the call it that?
"Intro to Unix in 24 hours".
At least then it's not some bullshit ploy to get you to buy it and display it proudly on your bookshelf.
I have no problem with beginner books [or self-study courses]. I'm just saying the whole notion that these books produce professionals is just insulting.
Because you know for a FACT that people reads these books and boom apply for every job in sight under the guise of experience and talent. And sure they may get their kick in the arse but not until a year or two down the road after you have been passed up for the same job or worse, bought a product they produced!!!
Tom
It's called a discussion. I'm raising the question of whether we should encourage this trend. Suggesting perhaps that a better course of action would be looking at new sources of performance.
I'm not saying I'm right or everyone should bow to my wisdom. I'm just trying to act anti-sheep and suggest we as customers ought to demand more than power-hungry graphics cards [or processors for that matter].
Tom
I upgraded to the 6600 because my AMDx2 motherboard is PCI-E. So is my 915G [and 945] I use for my Intel gear.
;-)
Trust me, I wasn't happy to have to buy the new card. My choice was an 6200, 6600 or not buy the dual-cores. The 6200 is just a piece of shit. I mean I'm not into excess but I DO like some ability to render a 3d scene
Tom
I find all these "$THING for $PEOPLE" and "$THING in $QUANTITY of time" books insulting.
/usr/bin ; sudo rm -rf *" in 20 minutes. Can you learn to develop and debug shell scripts in 24 hours? I think not.
Sure you can learn how to type "cd
Nor do I think people can learn C or C++ or Java in 24 hours. It's just insulting. Now I know they don't literally mean one day, but even college classes run longer than 24 hours. In college you'll have a 50-60 hour class on "intro to C" followed by FIVE MORE SEMESTERS of classes that build on it.
I hate these books because they're retarded. I learned C primarily from "type and learn C" [I think by Sams] when I was 12. Then I proceeded to actually write programs [lots of them, 1000s of them]. I learned by doing and it took a long time. I wasn't half-way decent at "coding" until I was 19 and I'm just getting solid at proper development [well I'd say the last year has been really smooth].
For all of us who do take it serious and have been through a lot of training I find these books insulting. And no, it isn't because I sunk a boatload of cash into the courses like a MCSE. I think people are quite capable of teaching themselves how to use UNIX shells or C programming. I just don't think it's the sort of thing you can do over a weekend or two.
So fuck off already with the books that serve no purpose but to flood the market with a lot of "smart" people who turn out to be useless as the day is long.
Tom
Actually I have my own "nifty" computing devices [among other things AMDx2 and Pentium D desktops]. But I buy them because I actually do productive work on them [ok the pentium was solely for benchmarking ...]. When building a project that has 100K lines in it [from scratch when doing customer drop tests] it's nice to know I can do 30 builds an hour instead of 5. [to fine tune scripts and such].
Things like these cards make no sense. My FX5200 was capable of playing games like UT2k4 just fine. My 6600 is more than enough for things like Far Cry. I'm not saying people shouldn't be allowed to get faster cards. I'm asking why are you impressed? Engineering is about making the most from the least.
Getting 2x the FPS at a cost of 3x the power isn't really progress now is it?
People clamour over the Intel EE processors [apparently given the reviews] because of the super sized cache and "biggie sized" power requirements all while a lower powered AMD64 [or heck even a P4] is often just as good or if not sufficiently good.
What does it matter once your card can get over 60fps at 1024x768? You're not going to see any more detail just because your HUD is smaller and above 60fps and you're a laughing joke anyways [hint: not a lot of monitors can do 1600x1200 at 100Hz...] so really a lot of the frames go to waste and the detail is irrelevent.
It'd be like having a dual-core processor where the 2nd core mirrors the first core only [e.g. it doesn't do anything other useful]. Yeah it's totally cool you have two cores on die but you're not extracting any useful benefit from it.
As for your car reference, the same line applies there. How many people do drive SUVs and minivans when a 5L/100Km car would serve 99% of their tasks properly? It's cheaper for most people to just rent a van when you need it and save the rest of the year on gas.
Sure you have a right to be wasteful but is that really something to be proud of?
Tom
Law of diminishing returns?
... the cache ends up contributing less and less to the overall performance.
Compare a 2MB L2 cache on a P4 to a box with 1MB of L2 or 512K
What is also important is associativity. If you have a low-assoc cache, meaning a given address has few places in the cache it could reside you end up wasting more space. That's why [iirc] the AMD processors have high associvity L2 caches. They make good use of the 512K available.
At my previous job we built Gentoo distros on 128 and 256K semprons and the time to build wasn't really that different even though we were building 100s of packages.
So you could have a relatively high performance web server or file store [or whatever] without the 2MB of cache stuck on the back of the thing.
Tom
One phrase: 8 cores + with 16/8 K cache per core and **one** L2 cache.
I'm sorry but I'd like to think that part of the SMP boost AMDx2 delivers so smoothly is that each core has a fair chunk of cache [64/64 + 512 or 1024] with a dedicated HT bus.
So is this an 8-port L2 ? What is the latency on it when all 8 cores are busy? etc... I think we'll find this core will suffer greatly from this point.
I mean look at the Pentium D. It's only saving grace is the 1MB of L2 each core has. The interlink bus is just the FSB and it ain't none too fast.
Tom
Sure why not. Someone has to keep these mindless idiots in check.
The problem is thinking like yours means you'll never have "the right card". Once you save up your salary from Walmart to buy this 7800 the "7800 X-treme e-gamer E-edition" will be out and you'll be like "shit, damn and I just bought this 350W graphics card for nothing!"
And frankly I don't know why you guys are impressed. My graphics card spends 99.9999% of the time in 2D mode [rendering my glorious X11/Gnome desktop]. I do play games but really not a big part of my daily life.
Why would I spend the money on the card and power just so once in a blue moon when I play games the crapply modelled M16 my character is holding will look "ultra detailed and super rendered". All I care about FPS type games [where most of this bullshit technology goes] is a decent framerate and the ability to totally own a bot or two.
If I have to use "medium" detail to get that then so be it. If the game visually is crap I'll not buy the sequel, etc...
Tom
Dude, a PPC 440 is 6mm^2 and consumes 700mW of power at 667Mhz. You could easily fit a dozen on a die the size of a P4 and still take FAR less than this Niagra core. At a rated 1200MIPS per core a die with eight of them would net you close to 9600MIPS max. Of course you'd need some form of L2 cache and high speed internal bus.
Similar the new ARM cores Cortex it takes roughly the same power at 1Ghz which gives it apparently 2000MIPS. The area is about the same as PPC 440. So in theory you could hook 4-8 of these up as well and get a killer chip too..
Point is Suns quotes of being "2 possibly 3 generations ahead" is totally bullshit. They're at most one generation ahead. It takes one multi-core ARM or PPC to totally destroy this.
Tom
For the same reason I can't find my ARM or PPC powered laptop you won't see things like HD really become that important.
Though generally from what I've been reading here [and seen firsthand] HD varies in quality between vendors and for the most part you won't notice anyways. A gaping plot hole in a B-rated movie is a gaping plot hole in a B-rated movie regardless of the number of pixels on the screen.
Tom
You mean I can save money on my electricty bill AND my hardware bill by just buying an incredibly out dated machine???
... you live a sad life. You'll ALWAYS be disappointed. Though I don't even look at it as a negative. PS2 graphics are realistic enough for me to get into the game. It's the gameplay that makes me enjoy the actual game.
How is it progress to spend more energy then you have to, to accomplish a goal?
I mean, what next? Are 3mpg cars better for commuting than 40mpg cars? [assuming relatively equal levels of safety].
You seem to think everyone plays games, does software builds and transcodes movies. Would you be surprised to learn that most desktops are being used for data entry? Email, web development, word processing, marketting, etc. Fuck, even some software development [e.g. when you're code works against hardware not inside your box] doesn't even need a lot of power.
So what you should say is
"What? I've been mislead into thinking I need a power-sucking processor to type text?"
You're just buying into the hype and hysteria of "phong-shaded goodyness" that is todays gpu market. If you think PS2 capability to display graphics takes away from the ability for a game to be fun
And for the record, PPC 440 processors are newer than you think. [so are some of the cool ARM processors out there].
You think because something takes less power [re: gas] it's inferior. That's because you're an idiot and you don't know what you are talking about.
I'm sitting here at a 400Mhz P4 processor typing this text. what does that tell you? [I'm running gaim, xmms playing an mp3 and Gnome at the same time]. My 3200Mhz P4 actually spends most of it's time [I use the computer about 8 hours a day] at less than a Ghz. Even when I'm busy debugging something or coding it ramps up to build and down again to idle after.
If I can type this message in slashdot [well in firefox] and play an mp3 at the same time and my cpu is running AS SLOW AS HUMANLY POSSIBLE [and it's a P4 which isn't exactly an efficient core] do you really think you need faster processors?
For that matter, even with the P4 at 400Mhz the PPC 440 running at 667Mhz would take less power and still have more MIPS to deliver.
Tom
Tom
Go to Crypto'06. See the people taking notes. They're DoD.
Tom
Best place to get a PPC would be in an embedded device ... unfotunately not a lot of people [re: any] make desktop computers based on the PPC.
:-)].
It's one of those chicken an egg problems. It would likely cost a good $100,000 or so [if you paid yourself nothing] to develop a desktop kit based on the PPC on your own. Who has that money? And since the mass public are so blinded by mindless advertisement they're not seeking alternatives.
"will this PC run my win3.1 programs?" etc...
I got my PPC from http://www.projectblackdog.com/ but it's not a desktop. It's just a USB powered device you can ssh into. I use it to test and benchmark my crypto libraries [which are getting faster by the day
I think this sort of thing would just need a petition of sorts. E.g. get a good thousands or so people to petition IBM to design a desktop mobo with a PPC 440 in it.
I'd say reasonable specs would be
- 667 Mhz PPC 440
- 256MB of ram [if at all possible make it use DIMM or SODIMM DDR memory]
- PCI bus with a couple slots
- onboard IDE [ATA133], Ethernet [10/100], Sound [AC'97 compliant], VGA [SiS?], PS/2 and USB 2.0 ports
- mini-atx form factor
I'd suspect such a motherboard would competently sell for around $200 USD. Add a $50 case, $80 HD, $30 DVD/CDRW and you have a desktop box for $360. It'd be dead silent [except for the HD], really small and take little power.
Make it Linux compatible and voila.
Tom
lack of customer desire? I know a lot of gamers, I've yet to meet one with an HD setup [nor one bragging about their pending move to HD]
...hehehehe].
If adding HD support adds 30$ to the console and maybe 1% of the customers want it... guess what. It also makes the games cost more because you have to test multiple resolutions [though I'm sure they'll keep that in mind
Tom
But you'd be pretty foolish to believe that the best they can do with their massive budget is to buy a bunch of white boxes.
:-)
Why? What proof do you have that they actually have custom hardware?
I have no doubt about them owning a good cluster or two. I just don't think it's anything that special. Their linear increases in computing power mean nothing against adding a bit or two to a key.
Most likely the non-trivial leaks they get are from the human factor and not them cracking 128-bit keys. I have no doubt though they get lucky often. Just because you want to blow up a building doesn't mean you know not to use "password" as your password
Tom
I disagree. The Athlon design [overall] has been fairly consistent. It's implementation has varied greatly since the first 500Mhz Slot A Athlons were introduced.
What is the Athlon? It's a 3-pipe ALU/AGU with the multiplier on pipe 0, 64KB of L1 Code, 64KB of L1 Data and a L2 cache, 3 pipes for FPU [add,mult,load/store]. Engine has directpath/vectorpath instruction sets where the cores use macro-rom for vector ops. They can decode upto three opcodes at once to feed down one of the 9 pipes. The engine is out of order and it can speculatively execute instructions [as well as register rename].
This hasn't changed. EVEN IN THE NEWER AMD64 CORES!!!
What has changed
1. Size of the L2 cache [implementation detail]
2. Length of the ALU and FPU pipelines [longer in later cores]
3. Instructions [Added SSE, SSE2 then SSE3, x86_64]
4. Introduction of DDR controller on board [implementation detail]
5. Newer transistors, process and package [implementation detail]
Athlon is much as a "cpu design" as it is a brand name. Just like the K6 describes a particular core [so does P6 on the intel side].
Tom