What is Digg?Sponsored by Guitar Center
When Heavy-handed Hints Fail--Guitar Center Wishlist Posters view!
guitarcenter.com - Cool personalized designs send a friendly reminder that a new Stratocaster would rock, thanks.
182 Comments
- reckage, on 04/14/2009, -4/+93I miss the days when computers had turbo buttons.
- anexanhume, on 04/14/2009, -13/+83OM NOP NOP NOP
- tian2992, on 04/14/2009, -4/+60Awesome, yeah sure... an architecture designed for 16 bits resized several times.
I feel sorry that X86 won, PowerPC or Alpha were significantly better but both are sadly dead to the mainstream. - zgf2022, on 04/14/2009, -3/+53"wanna play Doom?"
"Sure man, let me hit the turbo button." - LaughingMan11, on 04/14/2009, -0/+48PowerPC is still very much alive. All 3 of this generation's game consoles are PowerPC based.
- anexanhume, on 04/14/2009, -2/+46Yeah, I cry softly to myself each night that we still use x86.
- jejones, on 04/14/2009, -5/+42Awesome? Oh, please. More like 31 years of endless kludges atop a putrid, festering obscenity of an instruction set. A 2 MHz Motorola 6809 could outrun the 4.77 MHz 8088 of the original IBM PC, and was vastly more pleasant to write code for. If only IBM had chosen the 68000...
- LaughingMan11, on 04/14/2009, -0/+35There is a difference between a particular PowerPC implementation and saying outright that the PowerPC architecture is inferior to the x86 architecture.
The specific example you brought up ignores the fact that on the IBM side, no one was interested in building Apple a smaller lighter next gen PowerPC chip that they could use in a notebook computer. On the other hand, Intel was committed to developing their new architecture targeting the notebook space first for lower power consumption.
The fact is, IBM was more interested in building high power processors for servers and game consoles than to appease Apple. - pathouston22, on 04/14/2009, -1/+29The article didn't really touch on the disaster that the Pentium 4 really was. It blew past the price of the Athlon. Intel went the RDRAM route while AMD went DDR - we know who won that. Intel ended up releasing "EE" editions of the Pentium 4 to maintain the edge over the Athlon, overloading them with cache and jacking up the price. I liked to call them "emergency editions" as they had no answer to the Athlon. The initial Dual Core Pentium 4s were a joke, running way too hot.
- ePuck, on 04/14/2009, -1/+26Ah yes hitting the turbo button in the middle of a game was comedy the whole family could enjoy.
- spritom, on 04/14/2009, -3/+271 page:
http://www.maximumpc.com/print/5972 - Almightymole, on 04/14/2009, -2/+25I usually NOP after chewing something before the next byte.
- RealmDown, on 04/14/2009, -2/+24Silicon clean up on processor 2, aisle 5
- Scooby223, on 04/14/2009, -9/+29RISC ftw
- nullx42, on 04/14/2009, -4/+23push BOOBS
- pathouston22, on 04/14/2009, -2/+20I still remember my dad bringing home a 486 when I was 7 years old. Asked if we needed to cut a hole in the top of the computer (monitors sat on top of the computes in those days kids). Was programming DOS startup scripts/menus and running memmaker trying to get the damn games to run not too long after that.
- Rikushix, on 04/14/2009, -3/+19LUDICROUS SPEED
- anexanhume, on 04/14/2009, -1/+17CISC has more complex decoding hardware than RISC. That means more power consumption. You can't avoid that fact, and that's why intel will have a tough time making Atom go toe to toe with ARM on the numbers. It isn't impossible, but they're handicapped.
- syathish, on 04/14/2009, -7/+227 pages??
- ukblacknight, on 04/14/2009, -2/+17My sisters PC had a key on the front to "lock the HDD when being transported".
- trisweb, on 04/14/2009, -1/+16To be fair, disabling a portion of a chip means they don't have to throw away chips with errors on that part of the silicon, thereby increasing yield. It's not dishonest in the least, simply a way to lower production costs and decrease waste, and pass the savings on to the consumer.
- garrettg84, on 04/14/2009, -4/+17pop BOOBS?
- dhughes, on 04/14/2009, -1/+13Also known as the Benny Hill button.
- fuzzybad, on 04/14/2009, -4/+16Shouldn't a "Brief History of CPUs" include more than just x86?
- marx2k, on 04/14/2009, -2/+13I noticed someone dugg me down because my laptop has a turbo button. That's just bizarre.
- jejones, on 04/14/2009, -1/+12Yes it does, because compilers always write in assembly language. Irregular, non-orthogonal instruction sets make it that much harder to create a correct compiler, and in the generated code causes lots of gratuitous moves of values into the magic register that is the only one that supports a particular operation.
- Gibletoid, on 04/14/2009, -1/+12Then how would you comment?
- nullcodes, on 04/14/2009, -6/+17x86 isn't as bad as people make it out to be, actually the ISA doesn't matter as much as people think.
People think ARM is awesome for mobile devices ..and they are right.. but actually it's the power saving design of the CPU .. not the instruction set that actually gives it all the advantages. What I mean is that the x86 instruction set doesn't actually force a design that costs all that much more power (added transistor count is made up for by the added instructions so it evens out). No doubt we'll see x86 processors announounced as running with the same average watts per operation as the future ARM processors within the two years (they'll be competitive for mobile phones and handheld gaming/HD media devices against ARM processors out at that time). - marx2k, on 04/14/2009, -1/+12My laptop still has a turbo button. MSI GT735.
- topcat5, on 04/14/2009, -0/+9Nice history. Not unique. IBM's mainframes have maintained this kind of architecture compatibility all the way back 45 years to System 360 introduced in 1964. (the machine that made IBM)
- dhughes, on 04/14/2009, -1/+10Try the Firefox AddOn Autopager it loads the next pages in a series instead of clicking next to see it.
- laznhome, on 04/14/2009, -2/+11Oh please, once the PowerPC adopted Altivec instructions, it was actually more CISC than the x86 competition..
Ever since the P6 (that is the Pentium Pro to you) Intel x86 CPUs have been powerful modern (for it's generation) RISC CPU's with a x86 interpreter - front end for lazy programmers.
http://arstechnica.com/cpu/4q99/risc-cisc/rvc-1.ht ... - strictnein, on 04/14/2009, -2/+11Fast and cheap. Who cares if something is technically better? With the cost of the "good" stuff far outweighing the cost of a ***** load of cheap x86 gear, why bother?
Linux Cluster of cheap x86 gear vs proprietary SGI/Cray. 99% of the time, I'll take the former. - trisweb, on 04/14/2009, -1/+10Well, sorry to tell you you're wrong. Do you have experience in the silicon fabrication industry? Thought not.
You had every choice to buy a processor that wasn't "crippled." But it cost more for a reason -- because they had lower yields with more silicon to fail. The practice still exists today with many processors but with clock speed instead -- some lower-frequency chips are exactly the same as the higher-clocked chips, but "binned" into the speed they're most stable at.
As a funny anecdote, Levi jeans does the same thing. To account for flaws in fit, they measured jeans *after* manufacturing and binned them and labeled them to the size they most closely matched in actuality, thereby decreasing manufacturing costs and quality rejections and increasing yield. They were the first large manufacturer to do this, and it was one of the reasons for their consistent fit and market growth. All from simple understanding of process statistics. Genius.
None of these practices is ripping off the consumer -- it's very, very smart business. - wbkang, on 04/14/2009, -3/+11Mac? "price vs speed" fail :(
- wbkang, on 04/14/2009, -2/+10Sigh.. ia64 much?
- cnldelta, on 04/14/2009, -5/+13Backward compatibility is over rated but its what sells.
- aristotle0dude, on 04/14/2009, -3/+11Sorry but 50GB with 48.0 Mbit/s bandwidth (Blu-ray) is better than 30GB storage with 30.24 Mbit/s bandwidth. You must be using the "new" math where less is more.
It's over. Blu-ray won. Ironically, HD DVD was proprietary to Toshiba hardware while blu-ray was developed and supported by a group of hardware companies which included Sony and Panasonic. - anexanhume, on 04/14/2009, -1/+9Narishma, just because the architecture busts it up into micro-ops doesn't mean that's not decoding overhead. Besides, the current line of Intel processors are definitely not considered RISC, so I'd be interested to see where you're getting this distinction.
- fallingdamage, on 04/14/2009, -0/+7AMD has gotten too busy as of late to really focus on processors like they used to.
Seems they have shifted from trouncing Intel to now trouncing nvidia. - MrAbbas, on 04/14/2009, -0/+7MIPS ftw!!!
- SarahC, on 04/14/2009, -1/+8Millions of gameboys and master-systems can't be wrong!
- jbmcb, on 04/14/2009, -0/+7Everything following the release of the original Core 1 line of CPUs from Intel were based on the P3 processor core. The Netburst architecture of the P4's just couldn't scale well, and took too much die space to make multicore on one die feasible. The Pentium M processors were based on the P3's as well.
- Defiant001, on 04/14/2009, -1/+8Bottom of page 6 ?
"The first Core 2 Duos burst out of the gates with 167 million transistors, a 65nm manufacturing process, 2MB of L2 cache, and a 1,066MHz frontside bus. Despite debuting at just 1.86GHz and 2.13GHz (E6300 and E6400, respectively), Core 2's performance made it instantly attractive, and Intel's aggressive pricing sealed the deal." - chuckDontSurf, on 04/14/2009, -5/+12Yeah, the title should really be "A Brief History of *Intel* CPUs"
- gigamugged, on 04/14/2009, -0/+7Actually ARM is far better per/watt than x86...
The x86 instruction set is awful but it does get the same jobs done with ~30% less instruction bandwidth, something the RISC guys always discount. ARM thumb instructions seem to get ARM into the same ballpark (and predication helps a lot too). Having worked on a few GPU instruction sets in my time, that's where the real innovation and perf/watt advances are coming from. I predict the eventual next-gen CPU winner will have fused CPU/GPU instructions and dedicated graphics hardware for things like rasterization and texture sampling (something Intel got horribly wrong in their current effort IMO). - Rikushix, on 04/14/2009, -0/+7Core 2 Duos and Core 2 Quads have exactly the same architecture, man, just one has two cores and the other has four.
They're all just Core 2's. - tushyd, on 04/14/2009, -0/+6I loved my P3 866mhz. Actually, it still runs pretty decently well with Ubuntu.
- YoWhatDaFuxUp, on 04/14/2009, -4/+10my computer runs on magic
- trisweb, on 04/14/2009, -1/+7I have to agree with this. I've always thought of x86 as a sort of messy ISA, like they stuck a new instruction in there every once in a while just for kicks, or if they needed a better way to add so and so vectors or whatnot. And it's true, in a way. But part of its success has to be because it's been so adapted and evolving. And honestly, anything can still technically be compiled to run on any turing-complete chip, so the fact that you can compile for i386 -- and it will still run on a 20-year-old i386 chip as well as your brand new Core i7 -- that's at least slightly impressive. It's utilitarian, and it works pretty damn well.
But then, I'm not a compiler developer. They probably have a slightly more negative perspective. -
Show 51 - 100 of 185 discussions
© Digg Inc. 2009
— Content created and posted by Digg users is