Slashdot Mirror
"Budget" Chips go Head-to-Head
StewedSquirrel writes "Anandtech has published an article taking a look at the low-end of the CPU market today. It takes Intel's newest Celeron processors against the AthlonXP and Duron with a Pentium 4 1.8GHz thrown in for comparison. All of these processors will cost you under $120, but the article shows that the old Duron (at barely $40) can out-perform Intel chips costing nearly 3x as much. In addition, it shows that the performance of the Athlon XP is head and shoulders above the Celeron processors, while costing roughly the same."
I love my shiny new AthlonXP 2500+. $90 for retail packaging, scoring well above my old P3-500, with plenty of room to overclock.
-bZj
.sig
It seems to me that the number of market sectors may be the ultimate decider here, rather than the actual technology :-(
:-)
Intel simply have larger resources - they can push money at blue-skies research, and non-profitable lines, whereas AMD (although successful) have to "bet the company" on every major decision...
In a way, I think it's because AMD is such an underdog, that I like the company - although the fact that their chips are damn good helps a lot
Simon
Physicists get Hadrons!
It's time for people to stop rewarding the Intel marketing machine, and start buying the best tech - AMD!
At the high end, 64-bit addressing is just icing on the cake! :-)
Galileo: "The Earth revolves around the Sun!"
Score: -1 100% Flamebait
I notice that Anandtech describe an 800MHz machine as "chugging along". Hardly. One of my older machines is an 800MHz Athlon Thunderbird machine with 256MB RAM and a 40GB disk. It runs GNOME and WinXP without any problems and continues to be extremely responsive and perfectly adequate for the vast majority of tasks I throw at it (except Games).
The same is true for budget chips - if you want a machine to go online, to do Word Processing, play a few older games or whatever, these chips are perfect. Putting together a full-blown capable system for $400, or buying secondhand, is a great way for people to get in to computing who couldn't otherwise afford it. Getting them on the bandwagon is the important thing, and whilst the hardware is so far ahead of the majority of software (at least until Longhorn comes out...) getting more people using computers in their homes is a really good idea.
They didn't even tackle the the Green Celeron. After all, Celeron is derived from the latin word 'celer', meaning speed. Of course, celery is the fastest of all vegetables.
On a serious note, people, including myself, are starting to worry about power consumption. I'd like to pick up a low power device for a BSD gateway.
Life is the leading cause of death in America.
I see AMD advertisements on the web all the time, but they don't seem to have much of the "big name maker" market. Why not? Is Intel so intreched that their value doesn't even matter any more?
AMD seems to have been kicking Intel's butt for a little while now technically.
I'd love to see some brand name servers start using AMD chips, look at what AMD's doing on the low end!
-Pete
Soccer Goal Plans
Still stick there at 1.25GHz on your G4s apple? Tsk.
/.!
What does you post have to do with anything?
Oh, right. You don't understand chip performance.
That explains it.
You are just spewing numbers, and of course higher means better. Righty-o. What'ya expect tho.
This is
If anyone here didn't already know this, please raise your hand...
Anyone???
Anyone at all???
Hello???
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
The article had 3, yes 3, banner ads for AMD when I viewed it.
Conspiracy mod ~ON~
I don't know if you have heard, but apple is now using a "new" chip that they call a G5 in many of their computers. Sometimes it is know as the IBM PPC 970.
Second, what does a discussion of budget chips have to do with Apple and G4s.
Third, it is motorola that makes very expensive G4 chips.
Forth, why am I responding to this obvious troll that will not read what I write:-)
Anand's site is so biased towards AMD it makes me sick.
Back in the day I used to read his site all the time. All he did was was rave about the K6 processor so I put togethor several systems for myself and friends.
What pieces of shit! All kinds of compatability issues with my Lucent win-modems, Creative soundcards, and 3dfx or nvidia cards. Their processors also run *way* hotter than Intels'.
2 of uncles have had the same experiences more recently with Althlon processors. They both swear that they'll never use AMD again and neither will I!
My roomate in college had an AMD system that blew out its motherboard once a year. My trusty Intel P3 lasted all 4 yrs.
I'd much rahter pay more money and get reliable Intel quality. To hell with AMD!
(P.S. I'm not trolling or flaming, I'm talking out of many bad personal experiences. YMMV)
A decent number of companies took up dual Athlons because of their great price to performance ratio, and Opteron looks like it'll become even more popular with the same type of people. It also has 64bit going for it, which will be useful for getting beyond memory limitations. I haven't really been paying attention to prices lately but Xeons are expen$$$ive in comparison AFAIK.
yeah i'd agree with you...you will get modded as flamebait/offtopic. myself included
do you even realize though that the IBM chips that Apple uses are completely different from Pentium processors?
RISC vs. CISC google on it. then repost.
if you want to compare processors speeds you'd be better off comparing Apples to Suns (err hide pun in there somewhere...well maybe not)
if you look at the tech specs a 2ghz apple will outperform a 3ghz pentium. why? because of a couple of reasons
1. RISC vs. CISC.
2. Bus speeds. ever notice how pentium class chips bus speeds don't seem to increasing at the rate of chip speeds? the processor can't do squat if it's waiting for info all the time.
3. instruction speed. this is where AMD shines! as an example (number made up to infer point) an Intel multiple instruction takes 40 clock cycles (so on a 3ghz chip if my math is right you can perform only 75mega multiples) where as on an AMD chip a multiply is 30 clock cycles (which running at 3ghz will give you an additional 25mega multiples)
Intel is for HW what MS is to SW. they don't redesign to improve performance they just leave it there and add extra stuff.
Please mod me down now thank you
So to summarize the parent. Intel big, lots-o-bucks. AMD small, greater risk, "better" product. Sometimes money wins over quality
Well, with economic brilliance such as this, we all know who'll be taking over for Greenspan when he retires.
At home anyway. This bears out my experience with Intel vs. AMD. The only difference I've seen in equivilently performing CPUs is the 2x price tag on the Intel.
The lower end of the Intel chips has always been worse at everything PC related then AMD. AMD is the budget king, that is a known fact for some time now.
Waste of time by Anandtech imho
This is something everyone who has built systems and read any reviews in the past few years knows. The duron isn't really that great of a deal but the 1700+ and the 2500+ axp chips are unreal. Both perform exceptionally well, overclock like a dream, and unless compared to c varient (800mhz fsb) p4's absolutely rape everything performance wise.
Shadus
Sorry to hear about your bad experiences. But you start by saying "from my own personal experience", then you end by making this generalization.
Sorry but I've had my Athlon 1.4 system now for three years. The computer is on 24/7 and has been for pretty much the entire time (meaning I never turn it off unless I'm messing with it). I bought it right after the new cores came out and they started using their equivilency speeds along with a "cheap" ECS mobo. This combo has been rock solid despite a lot of tinkering.
Not saying that my case is the norm, but I wouldn't from my "personal experience" be so bold as to make some sweeping generalities about the chip in general. Feel free to go back to Intel if that makes you happy. Just know that your experiences is not one shared by everybody. BTW, I've been using AMD proccies since the 386 days and loved their 486's. I've never had a systemic problem with their chips (one can always find bad apples in the bunch, but generally they've been great).
... that whenever I bought a new motherboard + CPU, and then after 6 months decided to upgrade I would ALWAYS have to by a new motherboard + CPU.
They changed their CPU specs faster than I change between my two pair of socks. (almost..)
It was like whenever they released a faster Celeron or P3 you would have to buy a new motherboard because the number of pins were (your current pins) + 1, and then we had the Slot-1 to socket 370,371,372,373.... Dunno where we are now.
I'll be modded down as flamebait or offtopic, when this is stuff that matters (...) Still stick there at 1.25GHz on your G4s apple? Tsk.
May not be that much off-topic, actually. It's blatantly obvious nowadays that clock frequency isn't closely linked to performance, especially when comparing different architectures like PPC and 386.
I don't think it would be nonsensical to run a benchmark comparing PPC, Intel x86 and AMD x86: if you read a few of the articles at Ars Technica, you will see how incredibly complex the P4 is. In these conditions, it wouldn't be surprising that Intel's chip is as different from PPC as from AMD's chips.
Apple sells:
PowerMacs
iBooks
PowerBooks
eMacs
iMacs
only one has a G5. the rest are G4s. I wouldn't call that "many" of their computers.
But Intel will continue to bully themselves deeper into markets, threatening new chip makers and computer manufactorers with it's giant weight.
HURRY!! Someone slap them on the wrist!!!
*DrugCheese rants*
is the term "low end of the cpu market," as if to imply these chips are somehow less than adequate...
Bah.
Both of my current linux desktop machines run these "low end" chips, and they run just fine, thanks very much. They all have a bunch of RAM... but other than that they are very vanilla... 1.3ghz Durons all. It makes you wonder what's really driving the CPU market (other than wow-look-at-this-shiny-new-CPU marketing).
Even if a man chops off your hand with a sword, you still have two nice, sharp bones to stick in his eyes.
These Sub-$100 CPUs serve as decent upgrades for aging systems (e.g. the P3-800 that is barely chugging along)
I'm using a P3-550MHz, and it's fine for everything I do all day.
Can I have that 'useless' 800MHz chip when you toss it?
1. RISC vs. CISC.
Dude, go do some research on the latest Pentiums. They may still be saddled with having to support a CISC instruction set, but they are primarily RISC processors "under the covers". Plus one of the primary selling points of RISC way back when was that since it used a "simpler" instruction set, they could clock much faster. Well uh, that blows your comment out of the water. Fact is, a 2 Ghz PPC can outperform a 3Ghz P4 because of design decisions made by Intel on how to achieve performance.
2. Bus speeds.
How of the things that has hamstrung Apple for a long time WAS their lame bus speeds and general lack of performance in the systems surrounding the cpu (slow memory, slow support buses, etc).
3. instruction speed. this is where AMD shines!
Once again you blow your first point right out of the water with this statement, and reaffirms what I said earlier. Intel chose to go the route that says "we'll achieve performance improvements by architecting a system that allows us to easily increase clock speed in leiu of doing more per clock". Others, like AMD, choose to achieve more per clock at the expense of making it harder to scale their clock rates up. Go read about pipelines to get a better feel for where these tradeoffs are made.
Oh a couple more things. 1) Comparing Apples to Sun (by which I assume you mean to compare PPC to SPARC) is just as meaningless in the context in which we're speaking. 2) you should be modded down, but not because of your opinion, just the general lack of accuracy in your post.
The ads are random. I got a load of whoosing intel ads.
Is that AMD chips burn out 3x faster than Intel chips.
AMD procs run hot but that is why there is such a thing as a heat sink. Many people do not know how to properly use a fan with their heat sink, and many have never heard of silver compound. If you dont cool the AMD proc well you will have problems. So the word is heatsink... say it with me now h-e-a-t-s-i-n-k.
If I wanted easy I wouldnt be an engineer or a patriot.
(Page 8) is this correct? 18.4 Seconds to compile QuakeIII Arena Source Code on a 2600+ ? Maybe it's right, but I was expecting it to be longer.
Damn that's fast - all those months/years of id's hard work, only to compile in 18 seconds on a budget processor....
No Norm, those are your safety glasses; I'll wear my own thanks...
May I ask what the hell are you doing to your PC to burn out your processors? Am I in the minority in never having seen a case of processor "burn out"?
I think it was mentioned before but is worth repeating.
Not only are AMD great value for money, but you can upgrade them later quite cheaply too.
I have an 850MHz PIII laptop, and it is quite close to the point where the packaging changed for the +1GHz chips. So I can't upgrade what is essentially a perfectly good laptop.
I find this greatly annoying, and will be buying AMD next time round.
1) RISC vs. CISC is a moot point. x86 chips are RISC internally as well.
2) Bus speed is nice, but it doesn't help to have more bandwidth than you can use. If the bus has enough to max out your RAM, AGP and other IO it is fast enough.
3) It is true that AMD has higher IPC, but your figures are WAY off. Modern CPUs can do *several* multiplies *per clock cycle*. If think the theoretical number is 3 for the athlon and 2 for the P4 but I'm sure.
Anandtech is a big site, they have ads from every major computer brand known to mankind (almost).
If you watch closely or reload the page you will see Intel ads. On the left side of one of the pages there is a "Intel; Click here to get more performance" ad
Melius mori in libertate quam vivere in servitute.
This isnt flaimbait or troll....
Why is it that people only look at performance statistics of CPUs and not the reliability?
Would you do this for a car?
Would you buy a furnace for your house based on how well it heated your home but not how reliable it was? Or an air conditioner for cooling?
Hopefully not! So why would you do it with a CPU?
Many thanks Anand - this article will be a great help to many small system builders being beaten to death by Dell; many (such as myself) cannot compete with Dell on budget Intel kit of the same spec.
Having an article like this to show potential customers will mean I can provide better performing systems at competitive prices using AMD.
No Norm, those are your safety glasses; I'll wear my own thanks...
What are they doing to bias it then? They have so many benchmarks covering every possible usage pattern. Are they just making the benchmark numbers up? Or perhaps their pricing information is false? Give some evidence of the bias and I'll believe you. In the meantime, go find the other sites that reach the same conclusions. For example Tom's Hardware
"You can sometimes count every orange on a tree but never all the trees in a single orange. -A.K. Ramanujan"
WTF does that mean?
These green chips are mainly for the laptop, and then you have to buy the whole laptop. Where can I get me a green Celeron and an ATX board for it?
What it sounds like is that you didn't know or bother to find out what the hell was going on with those machines.
Wow, compatibility issues with winmodems. Imagine that.
Did you or your roommate ever stop to think that the problem might have been in the power supply (internals or cabling) that was burning out those motherboards? I doubt it.
The rest are configuration issues, except the heat. So make sure the heatsink is set properly. Done.
If I buy a boxed AMD processor with a heatsink, then I shouldn't need to buy another heatsink, or special "silver compound" or do ANYTHING special. The CPU should go in, with the heatsink provided, and work at the CPU clock rate advertised without any problems. What you're saying is that I need to take "special" steps to ensure my stock CPU running at the stock clock speed, core voltage, etc runs normally and lasts longer than 9 months. That's bullshit.
I'll stick with Intel.... I've had Intel machines running just about 24/7 for 7-8 years without any issues, even when the heatsink fan busted, the processor didnt croak like AMDs do even with fully functioning fans.
AMD chips run super hot. My Athlon box sounds like a buzzsaw with all of the fans it needs to keep from melting down into a puddle of silicon goo...
3) Several per clock? eh... pipelining helps push the average up. Don't we need to consider the size of the numbers being multiplied as well? Then perhaps a ratio of P4 to Athlon is more appropriate.
? I got a crapload of Intel Centrino ads
I've just received 5 P4 celery systems to install as part of a small cluster. The install is going down today.
Damn you Slashdot and Anandtech....
i have had plenty of hd's burn out, i have had monitors blow, and i have seen a proc/motherboard fried by lightning.
but i have never seen a proc or motherboard burn out on its own from normal usage. i run both athlon and pentium systems.
i have never seen a proc burn out and i still have an old pentium 1, and a 486 that i use occasionally. heck i even have an old computer with just 512k ram that still works...wish i knew what proc was in there.
I have had more Intel processors die than AMD.
AMd processors do not die faster... only people that have ZERO clue as to what they are talking about say that.
I don't understand why people still gets surprised by this kind of things.
Doesn't anyone know that Intel processors are inferior to AMD processors since about the K6-2 era?
The only reason why P4 keeps winning some benchmark comparison is because it runs at absurd clock speeds, but same clock speed performance are largely lower.
Personally I never even take Intel into consideration when I plan about buying a cpu for my personal usage.
At work, I use a similar system, with a cheap no-name heatsink. I got back from Japan this week to discover that the second CPU in three months had burned out. In the store room, we have a load of machines that have burned out in the same way. AMD chips get hot, and will burn themselves out less than a minute after the fan dies, so if you want to protect your investment then spend a little extra on a good quality heatsink and fan - it'll probably still work out being cheaper than an Intel equivalent.
I am TheRaven on Soylent News
I'm Not sure what he did, but it might have been the same thing I did...Except mine was an intel P3 800.
How did mine die? Actually, I don't know what the hell happened to it...It just stopped. It was half of a dual system, and the other proc STILL works just fine, but that one burnt out just like a lightbulb... It's a very expensive paperweight now.
I suppose he could be talking about a lot of the earlier generation Athlons...Those ran pretty damn hot and didn't have much of any thermal circuitry at all...Remember the Tom's Hardware Flaming Processor Video? That was some serious bad press for the Athlons...And probably the reason why they beefed up their thermal management stuff.
Personally, I'm impatiently saving up for an Opteron....I miss my dual system. I just home those dastardly cosmic ray CPU-killers don't visit again... Or whatever caused it.
Sig currently under construction. Mind the gap....
I'm running an AMD K6 1.1 GhZ chip on the machine I'm using at home, and I feel no need to upgrade right now. Nothing I'm running demands the extra performance... even though I know the cheapest eMachines on the market today is running at double the clock speed.
It seems to me that most consumer users don't need 2+ GhZ chips, but marketing over the years have told consumers that higher clock speeds always equate to better chips, even though that's a myth that Apple has worked hard to counteract.
Maybe it's time for a new generation of "budget" chips that don't headline their clock speed but instead compete on cost, power consumption, or other factors....
Comment removed based on user account deletion
what? You get heatsinks with your cpu's? Nice! I buy mine seperately no matter what proc I am using. Silver compound is not special it is normal. Welcome to the land of computers. Regardless of what heatsink you use, you need to use silver compound. Oh yea and a fan might help too. Different fans should blow in different directions when used with specific heatsinks. These are the basics man, if you want I charge $50/hr for training. I would be happy to tell you how to install RAM.
If I wanted easy I wouldnt be an engineer or a patriot.
I don't know what you mean by "burnout", but if you're overclocking that may be the cause, stick to motherboards that support it. Check out OVERCLOCKERS.COM or the chaintech line for some nice toys.
If you're actually cooking these things then you'll probably have the same problem with BOTH INTEL and AMD. INTEL certified cpu fans and power supplies tend to be beefier and move more air than the cheeper ones sold for AMD, but that's because you can usually cook eggs on pentiums. I try to keep systems I assemble to a case temperature of 75-90 degrees after 1 hour runtime. It seems to help all the systems but I've found pentiums tend to perform better when you add 2-3 fans to help. That one power supply fan just doesn't cut it.
Watch your airflow in the case also, the airflow should move over your chip location and not leave "dead air". An extra exhaust fan behind the chip pulling air out as well as a intake fan pulling air in the front should help. If you're using the new ULTRA ATA drives you might want to add drive coolers also. If the case is too jammed up with cables (like servers) try the new rounded cables, they really cut the case temp down.
You can check your runtime temp usually in the bios.
Good luck guy, hope the next chip you roast is idaho and not silicon.
It doesn't matter what you wrap your emotions around, Reality is a brick wall specifically designed to scramble eggs
No, pipelining helps push the clock frequency up. A non-pipelined processor would be faster at the same clock frequency because we would have no pipeline stalls.
What pushes the number of operations per clock cycle up is the fact that the processor has multiple execution units and the ability to execute several instructions at once.
We don't need to consider the size of the operands except is we run into the limits of the RAM or cache bandwidth trying to get them into the CPU, or they are larger than what the hardware can handle. For exmple 64bit integer multiplies are way faster on the G5 and the Opteron than on the Athlon and P4 simply because they have 64bit integer multipliers.
In real life we never reach the theoretical peak values of course, but we can still get a couple of multiplies and an add or so done each clock cycle with optimized code...
that is a bad comparison because normally you don't die if your computer crashes...
I'm surprised your chips burned out, as when the temp goes above a certain level, the on-chip diode shuts it off to avoid damaging the chip.
A blog like any other.
Its depressing when the "low end of the CPU market" beats all four of your machines.
This comment was generated by a squadron of trained super elite albino ninja chickens for you.
Has anybody noticed that Anandtech is testing this budget, $60 to $90 processors using a radeon 9800Pro w/256mb of video ram? That's a $400+ video card!
Is people really buying this kind of video card on a budget PC? I'd rather test the processors using a budget video card instead. It might become the bottleneck in some games, but I think that's what the consumer wants after all... an idea on how much faster their game will run on a realistic machine, not in this monstrosity.
VIA VPSD, the motherboard-end of VIA has a great board, the C3M-266, it's a socket-370 board running the CLE266 chipset (native VGA support coming soon in XFree86 4.4, VESA for now). It's got integrated-everything and it's the only DDR chipset for the Pentium3/Celeron/C3 that I know of.
The board is designed to run a C3 at full-blast, which is very stingy on power. Also, the DDR makes it less expensive to own, as SDR is getting pricey. If you need more power than the C3 you can slap a celeron or PIII into it and cruise away.
I'd look for someone making a box running this board.
"Sometimes, I think Trent just needs a cup of hot chocolate and a blankie." -Tori Amos on Nine Inch Nails
Is Intel so intreched that their value doesn't even matter any more?
Sadly, yes. From today's Inquirer:
Chipzilla bungs IBM $18 million to keep Intel Inside
The awesome power of marketing coop funds
By Eva Glass: Friday 05 December 2003, 08:33
NAOMI SAYS that IBM's attitude towards AMD and its pesky Opteron chips is lackluster, at best.
Just like in the old days when different divisions of Big Blue used to compete with each other to sell RS systems, AS/400s, big tin and PCs, Sam Palmisano doesn't have a corporate wide policy towards AMD she says.
So every marketing manager gets to have her or his say in what goes down. Take, for instance the x325. IBM engineers, she says, did little more than to do the metalwork, while a Taiwanese company made the living giblets at the heart of the machine. MSI she murmured, MSI.
She also says that the small server division kindly accepted something like $18 million from Intel's capacious marketing funds to stop whingeing about Opteron chips and get with the Xeon game.
If you're expecting 2U or 4U Opteron boxes in the next three months or so, said Naomi, think again. The 1U system won't go in a 2U box easily because the mechanicals are all wrong.
And Fortuna 500 folk think there's not enough management features for these machines anyway.
Sheesh!
OTOH, the Tier 1 vendors keep losing market share to the white box makers. Golly, wonder why.
What kind of a sad joke is that?
That's okay, I won't really miss any source of info that stretches the graph to make a 1% difference look like a 200% difference.
But the best thing about AMD is that a modern Socket A mobo will still take pretty much any SoA chip and most older mobos will take chips that weren't even thought of at the time. Compare that to the Intel alternative, where upgrading is a painfully expensive business. Any coincidence that Intel sell a lot of mobo chipsets? Never.
When I am king, you will be first against the wall.
PC builders put a $3 fan and heatsink on a $100 chip. Spending the extra $10 for a real fan and heatsink makes a heck of a lot of difference.
My 1.3ghz Athlon@1.4ghz has been running fine for about 3 years now. Very good uptime, I don't think I had a single hardware related problem(AMD related anyways). It even survived my first attempt at using arctic silver. Man that was a pain in the ass to get right :p
I think most people who had serious problems with AMD simply got a lemon, or it was really something else causing problems. It happened to a friend of mine, and he swore off AMD even though he now knows that the problem wasn't AMD. Oi.
Was kinda funny, he got fed up and bought a P4 cpu and mobo, which was a downgrade compared to the AthlonXP he had(and without saying anything to me, so I couldn't atleast steer him away from the old socket type intel was phasing out), and then STILL had the problems.
Even if a man chops off your hand with a sword, you still have two nice, sharp bones to stick in his eyes.
:-)
i'm crackin' up at your sig
$8.95/mo web hosting
I agree about never running the top of the line stuff. I'd rather hold back and get stable performance at a lower cost. Same with graphics cards. Sure, I have to play newer games at either 800x600 or 1024 on my geforce4, but I just can't justify spending hundreds of dollars on the newest card. Besides, every time I've gone back to a game to see what I was missing at 1600x1280, it really hasn't been much.
The sending of this message pretty much inconveniences everyone involved.
I can see through time and space, so I can count all the trees in a single orange, but not all the oranges on a single tree.
" the old Duron (at barely $40) can out-perform Intel chips costing nearly 3x as much. In addition, it shows that the performance of the Athlon XP is head and shoulders above the Celeron processors, while costing roughly the same" -
isn't this what benchmarks and processor-buying-guides have been repeating for some years now?
Actual prices and megahurts levels change over time but putting it into relation the picture seems stable: getting the same performance from an Intel cpu is way expensive.
605413? Yes, it's a prime.
You can get a good nforce2-400 board without onboard video for about $80. You can get a retail AthlonXP Barton 333mhz fsb 2500+ cpu (with fan) for $90. You can get a Radeon 9100 video card for about $60. Throw in some good quality 2x256 ddr 3200 ram for dual-channel goodness for less than $100 and you have the guts of a machine that'll run all but the very latest and most cpu-intensive games with total ease.
I figure the whole thing with 120gb hard drive, burner, dvd, case, monitor, etc. will run about $800. Imho it's the best deal on the market right now, price/performance wise.
I had an Asus A7M266 and tried putting a Thoroughbred in too. Wouldn't POST. The reason is not to do with the core (Thoroughbred) as such, but with the clock multiplier!
:o(
Turns out, many motherboard manufacturers incorrectly implemented the clock multiplier circuitry for high multipliers (above 12x I think). I got the impression from the internet that AMD released the full specs for all multipliers (future-proofing) but motherboard manufacturers skimped on a trace and some logic! Even Asus
I think they should replace/fix motherboards free of charge for such negligence, but know that the weak regulation of trading standards we have (UK and elsewhere) means that'll never, ever happen.
Ben Dover
Meanwhile all the cool motherboard stuff coming out seems to work best with Intel chips. I'm not sure I'd trust another AMD system to reliably push 3D, and didn't Carmack promise a Linux port of whatever next big game it is he's working on? I miss tribes2, and kernel-crashing half an hour into a game was never any fun...
Oh well, I'm sure the prices will all drop again next year :-)
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
anybody else notice that there were huge intel advertisements at the top of each page, but amd killed them in EVERY test?
i sell illegal drugs
It does everything I need; when I crunch a big job that's gonna take a few minutes, I use it as an excuse to go get more coffee. It's good to see they're thrashing Intel performance wise. Now all they need is the sales.
My mileage varies completely. In my case, I've never had a CPU die that wasn't related to a power supply blowing out and taking the motherboard and CPU with it. AMD processors have lasted just as long as Intel ones-- and that's on stock heatsink/gunk combinations.
If your experiences vary, it's likely where you're buying your parts from. All mine have come with the heatsink seperate packaging, meaning the reseller has some discretion on what heatsink and fan is sent.
Heh heh - - I pushed it to around 160 on a generic VLB mainboard and thought I was in heaven. I've built more systems for myself, friends, family, and work than I can remember, and every one has been built on AMD. CPU related stability issues have never - - and I do mean NEVER - - been a problem. My years of system building have convinced me that, when stability is a problem, you should eliminate drivers, physical connections, adapter cards, and mainboard components in that order. I know bad CPUs do surface occasionally, but I think that most people get themselves in trouble through pushing voltages/clock cycles and not compensating with good cooling.
I hate seeing money wasted, and the Intel name to me has the same connotations as "BMW" - - it's more about hype than bang-for-buck.
It's only funny until someone gets hurt. Then, it's hilarious.
These low-end chips are very much for general office and light retail/home use and not gaming or fancy rendering. Dollar for dollar you can configure a system thats more "productive" when using AMD then you ever will with Intel. If you want serious office or web surfing productivity then invest the savings in a dual-screen setup and more memory and NOT more CPU.
Your problem is that your motherboards are NOT up to AMD-spec. AMD spec now references that the board now HAS to cut back on CPU speed to handle a fan failure and protect the CPU from burnout.
AMD didn't enforce this, but changed their spec to enforce it after complaints. As far as I understand it, it was ALWAYS a reccomended board feature.
since we're sharing anecdotes, for me, the brand of hardware doesn't matter. I'm running a used SMP intel/via setup now, and my only stability problems were from an old video card. My previous setup (that I use for a secondary system now) was AMD/via.
Both systems are rock-solid stable, without any problems. The reasons for instability could always be traced back to the operating system and its drivers.
The reason I like Open Source OS's so much is it is much better to tell what's broken so i can fix it. But for these two systems Windows XP is up just as long as Mandrake Linux, and they'll both be shut down at the same time when I go home for break...
Who is this Anonymous Coward character, how does he post so much, and why is he always such a whore?
Of course, the points you raise may be pertinent for AMD users. Stock Intel retail processors have a basic modicum of quality assurance.
And how do you explain that I have a AMD 486-DX4-100 still running a firewall, a K6-233 still running a fileserver, and a Duron-800 that is still used as my daily workstation. Sounds to me you've had a streak of bad luck or you're combining those AMD procs with some really crappy MB's.
If somebody asks me for advise for a budget computer, I answer Athlon 2500+ (Barton core) with a decent MB and decent memory.
If somebody asks me for advise on a high perfomance computer, the answer is exactly the same. Why pay 3 to 5 times more for only 25% extra performance!
As a very small system builder, I agree. I've been making mostly Celeron systems lately for my low end boxes (which is mostly all of them). There are two reasons for this:
(1) Heatsink design. The Intel retail heatsink is foolproof, quick, and hassle free. My time is money.
(2) Stable value motherboards. I generally will build a value system on a SiS651 based board. In my experience, these things are completely rock solid, and MUCH cheaper than any Nforce2 based boards that I can get from my distributer. I had some bad experiences with VIA based all-in-one boards a little over a year ago, and am still somewhat leery of them.
Of course, if someone mentions gaming at all, I immediately steer them to what I call a mid range system which is AMD+NForce.
But if someone is just going to surf the web and do a little word processing, then I would still recommend Celerons. They end up cheaper with better motherboards.
I wish my distributer stocked Durons, then I could use a much cheaper CPU, and get a good Nvidia-based motherboard for around the same.
Mark
------- Mark
Remember the floating-point-calculation bug in the early Pentiums? Intel had found it, fixed it in their new Pentiums, and kept quiet, hoping nobody will notice. Of course someone did, (someone always does!) and it turned out that lots of people were using faulty calculations in their work.
The decision not to extend the Pentium processor line to 64-bit, and to stick with the Itanium ( which does not natively support 32-bit operation) is another shot-in-the-leg. Market signals (very few Itaniums sold) and voiced concerns from people like Linus Torvalds did not cut into their strategy.
AMD on the other hand, is building a new factory (Fab 36) in anticipation of the increased demand for Opterons and Athlon64s. The desktops we will be buying in 2005 (2004?) will be 64-bit, and it looks they won't be "Intel inside".
And not "Windows outside", I think :)
But right now, I'm stuck with Celeron/WinXP boxes at work, unfortunately. Gasp.
The Inquirer: Linus Torvalds, Itanium "threw out all the good parts of the x86"
AMD Breaks Ground on 300 Millimeter Manufacturing Facility in Dresden, Germany
I have many old(er) computers around here. My fastest is a Pentium III laptop, and my slowest is an old Pentium 133 box. Basically, whenever I got to the point where I needed an upgrade, I just got more RAM. Cutting down on swapping has brought me significantly greater performance improvements than having a "faster" processor.
Further, I am sick and tired of the market hype that surrounds clock speeds. It's not the processor, but the software that needs to be made more efficient. And because many programs spend a lot of time processing graphics and GUI stuff, I think that making video boards "smarter" by adding GUI-specific processing features would bring a significantly greater performance improvement, by offloading crap from the main processor, than speeding up the main processor.
All of that said, it doesn't surprise me that an old processor is "faster" than a newer one. There are hundreds of variables affecting their respective performances, the biggest one being the software used to test them. In most cases, I think it's like comparing apples to oranges.
Personal computers have gotten so fast and powerful in the past couple of years that I think what's under the hood is totally irrelevant to 90% of the users. The other 10% have specific needs because of high-end applications or something.
Sad to hear that so many people were disappointed in a 2.6GHz computer (yikes, how could you be so miserly, you Scrooge!), glad at least the recipient was grateful.
...) that (and again, this is a few years ago) that a 400MHz Intel chip is a lot more than "sufficient for running basic word processing apps." This claim drew blank stares, mostly. "But ... but it's on the low end of the line. Therefore, it's 'sufficient for basic word processing apps.'" Complete lack of remorse or comprehension ;)
;))
Note that when I worked (just a few years ago) on the dreaded and mostly-dreadful Marketing side of things, the same sex-organ-envy logic applied, and it's endemic to the marketing side of computer companies. We (at least I) on the ad-agency side would point out occasionally to the marketing managers (on the computer company side -- 4 letters, in Texas, rhymes with "Prell"
Not that people don't sometimes decide that their current computer is plenty (marketing is a broad activity, and thank heavens it's not a one-way decision maker), and not that new and ever-better computer guts and computers aren't something I'm very glad to see, but the marketing machine selling computers both personal and corporate is fueled by buzzwords and appeals to envy and other insecurities.
(That said, Hey, I am not immune to envy or curiosity, and would love to have a latest-n-greatest workstation -- an item which would of course change as the years go by -- at any given moment. My actual current work machine is way overpowered for anything it's called to do, though -- and it's a hand me down from a friend who really does have use for a more powerful machine, so asked me "Hey, do you have any use for a 1.2GHz Duron?" 1.2GHz!? I sometimes still have trouble believing that processors got past the 100MHz mark
timothy
jrnl: http://tinyurl.com/c2l8yr / foes: http://tinyurl.com/ckjno5
Best deals on the marketplace:
- get a cheap Dell 400sc server and upgrade the hell out of it
- get a cheap, decked-out Dell 4600 when they're on sale big time
Either way, you can end up w/ a 2.4->2.6GHz P4 w/ HT and all the goodies you want, and a graphics card far better than a Radeon 9100, for $600
The fact is, Dell sells machines below cost. They do this with a fair amount of frequency. If you jump at the right time, you can get a Dell far cheaper than something you build yourself
Only one of my systems isn't below the article's low end. The kids' system is decent, with an Athlon XP 2.1, but my desktop is an old K6-3.
Unfortunately it's AT, not ATX, so it's not a matter of a few hundred to replace the CPU and female-parent-board. It's the case, power supply, memory, etc. (But not the 1987-vintage true-blue IBM keyboard.) By the time I sink that amount of money into it, I'm NOT going for the low end.
I usually keep an eye on CPU prices, and buy around the knee of the price curve. Of course I agree that the other considerations, RAM, disk speed, etc, are at least as important as the CPU.
The living have better things to do than to continue hating the dead.
I've got a desktop system with an AMD K6-2 500 Mhz processor and 512 MB of RAM. The hard drive is a Western Digital 7200 rpm with 8 MB of cache.
And Dell still ships new machines with 4200 rpm hard drives.
Sure, I could buy a new 3.6 GHz system, but it would be slower than the one I've already got.
I've been building fast machines on a budget for the last 7 years. What most people fail to realize is that the average desktop user never uses more than about 300Mhz of processing speed. The rest of the clock cycles are spent waiting on the hard drive, memory bus, ethernet card, or the modem. My system building strategy is this:
- I buy the fastest hard drive I can afford. I get one with the largest cache offered.
- I use motherboards with the fastest system bus offered.
- I buy as much memory as I can afford.
- I spend the rest on the processor.
Anything above 1 GHz is simply irrelevant; I'll never use the processing speed. However, adding RAM and a faster hard disk does noticeably improve performance.And I always smile when people compliment me on the speed of my Macintosh (I've got a blue case) and I tell them it's a 500MHz PC. They can't believe that a processor "that slow" could be so fast. As if the processor speed made any difference.
It's not the hardware, it's how you configure it...
The society for a thought-free internet welcomes you.
I have to tell you, heatsink compound should always be used no matter what type of processor/fan you have. Just because you haven't used it doesn't mean the other guy is wrong or that somehow Intel's stuff is better than AMD's. Go look at any article from a reputable source explaining how to put a PC together and they tell you to use it. Can you get away with not using it? Probably, you could probably get away with not changing the oil in your car too but it's not recommended for maximum life span.
I have used both Intels and AMDs for years now and built a fair number of both, including my current Athlon 1.2, P3 550, P3 450, my father's Athlon 900, my brother's P3 600 and never had any problems with either "burning out" or any other heat related wierdness, and that generation of Athlon is supposed to run hot.
For the cost savings of AMD vs. Intel(performance/price) I think it is very little to ask to go and buy a slightly better quality fan, say $15 vs. $6 and some compound which you should be using anyway. There are many reviews which tell you exactly which fans cool the best and are quietest. Of course if you want to pay double the amount for the CPU so you don't have to buy heatsink compound( $3.95 and can be used on probably 20 machines ) that is your right.
Maybe I'm paranoid, but every page I clicked on had two sprawling AMD advertisements. I just can't trust these websites where objectivity can easily be called into question.
No, pipelining helps push the clock frequency up. A non-pipelined processor would be faster at the same clock frequency because we would have no pipeline stalls.
That paragraph is a bit confusing... it's correct, but one of the key things that needs to be pointed out is that it simply would't be possible to have a non-pipelined processor running at anywhere close to modern speeds. I think the last x86 CPU with no pipelining at all was the 80286. Certainly you could get more than 12 MHz using modern techniques, but you wouldn't come anywhere close to even 1 GHz. I'm not even sure if you could break 100 MHz.
If anyone out there wants to know what the hell pipelining and the rest of this crap is -- go over to Ars Technica -- they have some excellent CPU design articles. HowStuffWorks may also have some stuff, but I've never looked for it.
Now granted, I've only installed roughly two hundred AMD Athlons in the last six months, but every one of those worked exactly as you described. Out of the retail box, installed on the board, I use regular heat paste just to cover my ass (it's optional), attach the provided heatseak & fan, and it runs flawlessly.
I've seen a few manufacturing errors with the chips, but no more unusual than with the 200 or so Intel chips I've installed in that same period. I have never encountered a situation where an AMD chip with the provided heatsink & fan did not run at the advertised speed or voltage.
As for the 9 months, you may be interested to know that a large volume of components goes bad in the first few weeks of use. Where I assemble machines, we run a burn in test on every processor that is sold. This means that our customers rarely see a new but bad chip, however we see them often enough to not get excited when it happens. If you buy a chip from a place that doesn't do these burn in tests, then YOU have to carry the burden of dealing with a manufacturing flaw.
But anyhow, stick with Intel if you like. It's my opinion that they're overpriced and (as a graduate student in computer science) I'm convinced that HyperThreading (tm) is 85% hype and 15% feature.
Whatever "insightfulness" the parent post contains is simply wrong. I work in a shop that sells both P4 and Athlon machines (among others) and we obviously produce a lot more profit when we sell an Intel chip because of the relatively exorbitant price.
The situation the parent post describes simply -does not happen-. If you take an Athlon out of the retail box and install it correctly, it requires no special additional parts and it will run exactly as advertised, barring a manufacturing flaw as I mentioned in my other reply. The only "special" steps necessary to make an Athlon run properly are to install it "correctly".
If that's reason enough to kick AMD in the teeth and buy Intel, suit yourself. I have installed WinXP on a P4 with a heatsink resting on the chip (but not latched down) so maybe there IS a real advantage for Intel in the "cannot install properly" crowd.
My first AMD was a K6-233 on a Shuttle board (603? they couldn't list it on their web site because Intel threatened to cut them off if they did, since it used the AMD (640?) chipset)
:)
While it worked fine, I had nothing but trouble when I got an ATI all-in-wonder. The card would work until you tried to capture video - then it would lock up. At the time there was something about this and non-intel chipsets. While I'm more than willing to believe it was crappy drivers on ATI's part - wouldn't be the first or last time - it has stuck in my mind that buying non-intel opens you up to potential future compatablity problems. Sure...it was probably a freak thing, but it left quite the impression on me.
When I do get a new machine, it'll probably be a dual AMD - though I am still nervous about it all. As you said - bang-for-buck, you get more. And while mine was not a CPU issue, it was still a side-effect of not getting Intel. But Athlon does sound neater than Pentium.
Of course, by the time I get the money together, AMD and Intel ma both be out of business
I have AMD processors since my first 286, then 386, 486 twice, K5, K6, Duron, Athlon, and Athlon XP (it appears I've never missed a single AMD desktop CPU). Just for the record, from 486 afterwards, AMD cpus have always been 20-30% cheaper and 5-30% faster than corresponding Intel chips. I have never had a single processor related glitch.
Remember when it used to be CPI? Heh. Anyway, your clock cycle counts are so completely bogus, it defies imagination. Not even the 8086 had cycle counts like that.
A multiplication does not take 30 clock cycles; the AMD pipeline isn't even close to being that long (although P4's is), so it can't be latency either. In fact, because of the pipelined nature of modern CPUs, each execution unit can pump out one multiplication per cycle (although there is a multiple clock cycle latency involved). Furthermore, the AMD chips can perform a number of multiplications in parallel (the P4 has fewer parallel execution units).
A modern desktop chip can push out billions of multiplications per second (integer or floating point), not millions. Throw in SIMD extensions like MMX/3D Now!/SSE/SSE2/Altivec/whatever, which can perform multiple multiplications per instruction, and you've got some heavy horsepower at your disposal.
I've actually run microbenchmarks on this, and I've certainly seen performance on par with 1 or more multiplications per clock cycle. Heck, I've seen that sort of performance on division, which is really impressive.
As for RISC vs. CISC, the continued success of the x86 line has pretty much put that old saw to rest. RISC is a lot easier to implement than CISC; the MIPS architecture is just about as simple a practical general purpose CPU as you can ever design. Performance-wise, however, it's just an instruction set, and the amount of silicon resources you can throw at the problem is a better indicator of the capability for performance than theoretical design points.
The main reason why AMD and Intel have the fastest microprocessors? Volume. Because they serve the consumer market, they can afford to invest more in making their chips faster and faster with each generation. The other companies started out with a lead serving the corporate market (and thus being able to charge more for their designs), but as the complexity level goes up, the number of companies able to actually design a competitive chip drops way down; it's natural market consolidation.
Let's even allow some overclocking into the picture. Now, I run Intel by preference (see my post above under the "AMD Blows" thread for why, tho by now it's probably been modded into the cellar). The usual way I discover that a CPU fan has failed, is that I have the machine apart for something else. My P90 (overclocked P75) spent a good deal of its life with no CPU fan, cuz those shitty little 486-style fans used on Socket5 CPUs didn't last for beans. I've got a P2-266==>300MHz that doesn't seem to mind a nonworking fan either.
Also, does an AMD recover from being cooked? Apparently a P3 can. Over yonder I've got a P3-500 CPU that came out of a machine that had been declared dead -- tho its only problem was that the former owner was a chain-smoker -- the CPU fan had literally smoke-corroded into crispy-crumbles, and the heatsink was clogged solid with smoke residue. Consequently, the CPU had been hot enough, long enough, that the mobo was warped (I had to pull the CPU out of the slot with vise-grips, I kid you not). On a whim, I cleaned it up, replaced its fan, and stuck it on a working motherboard. After a couple false starts, it fired up, and has been working fine ever since. With one minimal fan and a low-end heatsink. Running WinXP no less.
~REZ~ #43301. Who'd fake being me anyway?
ouch! probably the same guy that modded your post :(
2 weeks ago, I got a post that was moderated redundant - it was like the 5th post and _not_ redundant, and then, it was moderated overrated. Shit like this happens more often it seems.
I agree with the parent, if you buy a retail box that includes the fan and you follow the instructions in the manual while installing, then you shouldn't have to worry about the cpu dying. That's assuming you operate it as you're supposed to. If it burns up from overclocking, that's your own fault for not taking extra precautions.
Your car analogy doesn't work as most car manuals have a maintenance schedule that tells you to change the oil every so often. You don't have to mount a big fan on the front of your new car to keep it cool enough to operate properly.
Now, with that being said, I've never had an AMD cpu burn up. That claim sounds like a bunch of BS to me. As for the guy that thinks it's better to buy an oem cpu and super duper fan, I'd take a retail box any day. I've never actually had to use a warranty before but the retail box comes with a 3yr warranty (+included heat sink and fan) for maybe an extra $6.
I had one of those until two months ago. It was actually a 750MHz thunderbird, and I got a motherboard with a VIA chipset. It had a decent amount of RAM. It had pretty good performance, but I was really annoyed at its poor multitasking ability. I would run compilations under Gentoo Linux, and other large jobs, and the mouse would get jittery.
I decided to upgrade to an Intel 2.4 GHz with HT, and an Intel Springdale chipset. Now this computer is fast. It doesn't get bogged down at all. It is about twice as fast as my friend's 2400+ AMD. Both of our computers have a gig of RAM, but mine has dual-channel DDR. Still, I have no idea how my computer is so much responsive under linux than his. But I'm pleased with my Intel. And I really hate it when people cry unfairness about the advertising of megahertz speed. It's objective. I think AMD sunk really low when they pulled a Cyrix and started using something+ numbers. Plus there's no way they're true, when comparing to an Intel processor with HT. Since it's possible to run two threads at once, why not call my 2.4 a 5.8? I don't see myself buying an AMD again.
Dell accused of being merely a box stuffer.
OTOH, Dell is accused of using non-standard parts.
Which is true. One of my highest concerns when buying computing gear is that I can run Linux on it, upgrade it, and fix it, all from standard parts. I usually get my parts at the MIT flea market, Hosstraders' Hamfest, or mail order, and I want those parts (assuming they're good, which they usually are) to work in my systems.
Frankly, I'm scared of ANY brand name, for that very reason, and generally stick to the white boxes.
The living have better things to do than to continue hating the dead.
IMO, even a Celeron 2.0GHz will do for most jobs. Also, put a lot of RAM in the box and you are much better off than with a faster processor.
My Duron 1GH has 384 MB of RAM. My Celeron 2GHz has 512 MB of PC-2700 RAM. Plenty fast!
www.pawlitzek.com
>Your car analogy doesn't work as most car manuals have a maintenance schedule that tells you to change the oil every so often.
:-)
>You don't have to mount a big fan on the front of your new car to keep it cool enough to operate properly.
You are right, probably not the best analogy. What I was trying to say is that since most sources, including CPU and fan manuals that I have seen recommend heatsink compound & fan(or heatsink in the PII days), if you follow the manual you shouldn't have any problems. Similar to if you follow the routine maintainence on a car which is recommended in the manual, you should also be ok.
OEM vs. retail is a whole other issue. I usually buy OEM. The reasons for that are that I have never had a CPU go bad and fans are cheap enough I can run down to the local computer store and pick one up rather than waiting for the manufacturer to send me one. That being said I can totally understand if you prefer the retail packaging too, since it is as you point out only a few dollars more. It is strictly personal preference and my utter lack of patience dictates my choice.
Someday everyone will use Linux on an AMD processor, or will be running X on a Mac (whichever camp you happen to be in today), etc., etc., but until then "AMD Outperforms Intel" is going to be as ignored as "Linux is technically superior to Windows", and for the very same reasons.
I remember when the K6 first came out. Intel folks (fans and marketing) would claim that it would cause binary execution compatability problems. I never once had any such problems, either. I suspect that the majority of actual stability problems with K6-based systems was the result of bad chipsets or hardware failures. Windows surely couldn't have been at fault, either. Everyone knows that Windows - 95a in particular - were rock solid. :-P
:)
After all, as long as AMD implimented the x86 instructions as per the standard, everything would be OK. We (being consumers in general) know that now, due to the plethora of different x86 processor implimentations: AMD's 32 bit varieties, mobile, opteron; intel's itanium, P4, P3, P-M, P3-M; transmeta's crusoe; via's C2 and C3 processors. They all work just fine.
~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
If a cooling system is provided with the retail processor, it should be adequate. If there is no cooling system provided with the product, and the specifications for the processor call for such, then that is also reasonable.
What is not reasonable (and defies all but highly cynical or paranoid explanation) is why an inadequate cooling system would be provided by the manufacturer when they (of all parties) should know the specifications of the product best.
To conclude, this was not supposed to be a dispute over the quality of OEM vs. aftermarket accessories. Rather, it was mulling the question of just why any manufacturer would choose to include substandard parts when they could reasonably choose to provide no solution at all (e.g. "batteries not included" as opposed to including batteries that corrode and leak 7% of the time, even when the device is used according to spec [random analogy, ymmv]).
My only guess as how an old Duron or XP could beat a P4 is that the benchmarks that they ran were bias to those chips. Although I do beleive that you get more bank for your buck with AMD.
I think I think, therefore I think I am.
It isn't as perfect as it could be, though. I was once a die-hard Amiga fan (still love the old beasts), there were things you could do with them that at the time were *impossible* to do on a PC, even one with a processor 4-5x faster, with soundcard and VGA - it just wasn't possible. It wasn't until VESA and other improvements that we really saw the PC take off. The Amiga had a very tightly integrated chipset, with each facet (CPU of course, but also sound and graphics) being completely independently programmable and not CPU bound - each worked on its own. There are similarities in PC architechture today (DSPs, DMA, coprocessors on various graphics and sound boards, etc) - but it isn't as seemless as what the Amiga had.
Part of the problem is the need for the PC to support so many different standards and interfaces, legacy and otherwise - that holds it back. Part of it is not having a common base (there wasn't much difference between one Amiga and another - at least within models - but even across the whole line it was pretty consistent).
Reason is the Path to God - Anon
I hate the word budget. My computers would be considered budget because they are "only" an Athlon 1.33 GHz and a K6-2 500 MHz for the main ones (have half a dozen slower than that for other things).
;)
People that I meet, after telling them I'm a programmer and computer enthusiast, they always say "wow you must have that new ___ thing" (3.2 GHz Extreme Edition P4, 64-bit processors, Serial-ATA drives, DVD-RW etc).
Fact of the matter is I don't. They always kind of stutter like, "What do you mean, why don't you have the top end stuff?".
That is my rant. I'm not saying that Athlon 1.33 GHz is all you need. It is fast enough for me to do my job but don't think I wouldn't like a 3.2 GHz chip for faster build times because god knows we'd all like that on larger projects.
I can't keep up with tech. I typically buy processors when I can triple my speed at a minimal. So the next thing I'll get is 4 GHz which won't be available for at least a year. And even then it will be priced high. So tack on another year and I'll hopefully be able to get a budget 4000 processor
Please don't tell me you actually believe all that crap. Dell will ship you a very nice and balanced machine if you want it.
Um, but I still can't get a Dell with a faster hard drive. Until that happens, I won't notice much of a difference in program loading times. Anything in memory loads almost instantly, so the faster memory bus doesn't buy me anything either.
There's no way in hell your 500MHz wonderbox is going to beat a new Dell w/ an 800MHz bus
So? If I wanted to brag about system speed, I might be inclined to upgrade. But I don't need, nor would I use the extra CPU cycles that the Dell would give me.
The gist of it is this: the processor speed is the least influential factor when it comes to system speed. In spite of this, computers are routinely marketed according to processor speed, not hard drive or memory bus speed. The average consumer who goes out and buys a 3.0 GHz dell really believes his machine is 6 times faster than mine. But a visit to my place leaves him disappointed because:
- He's still working through a 56k modem, so his machine can't load web pages any faster than mine. Since I'm using broadband, my computer feels faster because the web pages load faster.
- Applications pop up faster on my machine than on his - he's got 256 MB of RAM, 3/4 of which is used for Windows. I've got 512MB, so entire applications are cached in memory. A read from slow memory on my machine is still faster than a read from disk on his.
- And even when my machine doesn't have it cached, our disk drives are the same speed. So the apps load just as fast.
Even though he's just paid out $2k for a fast machine, his user experience isn't any better than had he bought a properly configured garage sale PC.Once you get above 1 GHz, processor speed has absolutely no bearing on user experience. After this point, the amount of system memory and hard drive speed have the biggest influence over the perceived speed of the machine. Yes, I suppose you could buy a pretty well equipped machine from Dell, but you'd have to know what to ask for.
I'm a little tired of hearing that Intel CPUs are more stable than AMD CPUs. Having used a combination of both for over ten years, I'd have to say there isn't much difference in reliability.
However, if you put a good chip on a cheap motherboard (PC Chips, ECS, etc..), use cheap memory, or overclock it without proper cooling it's going to be unstable no matter what kind of CPU it is.
The article's about low end chips. Low end AMD chips pump out LESS heat than the equivalent Pentium 4 CPUs. Go look up the specs for an XP1900+ if you don't believe me.
Sheesh, people get this "AMD chips run hot" idea in their head, and they cling to it year after year, never even pausing to wonder if it's still true.
GCHQ Quantum Insert installed. If only our tongues were made of glass, how much more careful we would be when we speak
1) The P4 and A64 are RISC chips (wearing CISC clothes). They very first thing they do when they recieve an instruction is decode it from RISC to CISC. With the Trace Cache on a P4 this operation is "free" when the instructions fall into the trace cache. Besides, the primary advantage to RISC was less bus overhead (A64 beats G5 in this category) and better clock scaling (P4 beats G5 in this category).
2) A64 runs a bus speed at FULL processor clock speed. (Apple runs at 1/2 clock speed). The highest end G5 runs 1GHz bus, while the P4 runs 800Mhz. Seeing how the G5 is quite comparable to the P4 in terms of bus efficiency, the difference isn't that significant. In addition cache speed and size plays a huge roll in this. The A64 and P4EE (the highest end chips) both have arguably superior cache performance to the G5. While the G5 does have very good caches, it's definately not a point you can use to prove yourself right in this case.
3) Where do you figure that a multiply instruction takes 40 clock cycles? That was the case on the 8088 processor, but modern chips CAN (yes, in fact, they do) do an instruction every single clock cycle. this is why their pipeline is so deep, but they can do it. So can the G5. The biggest difference is the number of "parallel" execution units and how efficient their SIMD are.
All three cases, you don't seem to understand the root of your own argument, leading me to believe it is somewhat "canned" from the assemblage of someone else's information.
Eric
There are 10 kinds of people in the world. Those who understand binary and those who don't.
The way how 'modern' operating systems take over 100mb of ram after boot up, take half a minute to boot, run slowly and consume gigabytes of hard drive space is insane. The same lunacy has happened with software in general.
Remember the pedal-to-the-metal days when we had things like AmigaOS, MacOS7 and Win3.1, which ran happily in a couple of megabytes? Remember how much faster they were on a 50MHz CPU than these new computers are with WinXP? Remember how well everything performed in 30MIPs? Yet now we've got _3000MIP_ computers that PERFORM WORSE than our early 1990s antiques with 2MB of ram! Now you can say those old computers did less, and sure that is true, but imagine what computing would be like if their software practices had been maintained - boot ups in 2 seconds, 20mb of hdd space used, DX2/66MHz CPU running fine for browsing & email, etc.
Of course any average geek would snigger at this. Of course they have to be 'cool' and upgrade yearly. Keep up the MHz treadmill, or else be scorned and laughed at. And many try to say that because modern software is more complicated, that it naturally scales upwards in system load. There is some truth in this, but 10s of megabytes to do what would've taken 100k? The base of the matter is most programmers constantly have up-to-date computers, they program their computers with only the project's performance on their on system in mind, and are often happy to neglect an additional 10% more effort for massive sacrifices of performance and ram usage - because it doesn't make any difference for them personally with their perpetually modern computer. So brick after brick of foundations is layed on the don't-care, snigger & laugh at the old, mentality. And consequently desktop Linux runs like a dog on my 3 year old computer.
And hey, there're literally half a billion or something rejected computers because they're 'old', sitting idle in garages around the world (or worse, landfills in 3rd-world countries where they go to be 'recycled' - the 'recycling' workers are oft to get cancer & die). This is an economical and environmental catastrophy of a vast scale. All because programmrs and companies usually hold a don't-care attitude.
And do you really think Windows and desktop Linux will be able to boot in under 1gb of ram by the year 2010? Little chance. This means virtually every computer ever made to date will find the bin or the attic for no reason other than human sociological failings.
The fastest machine I have is a VIA M10000 with a 1GHz C3. It's been running non-stop for months.
The review CPUs cost you "under $120"? I should hope so, my CPU and motherboard together came to $150... If $120 for a CPU is "budget", what's mine?
My "plan B" if I decided against the C3 was to get an Athlon XP1900+ for $50. Yeah, a 1.4GHz Duron's only $36, but I was prepared to pay the extra $14. Why the hell would I spend $110+ for a Pentium 4?
You have to be a complete sucker to buy Intel. As I explain to people, those non-stop TV ads are expensive--where do you think Intel gets the money?
GCHQ Quantum Insert installed. If only our tongues were made of glass, how much more careful we would be when we speak
I used to work for a company that provides hardware support for over 25,000 computers. Basically, large corporations would send their computers to us when something had gone wrong with the hardware. During our yearly reviews we would check to see what our most frequent problems were. It turned out that for all our CPU problems about 3 out of every 4 of them were AMD processors. This is unbelievable too, considering that about 19,600 of the computers we handled have "Intel Inside" of them, compared to about 4,100 for AMD.
We also noticed that motherboards with the Intel chipset fair much better as well compared to those that don't. Doing a little math, it's easy to see why Intel is on top. Even though the processors cost more in the beginning, in the long run companies still save money by using Intel. Especially considering that their is not much of a price difference for corporations when purchasing large quantities of Intel or AMD computers.
Microsoft should hire me. I can write code that doesn't work faster than the guys they have doing it now.
It's a valid point that Athlons do not last long after a fan failure. A perfectly valid counterpoint is that any reasonably modern motherboard will detect that and power the system off to protect the processor. I really couldn't consider this topic a very significant reason for choosing one processor over another, though, since fan failure is basically harmless these days (minus the awesome term paper you were typing that wasn't saved.)
You and your fancy pants 32 bit chips :-)
I feel so much better now with my recent purchase of a handful of Athlon 1800, 512MB, 40G, Nic, no modem/etc boxes from my local neighborhood shop. I'm running them as app severs at home to try out a few things here and there. After reading this thread, it seems I'm not the only one who isn't caught up in the arms race to have the faster CPU... Lets see, I got three 1800s for around $1200 total. That's less than my sister paid for her high-speed Dell home spam/virus collector !!!
"Whoever would overthrow the liberty of a nation must begin by subduing the freeness of speech."--Benjamin Franklin
AMD K62 - 500 on an Epox main board w/ 128 MB of ram has been running for 4 years as my edgenet router. With zero stability issues. The only restarts i've had to do were either maintanance based - or of course the '03 black out.
___________________________
I'm not a geek, but I play one on TV.
For the things that most of us do all day, adding a second slow processor makes a bigger difference than doubling the speed. There's no way I'm going to upgrade from my dual Celeron 933s (in BogoMIPS, that's 400+something) until I can get SMP with budget CPUs. (And I'm not talking about maybe-it'll-work-with-some-solder-SMP, I want at least some assurance of success that I'll work before I buy it.)
I vowed I'd never buy Intel again after they intentionally crippled the Celeron IIs. Who knew I wouldn't be able to buy AMD, either?
On a related note, ExtremeTech has an article detailing how to build a fast PC for $800. The final recommendation uses an Athlon XP 2500+ CPU with 512 MB RAM, 120 GB hard drive, and a GeForce 5600XT video card.
Does anyone find it odd that the comparison is between a chip that was _designed_ to be a budget chip (celeron) and a _flagship_ chip (AMD XP)? It seems like a very stupid thing even to dream up that test. This woud be like comparing the Pentium 4 2.0 with the celeron 2.0. Of course there's going to be a major preformance on the pentium over the celeron because the pentius was _designed_ to be a maximum preformer. A more even test would to exclude the AMD XP processesors and use only Durons.
Dont whine by saying "But the AMD XP's are cheaper!" Well so what. Your still tipping the scales just be using an old top of the line chip with a budget chip
Would you race an 2004 accord with a 1990 corvette and expect the accord to even have a chance?
budget == $ !!!
budget (here anyway) means cheap.
When a chip that was designed to be a budget chip is beaten by a non budget chip, on both performance AND price, it lost bigtime. If it won on performance or price I would buy your argument.
I forgot to mention that in my post. Retail Athlons now come with a patch of thermal paste already on the heat sink. Much cleaner than when people like myself put too much on. :)
But like you said, personal preference and all. My gripe was really with rudabager's condescending attitude.
(see subject line)
intel p2 350mhz
128mb pc100 ram
16mb video card
NT 4 server with everything turned off
The machine runs at 24mb ram is rock solid stable and great for web surfing, maming, mp3.
A well tuned machine is much better than a lame fast machine with dozens of unwanted processes (e.g., most stock linux installations, win 2000, win xp).
Forth is a language, not a number.
Contribute to civilization: ari.aynrand.org/donate
One would hope that CPU overheat protection would be high on a good motherboard's feature list, tho from what I've read here today, seems it wasn't exactly considered a selling point until recently.
Also, I'd prefer not to have to rely on that as protection unless I know how they implemented it. Frex, I've got an early P2 ABit mobo someone gift me, that insists that a CPU fan be plugged in before the board will power up. However, it doesn't bother to check beyond that -- plug in a DEAD fan, and it still powers right up! Ooops.
~REZ~ #43301. Who'd fake being me anyway?
Why's there never a pr0n benchmark? That's all anyone cares about.
My friend always buys the latest and greatest hardware, yet he just uses it for Kazaa...