Slashdot Mirror
AMD Athlon MP 1800+ Processor Review
Lars Olsen writes: "Amdmb.com has posted a review of the new AMD Athlon MP 1800+ processor -- a big speed jump for the dual Athlon processor family with the new processor running at 1.53GHz. There are also 1600+ and 1500+ Athlon MPs available as well right away at stores around the World.
Dual AMD Goodness is now running just as fast as its desktop counterpart ! Here's a quote: 'Those of you who want to jump into the dual processing Athlon world will finally be able to do so with the knowledge that your processors are the top speed that the Athlon family has to offer. And for anyone who already has a Tyan Thunder or Tiger MP board and a pair of Athlon MP processors, you may just want to pop a couple of these new Athlon MP 1800+ CPUs in your system to boost performance.'" Some of the comments following yesterday's "dream system" article addressed dual-Athlon complications, so make sure you read before you buy.Update: 10/15 15:14 GMT by T : Check below for LinuxHardware.org's take on this chip, and Athlon MP systems in general as well.
Augustus writes "LinuxHardware.org takes a look at the Athlon MP platform under Linux and the newly released Athlon MP 1800+ is included. Covered in this article is not only the technology and performance of the AMD-760 MP chipset and the Tyan Thunder K7 motherboard but we also look at why anyone would consider a multi-processor system."
Well at least I can still count on my fingers how many GHz we have achieved. I suppose when/if these quantum-based computers come about (on a large scale), I'll have to have an infinite number of fingers all representing the possible states of the processer
I'm running at home on a PII 400Mhz and it runs everything I possibly need. My Mother and Father in Law are on a P166 and wondered if they should upgrade. I said no as it really doesn't need it, they just do basic database, spreadsheet and word.
All too often developers use the increased memory and processor speed to write worse implementations, or to create pointless bloatware. I know this will continue no matter what I say but at the end of the day who really needs this much power, QuakeIV players ? QuakeV ? QuakeIII runs fine with my upgraded graphics card, and top of the line sound card, the processor does bugger all.
Moore's law is great, it means computers can do more and more, but for the home market its just silly, 90% of people would be fine not changing their machine for 4 years, but they are forced to upgrade by market perception.
Faster this, faster that.... but never ever actually "better", "more reliable" or "stable".
Hardware is the excuse for bloatware, its not H/W engineers fault but it isn't an excuse to use....
(and yes this is partly a dig at the huge swap requirements on the 2.4 kernel)
An Eye for an Eye will make the whole world blind - Gandhi
Firingsquad has an excellent review comparing dual Durons to dual Thunderbirds using both Palamino and non- versions of both chips. They conclude that the Palamino Duron is the best bang for the buck.
Computer games are the ONLY applications that tax a home-users cutting-edge machine... At the moment, systems are a little ahead of gaming technology, but in a few months that won't be the case. Just because your parents don't play Dark Age of Camelot or AquaNox, don't assume Joe User doesn't want to.
-Berj
Note to web programmers, MySQL doesn't like it when it runs out of connections. Try increasing the connection pool size. Also, instead of having the page try to open the connection just once, and fall all over itself if the connection fails, try putting the connection request in a timed loop with a timeout of around 5 minutes, and a sleep(5) in the middle to help throttle a little. Your MySQL server will thank you, and your web page viewers will thank you.
I like how the little guys are going to benchmarks to indicate how their product actually performs while the big boys (Oracle, I'm looking at you) are recusing themselves from it.
Too bad that IT managers go with what they know (everyone else is using) and what's worked for them in the past.
It may be confusing for Jane Consumer, but it's nice to see that AMD's finally gotten a marketroid with a clue as to what works. Now if only their stock would start working, too...
Easy does it!
This comment has been submitted already, 276865 hours , 59 minutes ago. No need to try again.
It always happens... you jump in and build your dream system, and immediately it's out of date. Oh well, a duallie 1.2ghzMP isn't anything to laugh at! Glad to hear that the TigerMP supports the new chip speeds out of the box, anyone know how high it will go?
A few notes on the TigerMP though: VERY picky on RAM, very picky on how it's seated (read: install memory before board is in your case, so you can wedge it in on a flat surface!), but since getting past that, it's been ROCK solid! Beautiful system I must say!
MadCow... always 500mhz behind the curve.
I used to have a sig, but I set it free and it never came back.
Losing the fan but not the heatsink is not going to cause flame-out. And losing the heatsink is really only an issue for people in earthquake zones and LAN partiers.
If you're really really scared, get one of the heatsinks that bolts onto the motherboard instead of clipping onto the socket.
Why don't you take a look at the AMD Processor Roadmap to see more on their processors.... http://www.threecom.de/artikel/amdround1/ ... though the site is in german.... translate it with babelfish: http://babelfish.altavista.com
===> An eye for an eye makes everyone blind - MG
From your tone I'd expect you woudln't buy AMD anyway. However, if you did any research, you'd find the AMD's new numbering plan is actually conservative. Independant benchmark reviews have shown that the AMD 1800+ is actually more of an equivalent to the Pent 4 2GHz chip. But AMD chose a conservative threshold. Granted, the new Intel cores will boost performance a bit, but even then the AMD numbering plan is expected to be on target. Honestly - who cares what they call the chip - anyone with half a brain can find out the MHz value. But to what end? Me? I want to buy teh system which gives me the most performance for the least $$$ and right now that is an AMD chip hands down when you account for other CPU specific system costs and impacts (chipset, memory type needed, etc)
I honestly think AMD did what it HAD to do - their chips are faster at slower clock speeds and Intel managed to get folks thinking MHz was king. Now AMD has ot try and chance that thinking.
Top Most Bizarre/Disturbing Error Messages
I've been an AMD fan ever since, erm well always actually even my 486 was an AMD ;-)
I really think AMD will have te expect some problems with this. Back in the good old days (r) of the pentium and the cyrix 6x86 I worked in a computer store and we also sold cyrix computers to customers that didn't want too spend too much money (so sue me)
Very often people came back because they saw that their Cyrix PR200+ wasn't actually running on 200Mhz and demanded a refund (which they didn't get ofcourse) we had to explain the whole thing and it costed us a lot of time
That's why we stopped selling them back then
Another thing is that the semi-geeks (the dudes that THINK they are geek but basically know nothing) won't buy them because "they are already overclocked"
Fighting for peace is like fucking for virginity
While it's true that AMD CPUs are, uh, sensitive to cooling, I don't see that as a show-stopper. When you buy the parts to build your own Athlon system, as I did recently, you get plenty of warning to NOT TURN THIS ON WITHOUT A HEATSINK (yes, they shout, as they should).
Other CPUs are also very sensitive. What's rather surprising is how well Intel's P4 thermal shutdown works. I suspect AMD will get around to doing something similar. But in the meantime, I've attached a nice quiet (3800 RPM, not the 7200 RPM version) ThermoEngine to my Thunderbird, and it cruises at around 100 degrees F. Some newer/bigger heatsinks bolt to the motherboard, rather than clip on to the socket, which I suppose helps if you're really paranoid about its falling off. I use Motherboard Monitor to keep track of the temp via the Win98 system tray, and wish Linux distros would include similar capability out of the box (yeah, I know there's a way to build it in yourself...).
But then I do admit to using a 1 GHz Tbird rather than a faster one because I don't want that excess heat or power consumption.
Yes AMDs will incinerate themselves if the heatsink alls off - but funny, you don't see many people saying this has happened - yes it has to a few, but honestly - I'd rather get the higher performance for my dollar and risk having to replace the CPU if the heatsink fell off - something very unlikely. But if it did, the replacement CPU would be pretty cheap given how prices on processors fall over just a few months! And total cost would STILL probably be chaeper than an equivalent Pent 4 system (not CPU, system) Hell my 1GHz Athlon has been chugging along for months and the heatsink is still on solid!
Top Most Bizarre/Disturbing Error Messages
Computer games are the ONLY applications that tax a home-users cutting-edge machine... At the moment, systems are a little ahead of gaming technology, but in a few months that won't be the case. Just because your parents don't play Dark Age of Camelot or AquaNox, don't assume Joe User doesn't want to.
But the kicker is that these games really don't need such horsepower. I'm willing to bet that if there were any pressure to get any of these games running on a more resource constrained system, like a game console, then lots of unnecessary internal fat would be trimmed right away. But there's no pressure to do so otherwise. And even if a game that could run just fine on a PII 400 requires a 1GHz processor, certain people seem to _like_ the justification for upgrading.
Hey all you /. people with a fab, here's a fun idea to piss off intel and AMD. Make the clock/speed irrelationship totally obvious.
:)
Imagine an x86 compatible processor that runs at a clock speed of 50ghz? That's right, fifty BILLION hertz! Now, that clock only ever hits a counter that lets the 8086-compatible processor cycle once every half to full second. You could get a whopping 1-2 IPS
You'd be able to make millions selling 8086's that use the first 640k of a bunch of 128 meg chips, and the first 40 megs of a 400 gig hard drive. Think of the possibilities!
"Look at me, I invented the stove!" -- Ben Franklin
That is assuming of course that their time has no value.
Sanity is a sandbox. I prefer the swings.
A console is a very different environment. You can tune exactly for the hardware because there will be no variances. A PC game has to allow for 30 different graphic cards using APIs that supposedly make the different cards look the same to you but fail miserably. By the time you get done tweaking for the current morass of cards, a new generation of them is present with their own damn bugs. In the console world you deal with 2-3 environments IF you are allowed/it is practical to port given the current state of exclusive games. Also, if you've ever developed for a console, its very different, with a PC you have a lot of freedom to build how you want and what you want, in the console world you pretty much have to build around the hardware. This means you are constrained to build the same kind of engine for most every game you build on that console. If you don't you are just looking for different ways to cull the scene down to fit into the same miniscule space.
.</RANT>
The two environments are very different, and most of that fat can't be trimmed by wishing it away or blaming on programmers
As for bloatware, start modelling cloth, hair, IK, bump maps, and the hardware gets used again. The reason the games aren't doing it now is because they want the comfortable sales window.
Honestly pushing ultra-high-end features that cut your market to 4% of what it could be isn't a big selling point - good luck convincing your publisher to bring the game to market - and trying to build an engine to scale between low and high-end aggravates the bloat of PC vs. console problem even worse.
Sanity is a sandbox. I prefer the swings.
Except that the high-end game curve has not been keeping up with the processor speed curve. Back in the day, the top of the line processor was need to run the game. You needed a 386/25 or a 386/33 to run games that came out within a month or so of the shipping date of the processor. Now the specs require a PII 233 or a PII 400 or equivelent. I've yet to see a game that on the box requires a PIII/1G. Have you? They do require some sort of 3d graphics card, but technology wise, the games try to aim for the greatest penetration of market share, which is about at the 300MHZ-600MHz. That is unlikely to change, since most of these people won't even think about upgrading any time soon, even if the newer computers make the internet "go faster!"
http://www.tech-report.com/onearticle.x/2994
They only compare against the 1.2Ghz Athlon MP though... although they intend to do an expanded article soon.
shut up man
Even if it was the Athlon "Bill Gates Special Edition?"
- Freed
"Coffee should be black as hell, strong as death, and sweet as love." -Turkish Proverb
For your convenience, here is a list of other things you should avoid buying because they have "fatal flaws":
I am constantly amazed by people who claim that faster hardware leads to bad code as if we've been living in the Golden Age of quality code for the past few decades.
With current hardware, people are still writing code a lot of code in C and C++ for performance reasons which has lead to buffer overflows, segfaults, core dumps, general protection faults, and blue screens becoming generally accepted aspects of computer programming. Now that the hardware is finally becoming fast enough, maybe we can wean ourselves from C & C++ and move over to writing apps in Java or even C# instead of still dealing with the same issues that were solvable problems 20 years ago. Programmers have shown that it is practically impossible to deliver significantly problem free C/C++ code in a decent timeframe while programming environments like Java have shown the opposite. Once hardware creeps up enough we can rid ourselves of the problems of C & C++ once the performance gains are not worth the amount of bugs one has to deal with, which is already happening in lots of server applications.
Also once, hardware creeps up enough maybe some of the stuff that has been in research labs for the past 20 years can finally see some use. For instance microkernel are generally seen as a superior way to design an OS but have had difficulty taking hold due to performance reasons (although Windows NT is based on a -kernel architecure and MacOS X is also built on the Mach -kernel) which wil change once hardware advances make it possible for the performance difference to become acceptable.
A.I. being built into applications as well as the OS is another place where hardware performance and memory availability would play a big part in helping come to fruition.
How about voice recognition and face recognition being built into the applications you use?
How about bringing virtual reality to masses?
Or do you think that a 1 GHz CPU and 128 MBs of RAM is all the power a computer user will ever need?
The gist according to firingsquad.gamers.com is that there was a initial batch of XP's that were
SMP-enabled and mistakenly shipped. AMD supposedly will be disabling SMP in the XPs very soon.
To-do List: Receive telemarketing call during a tornado warning. Check.
Before AMDMB went splat, I read enough to see that in some tests (most notably memory), the Athlon XP (yes, SINGLE) beat the dual Athlon MP setup soundly. This is because the XP tested in a VIA KT266A motherboard, which has the edge in performance over the standard AMD760MP.
I think the Athlon MPs are awesome, but having a much cheaper, single-processor setup beat out a dually in some tests throws a bit of cold water on my upgrade lust.
shut up man
You're right, they could cut their polygon count down to a quarter of what it is now, precache almost everything (quadrupling the amount of hard disk space used) and probably use 50% of the CPU they use now. Game developers really are into severely optimizing their code, especially those programmers dealing with graphics; They're usually trying to find ways to optimize every single action.
On the other hand, as others have pointed out, the only way to really optimize the hell out of something is to write it in assembler. That makes any large codebase pretty much unusable.
The biggest thing game developers could do right now to improve game performance is to use really excellent multi-res in a game. Multi-res is a process where, when used to its fullest, lets you start with very high polygon models for everything, and the game engine will reduce the polygon count one vertex at a time, in some cases all the way down to a single polygon. When done right this will let you draw amazingly complex scenes without slowdown; The computer can tell more or less what you're looking at and decide what needs lots of polys.
Unfortunately, even those games which are using multires are using a low-rent version where they pre-reduce the vertex count, so you still "pop" from model to model. It's getting better, though.
The best thing about multires of course is that you don't have to precompute things, like BSP-based schemes, and that it will make the best use of your graphics hardware, while still running well and looking good on lower-end hardware. On the other hand, your graphics card had better handle lighting pretty damned well. Since you can get a GEForce MX400 card for less than $100 (Or a GF2 for about $150) that's really not much of an issue these days.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
As a side note - an Intel motherboard will short out of you let the floppy drive slide onto the board whith the power on. Pouring coffee into a laptop makes interesting smells. Putting a CD-ROM in the microwave for 10 seconds if you want a real show.
Seriously here, you are missing out if this kind of thing actually sways you away. The biggest flaw, IMHO, is the AMD cores chips way too easy. I would really like a coating of nickel or copper like the Intel chips have. As an early adopter of the Chrome Orb (rev 1), the hard part was safely getting the heat sink on.
I've found that an AMD CPU will give you warning signs like lockups, kernel panics, and other goofy things when you loose a fan. My mainboard will shut down 5 sec after the post if the CPU fan is not spinning fast enough! Since they are good up to ~100C, using a motherboard monitor prog will go a long way to making sure it runs safely and shuts down before it gets into deep weeds. A copper heat sink goes a long way to passive heat removal as well in an emgerency situation.
This is like buying a car based on how well it runs without oil in the engine. I suspect my BMW would make for a fantastic video if I tried that too. DON'T DO THAT! I would not pay extra for an engine that would - like using synthetic oil to give an extra two minutes of use.
Buying a CPU that throttles back and paying extra for it -- that might be insurance, but I stopped buying retail boxed CPU's with the three year warr.... It would cost me more to ship an old 400mHz CPU back to Intel than to just replace it these days. I paid $99USD for a 1.4G CPU a couple weeks ago. At that price, these things are practically disposable.
+++ UGUCAUCGUAUUUCU
Thresh's Firing Squad has a review of the Tyan Tiger with dual AMD Duron MPs, which is probably of equal or more interest to us geeks. For those of you who weren't aware, AMD Durons work in multiprocessor mode as well, and they're very, very close to Athlons in terms of performance (and obviously cheaper.)
What's your damage, Heather?
Is the latest Athlon processor overkill for any normal computer user? Yup.
Is there any software currently available that requires this kind of speed? Nope.
Is there any sensible reason to upgrade your CPU? Nope.
Is my rational, analytical mind paying the slightest bit of attention to this argument? Nope.
It's all about the megahertz, baby! In an earlier generation, we were the people tinkering under the hoods of our Fords, trying to get a little more oomph out of a carburetor. Most of us don't need it, most of us have no idea what to do with it, but since when has that ever stopped us? More speed! More storage! More bandwidth! I want more!!!
Good job, AMD. Keep 'em coming.
My id is sneaking up behind my superego with a rock...
Here's the low down on the dual Athlon. It is incredibly fast for any server or workstation application. Of course the app has to be SMP capable which is why your seeing the new KT266A chipset single CPU system beat it out in some apps, but those are only non SMP capable apps. It is apples and oranges. Yes, I would like to see some chipset improvements to the 760MP. The latency is too high. Perhaps the 760MPX will address some of this. I would very much like VIA to commit to their dual Athlon chipset, but they have not as of yet. Another issue is heat. While they do use the cooler running Palomino core, they are still quite hot for say a 1-2U rack. The shrink to .13 micron early next year will eliminate that issue and should hasten adoption by larger computer makers. For the time being though it is a relatively cheap solution for those who need it, and is a blazingly faster web server for those who know how to set it up. Check out my review for more, my site is still up, and we didn't copy anyones site idea.
ignorance is bliss. googlefiberatx.com
I didn't realize that people took press releases as gospel. What's all this crap about a "big speed boost?"
We're talking about mhz increase of less than 10% -- in how many months? It's been over a year since the T-birds were introduced.
Yeah, they have a new core. Whoopee. It's not a dramatically new improvement, and apparently AMD has decided that if its chips, in name, are as fast as P4s, they should cost as much too.
I like AMD stuff, but the Mhz Myth shit hasn't worked for apple, ever, and it won't work for AMD. Apple tried the Mhz Myth stuff back when the ppc601 came out, and despite 6 or 7 years of PR bunko, it's not caught on.
No sig is worth reading.
Anandtech has a good review that compares all the latest p4's and athlon xp's. Check it out here.
To me, this is an apples vs. oranges analogy. On one hand, we have the apples, who are the auto and motorcycle enthusiasts. On the other hand, we have the oranges, or the vast uninformed PC-buying public walking into CompUSA and Circuit City stores. Two completely different species.
Intel's ability to con the public into buying into the MHz game is obhorrent, at best. They manufactured an inferior processor, the P4, basically to outmatch AMD in numbers. Intel knew that AMD wouldn't be able to ramp up their Athlons to the same level within a reasonable amount of time. The P4's inferiority is backed up by the fact that P3s outperform P4s MHz-per-MHz.
I feel that AMD's new effective/relative performance ratings are justified in this case, especially since the numbers are realistic (as opposed to their 486/K5 series or Cyrix's CPUs). If Intel wants to bloat numbers, AMD has to catch up in the marketing game in order to survive in this industry. People are walking into the major retail stores and being convinced by salespeople that the P4 systems are better and just as cheap (only because they bundle inferior components such as nVidia TNT2 graphics cards and generic sound cards to reduce the price) as an Athlon-based system. The regular Joe Blow will see a bigger MHz number and an affordable price, which is the killer combination.
The Linux/hardware enthusiasts are by far a minority in the PC market. Thus, the battlegrounds look ugly to those who are more informed, but I'm sure they look even worse within the buildings of Intel and AMD. It's a dog-eat-dog world.
Hammer is going to be unveiled today as well.
-- "Ever wonder why the SAME PEOPLE make up ALL the conspiracy theories?"
A console is a very different environment. You can tune exactly for the hardware because there will be no variances.
I agree in principle, but that's not it. It isn't polycount either, as someone else said. At the moment, the average PS2 game has more polygons than the average PC game (that's because if you assume hardware T&L on the PC then you have a severely limited market; lots and lots of mass market PCs still ship with the equivalent of a Voodoo 1 or worse, go to Dell's site if you don't believe me).
I'm talking about much larger issues. For example, on the PC you come up with a file format for something, then just keep using it because it works. With a little work, it often turns out that a 20MB file of world geometry can be knocked down to 5MB, just because there's so much garbage in there and no one ever thought about remove it. Or maybe there are thousands of keyframes of animation that make no visual difference and can be removed. Or some trifling module allocates 8M at load time and keeps it around, even though it isn't actually used. Or maybe there's poor collision detection code that does way too much work and could be made to run 4x faster. These kinds of things are _common_. I'm a game developer; I've been there.
Is Intel paying you directly or thur alternative sources? ;) Yeah, it was fantastic at E3 when they couldn't show me a Black and White demo on the P4 when it kept blue screening. That was fantastic! Anyway, go ahead, pay twice as much, get 2/3 of the performance in a server. I'll happily serve a million plus sessions a month on my little Athlon 900.
ignorance is bliss. googlefiberatx.com
The only thing that matters is as follows (in rank order):
1. Bandwidth - face it, email and the web are king. Unless you're a gamer.
2. Video Card - if you're a gamer, you're better off spending your money on this and making sure it has tons of cache.
3. Sound Card - if you're a gamer, you're better off spending the rest of your money on this. The rest of us don't care, so skip this.
4. Memory - more, more, more. Yes, even more.
5. Bus speed - more channel so those CPUs can actually send more data.
6. Hard disk - you really should have more RAM, but once that's crammed, get better seek and access times here.
6. Chip speed - WAY DOWN HERE! - yes, if you maxed on all the above, then you MIGHT notice the difference between a 1GHz and 1.8GHz system. Otherwise, unless you're a graphics artist, YOU SHOULDN'T WASTE YOUR MONEY!
Naturally, when people review systems, they compare older systems with slower bus speed, less RAM, slower HD, and cheaper cards to new systems with faster H/W. Buy the motherboard and cards yourself and pop in a slower chip and spend the extra money on RAM - you will get way more bang for your buck that way.
Aside - I own AMD shares, so sure, go buy these speed demons! But don't do it because you have to, do it because you know you just like BIG NUMBERS.
--- Will in Seattle - What are you doing to fight the War?
Engineers do this all the time. I am a mechanical engineer, and the analysis software packages I use to assist in design and analysis of my creations does just this: solve massive systems of equations.
This is very common and very useful.
Also, if I had a PC with 100 times the memory and speed, I could still bring it to its knees. As it is, I have to simplify and granulate my models to make them fit the computing power I have.
How do you think they predict the weather? Design cars and planes? Do thermal analysis? Do vibration analysis? Do electromagnetic analysis? Do displacement/stress analysis? Do computational fluid dynamics? Do transient analysis of all the above?
Every single AMD CPU from Duron to Athlon MP supports SMP. There are reports that AMD will simply try to disable SMP on the XP line. But the core architecture is SMP-capable.
___
If you think big enough, you'll never have to do it.
I have an Dual CPU 800 (Intel).. the CPU's won't shut the system down if the heat gets to hot, but the motherboard will.
.. later discovered that the motherboard was sinking ship after 75deg celc. from either CPU.
the case fan died on my box , and because I run dual-seti at night, the machine heated up and started beeping.. woke up me up, but found that the box shut itself off
The new Athlons (XP and MP) have thermal sensors on board according to AMD's site. I still can't find any information indicating whether/how they actually use these though.
I was under the impression that AMD changed to this new naming scheme to avoid the public's concentration on MHz. Why, then, do I read an article on Slashdot in which AMD's new naming scheme is broken down into MHz equivalents?
Seriously, I think we all agree here that AMD is making a bold and necessary move to diminish the importance of MHz. Unless we follow suit and stop using MHz as our measure of performance, the public will never catch on. I think the importance of attaching a "model number" to a chip name is that we will eventually forget about MHz altogether and focus on pure chip performance. Let's start that now.
The Mhz equivalents for each of these new processors had no place in this article.
Otto-matic