Anand Reviews Athlon 64 FX-53

trickofperspective writes "Anandtech has a review of AMD's latest processor, the Athlon 64 FX-53. Long story short -- the FX-53 is a "very solid processor," but you'd be better off waiting a couple months for Socket 939."

Comments

Get some PRIORITIES

[Anonymous Coward]

It's mere hours after the single MOST HEROIC EFFORT BY FIREFIGHTERS IN US HISTORY and you people are talking about television???? MY GOD, people, GET SOME PRIORITIES!!!!!!

first post

[Anonymous Coward]

maybe this time???

TOASTY

[Anonymous Coward]

toaster,toaster toaser, do you have toast in you yet i think
so!!!!!!!!!!!!!!!!!!!Im not a toaster!!!!!!!!!!And one more
thing........YOUR A TOASER!!!!!!!!!!!!!! AND A COOKIE WITH MILK SOAGE
MILK!!!!!!!!!!AND A BUTT WITH POOP IN IT!!!!!!!!!!!!!!!!

Anandtech

[Anonymous Coward]

Anandtech has a review of AMD's latest processor, the Athlon 64 FX-53. Long story short -- the FX-53 is a "very solid processor," but you'd be better off waiting a couple months for Socket 939. Does it sound cool, you might ask?...do you wonder how long before someone cries about civil liberties? --- Does this seem like an idea that would suck? Is it good, or is it whack?

Re:Anandtech

[abscondment]

You can always get a better piece of technology by waiting just a little longer--the only real reason to wait then is if the standard is going to change. If you buy this current chip, it'll be the best you can get right now. When they change to socket 939, however, you'll be stuck with what you've got--no upgrade for you!

It's always best to buy right when the standard changes, so that you have the ability to upgrade later if you want to. If you buy right before the change, you guarantee having to purchase a whole bunch of new stuff for the next upgrade.

Re:Anandtech

[yppiz]

A decent motherboard costs $100 or less. Is there anything else I would have to replace, besides the CPU, if I wanted to upgrade from the current chipset?

If not, I don't see why I would want to wait for the next chipset.

--Pat / zippy@cs.brandeis.edu

Re:Anandtech

[ivan256]

I can't find anybody anywhere that is saying you won't be able to stick a 939 pin Athlon in an 940 pin socket. Plus, Opteron chips will still use the 940 pin platform. You'll be able to upgrade still if you buy now. The only downside is that you'll have to use Registered ECC memory.

Re:Anandtech

[tiger99]

Personally I would wait a year, so they are cheaper. Certainly socket 939 is a "must", but I always take the view that where computers are concerned, if you wait till the last possible moment before you really must have something, you save a lot of money.

I definitely don't need one right now, but in a year, when it is a mainstream product, I will find some excuse to persuade myself to buy one. By that time the OS (Linux of course) will have been very well debugged.

I wonder when the move to 128-bit will come?

Re:Anandtech

[MindStalker]

At $733 for the processor and 200-300 dollors for the motherboard, I really don't think the cost of upgrading the motherboard should really be what you are thinking about, afterall I really wouldn't want to put my $733 processor out of commission, Id rather keep it running as a backup computer or doing some other job. Its not exactly something you throw out and replace.

Re:Anandtech

[prockcore]

Is there anything else I would have to replace, besides the CPU, if I wanted to upgrade from the current chipset?

Quite possibly your ram.

Re:Anandtech

[Rufus211]

you can be sure that 939 cpu's won't fit in 940 sockets. its not the number of pins that matters, it's the electrical interface. also 939 is going to use unbuffered ram whil 940 uses buffered

that said, amd has said they will continue to produce 940 socket cpus through the year.

Re:Anandtech

[sumdumass]

I have seen speculations that the socket 939 will offer perfoermance gains too. The evidence is kinda just speculation but i think based on what i read it will.

You are wise to point this socket change out. But I am already planing on getting one just so I can have bragging rights for a couple of months. It wouldn't bother me to pay extra later (maybe after pci express is out) to take advantage of better features. But I am not most people. I still agree with you though.

If I knew howto mod you up (informative) or was sure if I had points I would. i'll get this figured out one day

Re:Anandtech

[mczak]

Unfortunately the pin layout will be different - the whole point of socket 939 is to make it possible to build 4-layer boards, which supposedly requires a new pin layout - it seems to be impossible with the current socket 940 chips.

Re:Anandtech

[rootus-rootus]

Actually, there is always the Opteron 1XX processors, which is the same thing, and that line will be socket 940 for quite a while.

Re:Anandtech

[Anonymous Coward]

Actually, I do nearly the opposite. This summer I'll be purchasing one of the 'obsolete' motherboards with one of these 'obsolete' chips (or the FX51 depending on how thrifty I'm feeling.). Then I get the following:

- Marked down prices from vendors clearing their old inventory.
- Mature technology without the usual sharp edges associated with cutting edge technology.
- Middle to high end performance.

Of course, then I'm stuck with this cpu for the next few years. That is, of course, until the newest, whizbang socket 1039's hit the market and I pick up a socket 939 mobo priced to move ;-)

BTW, Thank you for encouraging everyone else to adopt the newest techology as soon as it hits the streets. _Someone_ has to beta test the hardware after all...

Re:Anandtech

[Junks Jerzey]

Personally I would wait a year, so they are cheaper. Certainly socket 939 is a "must", but I always take the view that where computers are concerned, if you wait till the last possible moment before you really must have something, you save a lot of money.

And if you had done that for the 3GHz P4, then you'd be buying one right about now, when the prices have finally dropped to mainstream prices. But then you'd see some fancy new processor on the horizon, like the latest Athlon 64, and decide to wait for that one to become cheaper...

Re:Anandtech

[abscondment]

and if you're buying a bleeding edge upgrade, you probably won't be getting the mb for $100

Re:Anandtech

[athakur999]

Besides socket 939, there's a few other goodies that are supposedly going to be in wide use this year, namely BTX and PCI Express.

For those of us like me that like the keep our hardware for a good long time, now is not a good time to buy since obsolescence is looming on several fronts.

Re:Anandtech

[mrdavidk]

BTW, Thank you for encouraging everyone else to adopt the newest techology as soon as it hits the streets. _Someone_ has to beta test the hardware after all...
So true. Granted, if I had the money I'd be updating every few months but I don't. My current system: Athlon 1600, 512 MB RAM, CD-R/RW (Anyone remember when they just sold stand alone CD-R drivers? lol), DVD-ROM,40 GB HD (Seems small compared today's standards), and GeForce2 video/sound. Last system before that: 266MHz PII with MMX, 128 MB RAM, CD-R/RW, DVD-ROM, 5 GB HD (!), Ati Rage Fury Pro and floopy drive ;). That system lasted me a good 5 years. I'll be ready to jump on the 64-bit train later this year, when the hardware is common place :).

Re:Anandtech

[TheGratefulNet]

power supply, for sure!

Re:Anandtech

[niko9]

You would have to get new RAM. Socket 939 lets you use the unbuffered DDR RAM you probably already have.

Re:Anandtech

[Retric]

I see little point to upgrading the CPU on a mother bord. Then again if you were plaining on buying a CPU for a grand then you may just be the kind of person who buy's a new CPU every 6 months but in that case ya may as well buy a new mobo becouse the difrence in the top end CPU on an old mobo and the 2nd cpu is normaly enof to buy a new mainbord which tends to give you better overall preformance for the price.

IMO it's best to buy CPU, mainbord, and RAM then in a double the ram year 1 or 2.

Repeat as needed.

Re:Anandtech

[PingPongBoy]

If your trusty P166 isn't too slow.

The biggest thing I noticed besides processing speed is hard drive speed. Some people I know have some 1700 MHz systems they bought about 3 years ago. I see the disk drives appear a lot faster now, and it makes a big difference to upgrade later. I bought a computer a year later, and the disks are a lot faster. Now, I don't really need an extra 500 MHz, although I can use an extra 50000 MHz.

Re:Anandtech

[Rich Klein]

Personally, I'd wait until I could afford an Athlon 64. :P

Re:Anandtech

[tiger99]

Exactly! I did once buy a 486DX33, when it was very leading edge. Mine had a 340MB SCSI HDD, 17 inch monitor, lots and lots of RAM (16MB!), etc, because I wanted to run CAD and circuit simulation. It cost a huge amount of money and was barely adequate, although very leading edge at the time, thanks to the bloatware of the day, the absoluitely new Win 3.1. DR-DOS improved it noticeably, especially where use of the bottom 640Kb was concerned.

Hardware has improved faster than incompetents like M$ can add bloat and slow down the software, so although a leading-edge machine would be nice, it is by no means essential now for any serious use that I can imagine, and I don't see why games players get excited about getting more than 70 frames per second, your eye can't tell the difference. For larger problems, a Beowulf cluster might be the cost-effective way to go.

The P4 never has been on my agenda, I have used Athlons for some time now, there is no point in paying top money for a product that is not any better.

I do tend to upgrade my PCs throughout their life, it might not save anything overall, but it spreads the cost nicely. The 64 will have to be entirely new of course, I don't have a case and PSU that will become spare in the near future, and it is time to move to serial ATA for the disks (I always fit two, at least, for efficiency, swap on the one with the least load).

I don't bother with state of the art graphics cards either, just a well-supported model of a year to 18 months ago is fine, because serious software tends to not use 3D textures and bump mapping and all that, it mainly spends its time drawing simple 2D graphs and diagrams. Opening and closing windows (of the X sort, M$ is not on my current best PC, and will not be on the next one either) is about the heaviest graphics load that I envisage. OK, maybe I will watch the odd DVD, even that will not push one of last year's cards very hard.

But when it is cost-effective to make the jump to 64 bit, I will enjoy doing so, although I expect that my heaviest application (SPICE) will have to be recompiled to take advantage of the upgrade.

This will be a major jump in performance, there will not be much incentive for a big upgrade again until 128 bit, or something radically different, without the old x86 baggage, comes along.

I expect that my biggest purchases in the next year or two will in fact not be PCs, but LCD monitors. Having 1600*1200 on my laptop, I would like it on the main PCs, but probably about 21 inch size, that will as of now cost serious money. If I can scrape up the money for an even higher resolution I will do so. My first monitor, a NEC 17 inch, which runs at 1280*1024, original cost about 700 pounds (UK)is still going strong, but is big, heavy and ugly by todays standard.

The point is that if you buy leading edge, it loses value very fast indeed, if you wait a bit till the price has stabilised at a reasonable level, you lose less, if you don't immediately need the performance.

Addendum

[Anonymous Coward]

Tom's review is here.

Re:Addendum

[Loki_1929]

"Tom's review is here."

Why bother? Is this just to show how well you can rig benchmarks with enough driver vetting and settings tweaking? All you have to do to get the Tom's results is to take every other reviewer's work, reduce the AMD chips' performance a bit, add to the Intel chips' performance a bit, and then create a conclusion to support Intel's total dominance, power, and magnificance, regardless of what the (tweaked/rigged) benchmarks show.

Re:Addendum

[coughman]

Extremetech also has a review too. They have a more negative opinion. link

Re:Addendum

[webtre]

Here are some more benchmarks

AMDzone

AnandTech

XbitLabs

Ace Hardware

There are even more at AMDZones main page.

Re:Addendum

[Anonymous Coward]

But Toms review

is a little

more drawn out

Re:Addendum

[bhtooefr]

Nice FX-51 reviews, but we want FX-53 reviews.

Re:Addendum

[Anonymous Coward]

To quote homer : "It's funny because it's true!"

Re:Addendum

[niko9]

But

/flash/Gaaaaaaaaaaaaaaaazzzzzzzzzillion Fasssssssssssssssssssst! ASUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUS!!!! BUY NOW! /flash/

Toms

/flash/ AOPEN! New! ILLUMINATED PC SCREWS! So bright gaunrantee abrasion corneal. Super Bright many Watt!/flash/

review

/flash/ SOLTEK! /silver_hydrocephalus_umanoid_with_tumor_size_of_m elon.gif/ Speed Insane!/flash/

is

/flash/CENTRINNNNNNNNNNNNNNNNNNNNO! Post on slashdot while making a poop! Look ma No hands! INTEL INTEL INTEL!/flash/ /flash/ Do more! Email at 4.0Ghz only with INTEL! Don't be left behind, ASK TOM!!/flash/

little

/flash/Tom's LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOW price finder /the_hammer.gif/ /flash/

more

/flash/ New from Coolermaster, coolermaster refrigarator! With BLUE LED! Match case in color. To keep food cold!!!!!!/flash/

drawn out

A little?

Re:Addendum

[eggstasy]

Shhhh! Look, you're new here, so let me give you some pointers. This is slashdot. We love AMD here. Everything from AMD is Holy. There are no negative reviews regarding AMD, because there is nothing bad to say about AMD. It's all a bunch of Intel-sponsored FUD.
AMD is winning this war: Intel employees are committing suicide by the hundreds at the gates of AMD's HQ.

Re:Addendum

[Penguin Follower]

Offtopic? Personally I found that to be funny, and would have moderated as such if I'd had mod points...

Re:Addendum

[Anonymous Coward]

FYI - Tom's Hardware also purchased years ago and is very biased in their reviews.

Re:Addendum

[cuban321]

Actually the more I use Intel workstations and laptops the more I dislike AMD.

My new centrino laptop @ 1.6ghz blows my 2ghz 333FSB desktop away.

Re:Addendum

[Anonymous Coward]

Tom forgot to take a screwdriver and pry away the headspreader on the FX-53 during operation! Has he been bought by AMD? This review is worthless without the usual THG Real World Testing! ;)

Anand DIDN'T review it

[Anonymous Coward]

Derek Wilson did for AnandTech. Anand is a person, AnandTech is a site.

Re:Anand DIDN'T review it

[trickofperspective]

Sorry, my bad. Won't happen again.

-Trick

Re:Anand DIDN'T review it

[Anonymous Coward]

It BETTER not. I almost had a heart attack from such a glaring and horrible mistake.

It matters so fucking much! AAARRRRRGGGGGHHHHHuuugggghhh.....

the FX-53 is a "very solid processor"

[lofoforabr]

I am yet to see a very liquid (or even a gasous) processor. Wouldn't it be cool? Hey, if it would be cool, it would solve the thermal problem inherent to solid processors. Why not make them?

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

Who modded this interesting? Sheesh.

Re:the FX-53 is a "very solid processor"

[DR SoB]

Back in the 70's IBM mainframes were liquid cooled, the problem was, they often sprung leaks...

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

isn't air also considered a gas ? plus i dont fancy water running very close to my circuit boards...

Re:the FX-53 is a "very solid processor"

[afidel]

The Cray-2 supercomputer was also liquid cooled. They used a fluid called flourinert which is electrically non-conductive but a good thermal conductor. Flourinert was origionally developed as an artificial plasma substitute for heart surgery. It is also insanely expensive, around $500/gallon!

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

Quantum cumputers would fall into the very liquid category. So far they can solve simple arithmatic by mixing two liquids. I dont know how they put in inputs or read outputs though :P

Re:the FX-53 is a "very solid processor"

[nomadic]

It is also insanely expensive, around $500/gallon!

Wow, that's almost as much as a comparable amount of Starbuck's coffee...

Re:the FX-53 is a "very solid processor"

[dubiousdave]

I am yet to see a very liquid (or even a gasous) processor.

Clearly, you've never run your Athlon without its cooler in place.

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

The newest Intel chips run hotter than the newest AMD chips. Time to update your jokes.

Re:the FX-53 is a "very solid processor"

[bhtooefr]

It's THG, but... http://www20.tomshardware.com/cpu/20010917/

BTW, that only applies to non-64-bit Athlons/Durons. Yes, Intel is hotter than AMD. No, Intel processors don't hit 370 C when their heatsinks are removed, as the Pentium III shuts down, and the Pentium 4 slows down. The Athlon MP/XP had thermal overload protection in their spec, but some boards don't include it.

Re:the FX-53 is a "very solid processor"

[flewp]

But I bet it tastes better...

Re:the FX-53 is a "very solid processor"

[dubiousdave]

Not until we have an article on Intel chips. I wouldn't want to make a useless, smartass comment that's also offtopic. Besides, I love my smoking hot Athlons.

Re:the FX-53 is a "very solid processor"

[NanoGator]

"The newest Intel chips run hotter than the newest AMD chips. Time to update your jokes."

Okay. Hey! My Athlon self destructed because it doesn't clock itself down when it overheats!

Hmm well I need to refine that joke a bit.

Re:the FX-53 is a "very solid processor"

[juhaz]

See the weird line of numbers in that url? It's usually called "date". Magically you can deduce from that the information that the article in guestion is about three and half years old.

If some boards don't include something that's required by the specs, well, that's problem of the greedy bastards manufacturing such crap and the fools tricked into buying them.

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

Anyone know if flourinert is the same as galden fluid a.k.a. perfluoropolyether or propene hexafluoro oxidized ploymerized?

Are they in the same family?

Re:the FX-53 is a "very solid processor"

[puhuri]

The Cray X1 uses spray evaporative cooling. Dielectric coolant is sprayed over PCBs and then coolant evaporates. It makes possible to use something like 65 W/cm power densities.

One needs, however, hermetic chasis, so your average PC box is not sufficient.

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

Coincidentally, I can solve many of my personal problems by mixing two liquids - Jack and Coke.

Re:the FX-53 is a "very solid processor"

[DR SoB]

Hahaha, mod me -2 offtopic and someone who mentions a Cray get's +5 informative? What's wrong with this picture.. Well, I get the picture, time to start surfing a different website.. L8r /.

Re:the FX-53 is a "very solid processor"

[ZosX]

Actually one gallon is 128 ounces. An average cup of starbucks coffee is 20 ounces. So 6.4 cups per gallon. Local starbucks is $1.73 for a large so.... It is only $11.02 per gallon. That is roughly 50 times cheaper. I know I have too much time on my hands.

Re:the FX-53 is a "very solid processor"

[afidel]

They are both perfluorinated liquids. Fluorinert is the 3M brand name for their line and Galden is the brand name for Ausimont's.

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

The fact that you needed to explain that is a very good argument against imperial mesurement. :-)

Re:the FX-53 is a "very solid processor"

[niko9]

Yeah, that's a latte money.

I know, place a bright orange helmet on my cranium and put me on the short bus.

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

>>I can solve many of my personal problems

What's her name? :)

Re:the FX-53 is a "very solid processor"

[Jeremy Erwin]

Flourinert was origionally developed as an artificial plasma substitute for heart surgery.

Um, I'm going to have to ask for your source. After all, immersion cooling was an accepted, if not exactly prevalent method of thermal transfer. Some early IBM computers were oil cooled. PCBs were used to cool electrical transformers. 3M's flourinert was most likely marketed for a similar purpose.

Yes, mammals can breathe oxygenated perfluorocarbon fluid, but the experience is likely to fatal unless the perfluorocarbon has the right vapor pressure and is highly purified. In any case, the medical function of fluorocarbons is oxygen tranport, not plasma replacement.

The CRAY2 was cooled by FC74. To my knowledge, this particular chemical was not used for medical applications.

Re:the FX-53 is a "very solid processor"

[silicon not in the v]

That's interesting. I only know of a different use for it. I work at a semiconductor company (not AMD or Intel, though), and we use it in the lab when debugging failing parts in R&D development. We can test chips while they are still on the wafer, and if they have shorts on them, we can use flourinert to help locate the short. We squirt some flourinert on it while it's operating. The shorted area generates a lot heat, so the fluorinert will start boiling locally and generating a lot of tiny bubbles right where the short is.

Re:the FX-53 is a "very solid processor"

[HungWeiLo]

I thought that was funny, but reading it once more - I realized that it was also true!

What's a typical cup of, say, espresso there? $3-$4?

According to their website, a cup of espresso is 1 fl. oz. There's 128 fl. oz. in a gallon, so:

$3 * 128 = $384 / gallon (conservatively), or
$4 * 128 = $512 / gallon (more likely)

Re:the FX-53 is a "very solid processor"

[Jeremy Erwin]

The Cray 2 used FC 74 fluorinert, which is no longer produced by 3m. It is apparently identical to FC77, which is described on the msds as "PERFLUORO COMPOUNDS, (PRIMARILY COMPOUNDS WITH 8 CARBONS)". Nice and vague.

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

I think he was talking about Starbucks espresso. A single espresso is about 2 ounces of coffee and costs about $1.50. So, that's $.75/ounce, which is 0.75*128 = $96/gallon for Starbucks coffee.

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

And I'm told that the original flourinart would turn into Mustard gas if it was heated.

Supposedly, you weren't even supposed to smoke a cigarette after handling the stuff, on the off-chance that you had some on your fingertips, and it got onto the cig while you were smoking it.

I kept waiting for the headline about some crazy overclocker gassing his neighborhood, but fortunately it never happened.

The newest stuff supposedly doesn't have this quality; but it was an issue still during the early 90's.

Re:the FX-53 is a "very solid processor"

[DaSpudMan]

I work in agriculture where $500/gallon chemicals are not uncommon. It's only a few that cost to be sure but $200-$250/gallon chemicals are relatively common. Most applications for these are measured by the oz/acre.

Re:the FX-53 is a "very solid processor"

[Shinmizu]

The Wookie's measurement system doesn't make much more sense.

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

The newest Intel chips will clock themselves down if you remove their heatsinks, without any OS support. AMD still aren't there yet. (the chips won't fry, but they will lock up hard, reset required)

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

Go back to your parents' basement, zonk.

Re:the FX-53 is a "very solid processor"

[spun]

No, it's a design problem. Thermal overload protection should be based completely on the chip, not off on the motherboard! Intel does it that way, AMD didn't. Don't know about now. But if you leave it up to the motherboard to protect the CPU from overheating, that is just asking to let the magic smoke out, IMHO.

If some CPUs don't protect themselves from permanent failure due to thermal overload, well, that's problem of the greedy bastards manufacturing such crap and the fools tricked into buying them.

Re:the FX-53 is a "very solid processor"

[Anonymous Coward]

You should check out DNA computing. Formulate your problem in terms of a different strands of DNA, replicate them a zillion times, put it in a beaker and stir ---> wallah, massively parallel computing in a jar.

Re:the FX-53 is a "very solid processor"

[afidel]

Cray's next machine, uncreatively called the "Cray-2,"solved the cooling/plumbing problem another way: the boards themselves were swimming in a non-conducting liquid called "Fluorinert," a blood plasma substitute used in surgery that happens to have the right thermal, mechanical, and electrical properties. link

I know this isn't the only place I've seen this as well. In fact back in 1996 when I attended a computer camp at The University of Michigan and we were working on the Cray YMP-80 the prof gave us a little history of their supercomputer center and mentioned this fact about their Cray 2.

Re:the FX-53 is a "very solid processor"

[juhaz]

Thermal overload protection should be based completely on the chip, not off on the motherboard!

It's motherboards job to make sure that every thing required for the operation of the chip is right, why should temperature be any different than rest of the bunch? You'll end up integrating the whole computer into CPU if you start down THAT road.

But if you leave it up to the motherboard to protect the CPU from overheating, that is just asking to let the magic smoke out, IMHO.

Well, it's about equal to letting motherboard worry about CPU getting the right voltage, both result in magic smoke if done wrong, yet motherboards have done that pretty well forever. And will continue to do so.

Re:the FX-53 is a "very solid processor"

[Jeremy Erwin]

Yeah, it's a pernicious little meme.

Cray used FC-74 as a coolant. It's since been replaced by FC-77 which is mostly perfluoroctane. 3m implies in its msds that FC77 is a bit of a hodgepodge of various perfluorocarbons. The FDA frowns on medical use of such oleos, and will often demand that each isomer be tested separately...

There are some citations for perfluoroctane in pubmed, but most are of animal studies. The "blood plasma substitute" you're probably thinking of is either Fluosol (perfluorodecalin) or Oxygent (perfluorooctyl bromide).

Re:the FX-53 is a "very solid processor"

[ChaoticLimbs]

Read his post- he's talking about Espresso, which is served in teensy weensy shot glasses.

Re:the FX-53 is a "very solid processor"

[rnd()]

Nope, espresso is typically only around $1.30 for a single, so it's not quite as bad as you suggest, but I still feel foolish every time I buy one.

Re:the FX-53 is a "very solid processor"

[bhtooefr]

TOP is now on the K8 chips, in the "cut power to the chip immediately" form.

Ladies & Gentleman, We Got HIM!!!!

[Anonymous Coward]

Suck it, you terrorist-sympathizing bitches!!!

The lesson here, as always: don't fuck with the US or you're gonna get a good ass whooping!!

Hell yea!!!

Thank God!

[Anonymous Coward]

Taco is back. Michael was getting tired (both his physical condition and our patience with him).

Thanks Taco! g2g and nab one of these hot new, sexy slabs of silicon yumminess from AMD now!

Re:Thank God!

[Anonymous Coward]

Jeez, dude.

Don't kiss Taco's ass so hard, you might strain your lips...

Re:Thank God!

[Anonymous Coward]

Yeah, his ass is rather massive.

Or is it Cowboyneal's? Or both!?!

God damn it, I'm so confused now.

Explaining the difference...

[BJZQ8]

I had a conversation with a neophyte that was looking to "build their own computer" yesterday...He was obsessed with the idea that megahertz=performance...I tried to tell him that an FX-51, 52, or 53 would be a much better performer, all around, than any Pentium 4, "Extreme Gaming Edition" (as he put it) or not...but in the end, he was swayed by things like "Hyperthreading" and "Netburst"...AMD is having a hard time fighting against Megahurtz Madness and Buzzword Bufoonery.

Re:Explaining the difference...

[snarkh]

But surely AMD's HyperTransport technology with 3DNow! is a worthy contender.

Re:Explaining the difference...

[Anonymous Coward]

Well it's a good thing he didn't believe you because it's simply not true. The top end Athlons beat the top end Pentiums for gaming. The top end Pentiums beat the top end Athlons for most other applications. The top end from either line is ridiculously expensive.

Re:Explaining the difference...

[Anonymous Coward]

The only advantage the top end Pentiums have are where Intel's people have gone in an done specific Intel optimizations. Since the 64-bit chips support SSE2, they also support many of these optimizations.

Re:Explaining the difference...

[74nova]

some people are also scared of amd being cheap. my brother was convinced that there were some things his amd 600 wasnt compatible with just because it was an amd. i tried to convice him otherwise and, well, he now owns a 3.whatever dell with hyperthreading.

slightly OT, but 'Buzzword Bufoonery' is, in itself, a fantastic sort of anti-buzzword buzzword that i shall use from now on. a fictitious cookie for you if you invented that.

Re:Explaining the difference...

[bhtooefr]

Note that "Extreme Gaming Edition" was what this person called the P4 EE. He was going to be gaming on it, most likely.

Re:Explaining the difference...

[AnotherBlackHat]

I've often thought that processors should take the clock and divide it by 16 as the first step, just so the Mhrzt can be 16 times higher.

Later they could remove the divide by 16 and claim to have an internal "clock multiplier" and charge extra for the part.

-- this is not a .sig

Re:Explaining the difference...

[Awptimus Prime]

some people are also scared of amd being cheap.

A lot of people avoid AMD because of the heat, too. I will probably not buy another AMD after the XP2100 I got last year. With a lian-li case, arctic silver compound, an SLK800 heatsink and 7 80mm fans it still runs at 58C. Overclock it just a teenie bit and it will jump to 70C within minutes under load.

It's very loud, not all that fast and consumes more power than most datacenter servers. I saved a whopping hundred bucks by not buying a 2.4Ghz P4 at the time.

Re:Explaining the difference...

[sumdumass]

This is more common than not. AMD has to place the blame partually on itself. IT bombed at a couple of processors back when going to the k6 and the k6 2 -3 prcessors lack some instruction sets that are just now beguinning to be used in games and such.

The k7 line upto the present is a completly better processor in my opinion. But people still hold onto the past especially when they don't know about the tech involved with it.

I have a customer that even thinks that if you use an AMD proccesor on a network with an Intel, you will have problems becuase it is AMD. I laugh when something goes wrong and I place my AMD notbook on the network to figure out what went bad.

Re:Explaining the difference...

[shepd]

>With a lian-li case, arctic silver compound, an SLK800 heatsink and 7 80mm fans it still runs at 58C.

Wow! After reading that I will never buy arctic silver compound, an SLK800 heatsink, or a lian-li case ever again. Using the stock fan and a $25 case (no fans), and the world's most garbage PSU known to man my Athlon 2000 runs at about 45 degrees C. :-)

Re:Explaining the difference...

[Oliver Wendell Jones]

My guess is something is wrong somewhere.

I'm running an AMD XP 2500+ at stock speeds and voltages with the OEM heatsink and some cheap, white thermal compound and my system sits as 119F (48C) and under heavy load (say many hours of hectic UT2K4) it gets as high as 130F (54C). That's the reading I get on the front of my case from a thermal probe touching the side of the raised center part of the top of the chip.

It's also very quiet, even with three case fans in it.

Re:Explaining the difference...

[Too Much Noise]

same here in Windows ... while in Linux the average temperature is just below 40 deg. C and it jumps to about 55 under very heavy use. Ain't Linux cool ^_^

Re:Explaining the difference...

[hackstraw]

He was obsessed with the idea that megahertz=performance

Is a couple of months of having "the fastest PC made" really worth obsessing about? (Hint: there will always be a faster one coming out)

I would like to see PCs that are upgradeable. How hard is it to make a MB that is "overspeced" so that you can stick the fastest processor and ram in it for a couple of years?

Re:Explaining the difference...

[MikeCapone]

A lot of people avoid AMD because of the heat, too. I will probably not buy another AMD after the XP2100 I got last year.

Well, I guess that you should avoid Intel too, since they have been running hotter than AMD for a while (especially the Athlon 64's with they frequence throttling tech) and I don't think your next purchase will be downgrading.

Re:Explaining the difference...

[BJZQ8]

For him, it was the ONLY criteria for his computer...that it be "the fastest" available. I have stocked a school with 1 to 1.4GHz machines, and they work fine...most of his time is spent browsing the internet and sending e-mails anyway...but not buying "the fastest" makes him feel dirty, at least that's the feeling I get from what he's said to me. That's the thing that drives me nuts about people like that...they believe the marketing hype over reality.

Re:Explaining the difference...

[Just Some Guy]

arctic silver compound

Just for the record, you do know that it's a very good insulator, and that if you have more than just a bare film of the stuff then you've basically wrapped your CPU in a sweater, right?

Perhaps you do, but it seems like Arctic Silver is the computing equivalent of low-profile tires: if a little bit is good, a whole lot must be better! It's almost the computing equivalent of a whaletail on a Sentra, although some people do actually use it correctly and see some benefit.

Re:Explaining the difference...

[AKnightCowboy]

some people are also scared of amd being cheap. my brother was convinced that there were some things his amd 600 wasnt compatible with just because it was an amd. i tried to convice him otherwise and, well, he now owns a 3.whatever dell with hyperthreading.

Your brother is smart. It's wise to avoid AMD unless you don't care about compatibility. For example, an AMD Athlon XP 3200+ CPU in an Asus A7N8X-Deluxe motherboard with an Nforce2 chipset is absolutely horrible under Linux. I have to practically cripple the thing by shutting off APIC before I can even get it running for longer than a few minutes without hard locking up. Personally I feel this is a problem with Linux since I have the same board running Windows XP and it works flawlessly, but suffice it to say, there is definitely an incompability somewhere. If I had gotten a P4 system instead I'm assured it would've been working 100%.

Re:Explaining the difference...

[Tokerat]

If I where buying a new computer, I'd but the absolute fastest, most maxed out machine I could, but only because it's such a rare occurance that I have the money to make such a purchase, I'd need my machine to last me a few years.

Re:Explaining the difference...

[Anonymous Coward]

My Athlon 3200+ maxes out at 50c under full load on a warm day, using Artic Silver 5 with a Vantec Aeroflow (which is relatively quiet.) I'm quite pleased with the temperature.

Re:Explaining the difference...

[selfabuse]

How hard is it to make a MB that is "overspeced" so that you can stick the fastest processor and ram in it for a couple of years?

Pretty hard.. technology changes, and without knowing what's going to happen several years down the line, how can you possibly build a motherboard to support stuff that isn't out yet?

Re:Explaining the difference...

[BJZQ8]

Yup, I thought it up...I needed something with alliteration to complement Megahurtz Madness...so Buzzword Bufoonery sorta fit. I'm almost proud of myself. You can use it as long as you pay my $699 fee! Hahaha! j/k I'm releasing it under open source

Re:Explaining the difference...

[Anonymous Coward]

and the fact that 64 bit is an ABSOLUTE WASTE for him right now.

you can laugh at his idioticy in 12 months when the first 64 bit games show up... but right now, you cant touch my 2.8GhzE P4 processor machine with your FX-53.. not in dollars -> performance ratio...

and only fools buy the high end+high priced stuff.

so if your friend was not a moron and bough the 3.4E P-4 processoe for the ream-me in the rear price he made a smart decision...

I'm waiting 24 months befoer I'll tocuh any 64 bit platform... let's see this stuff mature first and become affordable. (Plus pci express will be there to kick the crap out of the mess that is today's expansion busses.)

buy cheap now.... get cool later.

Re:Explaining the difference...

[Awptimus Prime]

Hehe. Your post brings back memories of watching a computer noobie roommate I had try to build his first game system. He used about a 1/4 tube of the stuff before I came out and yelled at him.

Typically, I will scoop a wee little bit on the fat end of a tootpick and work it into a very fine film on the CPU core. I've been building PCs since the mid-80's, so I've had plenty of time to learn that slapping things together will occasionally result in disaster.

It's simply the particular chip I have. If I swap my CPU out with any other XP chip, they run much cooler than this one. It's been a chuckle reading other people's responses where they auto-magically assume I'm some retard who doesn't know anything, when I've been dealing with hardware since before most of them were born. :-)

The next system is going to be water chilled, so I don't care much about next gen Intels and AMDs overheating. My inital post was just to bring up another reason why folks are shy about AMD.

Re:Explaining the difference...

[jafuser]

I have an XP2100 as well. I found out that they were the last of a series of chips before the die size was shrunk, and therefore explains why they run so hot compared to faster Athlon XP chips.

Also, the fan that came with the XP2100 was not adequate. The fan I replaced it with sounds like a jet engine, but at least it keeps it at a reasonable temperature.

I can't close the door to the room my computer is in, or the temperature rises by 3-5 degrees. =P

Re:Explaining the difference...

[Anonymous Coward]

Ever consider a refridgerated inert gas like xeon or nitrogen? You can avoid the possible disaster of a burst pipe, and also continue using the fans inside your case.
And also if you keep it cold enough, a computer case with FROST on the outside, dude that's hot.

Re:Explaining the difference...

[Anonymous Coward]

My 2500 runs about 40 with the retail cooler and a cheapo case. Maybe all your fans are making the power supply run hot?

Re:Explaining the difference...

[PingPongBoy]

I know an e-machines AMD 2500+ running as a server. So far so good, but I'm watching it. Computers have really improved qualitywise. I think consumer awareness helps.

Re:Explaining the difference...

[owlstead]

Indeed. From the packaging of one of those silver emulsions: put little on processor, use credit card (and old one probably, they did not mention) to spread over processor. Worked for me. I don't everclock though, so I would not see melting processors too soon.

And note that the stuff is hell to get out of your clothing. It's just silver, it's not a solution on anything. The problem with that is that it won't dissolve in - well nothing really. It's silver, isn't it?

Re:Explaining the difference...

[freakmaster]

I know people who have had trouble getting their intel 875P to work in linux, particularly the SATA controller. this is not an issue w/ amd vs intel. just that it takes linux time to get drivers for new hardware. hopefully as linux continues to grow. the hardware vendors will be a scrupulous in making sure they have good linux drivers available as they are w/ windows (they are economically FORCED to release windows drivers w/ their hardware, but not for Linux).

Re:Explaining the difference...

[smash]

For him, it was the ONLY criteria for his computer...that it be "the fastest" available. I have stocked a school with 1 to 1.4GHz machines, and they work fine...most of his time is spent browsing the internet and sending e-mails anyway...but not buying "the fastest" makes him feel dirty, at least that's the feeling I get from what he's said to me. That's the thing that drives me nuts about people like that...they believe the marketing hype over reality.

No offense, but closed minded thinking like this is deserving of being ripped off.

smash.

Re:Explaining the difference...

[BJZQ8]

No offense taken...I would agree completely. My thought is that if he isn't willing to listen to reason and fact, then he gets what he deserves.

Re:Explaining the difference...

[Anonymous Coward]

That's a motherboard chipset issue, not a CPU issue.

Just like for all hardware, the current state of affairs is that you have to check for compatibility before selecting what to buy, especially if you're running anything other than the latest MSWin.

My Athlon XP and Athlon 64 on mobos using VIA chipsets run FreeBSD and Linux just fine.

The Athlon 64 is also capable of running in 64-bit mode. The only current offering from Intel that could do that isn't available at reasonable price levels nor with good 32-bit compatibility.

Re:Explaining the difference...

[Oakey]

Something is wrong with your setup. I have an XP2100, Lian Li Case, did have the SLK800 but is now an SLK900 (not a lot of difference) and I have a controlled 80mm fan. Idle temp is about 40c, hits about 50c full load and that's when the fan and SLK are gunked up with dust. If it's clean then it rarely goes to 50, averaging out around 47-48c. Oh, and it's overclocked to 2.3GHz, so yeah, you have a problem somewhere. It's not loud either, I have 4 19db papst fans, the loudest thing in the case is now the GF-FX 5900u.

Re:

[struckleberry]

Imagine...a Beowulf cluster of these! *sigh*

WWot?! 7p...

[Anonymous Coward]

plaInly states that

Virus protection on the chip?

[platypussrex]

The AMD website says the chip has virus protection against MSBlaster, Slammer, etc. Does anyone besides me think this is a bad idea? Not that virus protection is bad per se, but that all these "protections" built into the chip are harbingers of even more "protections" to come. I'll let your imaginations fill in the rest.

Re:Virus protection on the chip?

[Anonymous Coward]

What they probably mean is the chip has the capability to set segments of memory with a do not execute bit, for parts of memory such as the stack. That reduces the amount of things that a hacker can do if he finds a buffer overflow to exploit.

Re:Virus protection on the chip?

[garcia]

It's unnecessary as some of us don't need that crap in our machines. Unless this part of the chip is upgradable, for free, quickly, then what good does it do when a new one comes out?

Virus/worm/fuck protection is up to the people to use. Manu's can put all this shit into products and put it out there but that doesn't mean dick when the viruses get more virulent and the people don't use them...

Re:Virus protection on the chip?

[trompete]

Is that the automatic buffer overflow protection that Intel will be offering soon as well?

Re:Virus protection on the chip?

[RealErmine]

It just means that it has buffer overflow protection integrated into the silicon. This is just good engineering practice rather than an Orwellian plot. The article just dumbed it down.

Re:Virus protection on the chip?

[stateofmind]

I'm very much against this, if I don't want to be safe, thats my problem. If a doctor says I need anti-biotics for something, I'm not required to take them.

I understand that kind of thinking can hurt other people when my machine is on a network, and in my laziness my machine becomes infected and starts a DoS attack and such. But still, not every machine is on a network, and I don't want to have to pay for the extra security.

Let the processor do the processing.

Josh

Re:Virus protection on the chip?

[gokulpod]

If you knew how the protection is done you woudlnt be so critical of it. The core of AMD's Enhanced Virus Protection is the a NX bit which specifies whether a page of memory is executable or not. This way, even if buffer overruns occur in that area of memory, it wouldnt be executed. I am not really sure how this is a bad thing.

Re:Virus protection on the chip?

[tiger99]

Splendid, it will now just be up to programmers to use that facility, which I assume has to be configured by software. It has been needed for a long time.

Now who is going to have the first kernel which sets it all up properly to be secure? Linux? OpenBSD? FreeBSD? Or will it be that backward little company in Redmond who have major quality and security problems with everything they do?

Re:Virus protection on the chip?

[Anonymous Coward]

Microsoft announced it'll be part of Windows XP SP2, which is currently in beta.

Re:Virus protection on the chip?

[lederhosen]

Think it is allready in OpenBSD

Re:Virus protection on the chip?

[afidel]

Well since every other major architecture except IA-32 has NOEXEC or similar I would imagine that every Free OS has such code already, it might need to be ported and cleaned up but most of the work should already be done. Also XP SP2 should have it later this year.

Re:Virus protection on the chip?

[Short Circuit]

It should be simple to pass the flag into gcc and see if the kernel still works. (It should.)

Unless they do odd things like generate code on the fly, it shouldn't be a problem.

JIT runtimes might have problems, though, if you forced the flag in, say, your Gentoo installation.

Re:Virus protection on the chip?

[Short Circuit]

I'm curious if that bit is preserved when pages get moved into swap space. It'd be interesting to study the kernel mechanism for Linux.

Re:Virus protection on the chip?

[Anonymous Coward]

and you'd be 100% right

Re:Virus protection on the chip?

[Anonymous Coward]

It would have been an Orwellian plot had Intel come up with it first.

Re:Virus protection on the chip?

[Imperator]

Linux has done this for a while now. It's just a separate no-execute bit in the AMD64 extensions.

Re:Virus protection on the chip?

[Anonymous Coward]

PaX does this on Linux. It can also emulate it on CPUs that does not support it.

http://www.grsecurity.net/

Re:Virus protection on the chip?

[Too Much Noise]

from the x86_64 documentation in the Linux kernel:

Non Executable Mappings

noexec=on|off

on Enable
off Disable
noforce (default) Don't enable by default for heap/stack/data,
but allow PROT_EXEC to be effective

noexec32=opt{,opt}

Control the no exec default for 32bit processes.
Requires noexec=on or noexec=noforce to be effective.

Valid options:
all,on Heap,stack,data is non executable.
off (default) Heap,stack,data is executable
stack Stack is non executable, heap/data is.
force Don't imply PROT_EXEC for PROT_READ
compat (default) Imply PROT_EXEC for PROT_READ

Re:Virus protection on the chip?

[runderwo]

Actually, the new capability is to set execute permissions on a per-page basis in hardware. IA-32 already had an execute permission bit on segments. No 32-bit IA-32 OS that I know of uses a segmented model though, preferring a flat memory model, which means we only get read and write permissions in hardware.

Thanks to this minor oversight in the design of IA-32, we have gone a long time without the benefit of hardware execute protection. There are software kludges that try to work around this (like working around the 386 bug with page write protection), but a hardware solution will be more robust and speedy.

Re:Virus protection on the chip?

[Anonymous Coward]

so now, programmers can be even more lazy??

good god, software is overbloated slow crap already because the lazy sod's have overpowered, orver sized storage avaialble... now we have overflow protection so we will see any good coding go right in the crapper for sure...

nice... what's next a VB.net interpeter on the silicon?

Re:Virus protection on the chip?

[Anonymous Coward]

I mod people down who say "per se"

Re:Virus protection on the chip?

[Anonymous Coward]

Did Alphas in the days of Windows NT for Alpha have an NX bit? If so, did Windows support it?

Re:Virus protection on the chip?

[owlstead]

Doesn't save you from overwriting an important piece of data though. So it's not the final word on buffer overruns I presume. Makes it harder to infiltrate a program, but in no means impossible.

Re:Virus protection on the chip?

[tiger99]

Good. When the price is right, I will go the 64 bit route. It will be Linux, if it dual-boots with anything else, it will be one of the BSD family, or maybe BeOS if the free version is complete, and availabe in 64 bit form. (I would willingly buy a new copy of commercial BeOS if I could, money is not the issue).

I realise that this is not the entire solution to the virus problem, but at least if data areas are seperate from program code, such that you can't execute data, a mechanism for stupid things happening has been removed. It will IMHO do more to trap programming errors than viruses, and should, if I have understood correctly, help to stamp out some of the worst programming practices such as self-modifying code. All in all, it is a good feature, long overdue, but realistically, there will still be programming errors in the OS , even Linux. Work on preventing these, which can be tackled at compiler level, or with secondary tools such as lint, the old *nix program checker, will continue for a long time yet. gcc can do a lot, but only if you use it properly. It is quite depressing to tun a kernel compile, and get about 100 warnings due to comparisons between unsigned and int, for example. They might all be harmless, but effort to get rid of them all, and get clean complies with the strictest checking, would not go amiss.

When I used to be a C programmer, back in 1986 (part-time and as an adjunct to my real job of hardware design) we would never let anything out unless it could pass lint with no warnings.

My latest PC has never been exposed to the rubbish that comes from Redmond, nor will any others in the future. 10 years of misery, crashes, security holes, deliberate incompatabilities, and programs that try to be too clever and do things Sir Bill's way, when that is not what is needed, was quite enough!

MICHAEL IS FUCKING LAME...

[Anonymous Coward]

you know it. i know it. that motherfucking post was on topic (he's going to go get a new AMD proc....**GASP!** yes, the very one IN the story)....off topic my cock.

blow me, michael.

Just tell them this

[Anonymous Coward]

AMD will whiten your teeth. It's a fact, people love teeth whitening above everything else.

Re:Just tell them this

[Ralph Wiggam]

They should come out with teeth whitening cigarettes and coffee. Those would be big sellers.

-B

Re:Just tell them this

[YOU LIKEWISE FAIL IT]

Anonymous Coward got my teeth whiter than ever before! I would recommend Anonymous Coward to anybody!

$733 for 1000

[Nasarius]

Priced at $733 for 1,000 units...

Whooo! I can get one of these for 73 cents! :-P Yeah yeah, I know what they mean, but that's some horrible wording.

Re:$733 for 1000

[MonkeyCookie]

Yeah, it should have been something like:

Priced at $733 for orders of 1,000 units...

Re:$733 for 1000

[silicon not in the v]

I'll go in with you on that. 998 more volunteers?

Re:$733 for 1000

[MindStalker]

Believe it or not many online stores are selling it for 733 plus shipping and handling. So you don't have to do this.

Re:$733 for 1000

[silicon not in the v]

Yeah, good to know, but I run a little behind the cutting edge of technology. I'm still hesitating about getting a new video card to replace the 2MB one I still use that came in my Pentium 100 back in '95. If someone has a better video card than mine sitting around they want to get rid of, I'd be happy to take it off your hands.

Re:$733 for 1000

[gauchopuro]

You're asking that question on the wrong site.

Re:$733 for 1000

[MindStalker]

Wow, I and I thought I was doing bad with my voodoo 3 :)

No, you misunderstand

[Anonymous Coward]

The protection simply prevents buffer overrun exploits. It's a good thing, and has nothing to do with possible data/content censorship or DRM.

Very solid not good enough

[Anonymous Coward]

I prefer my processors to be very very solid.

Re:Very solid not good enough

[Anonymous Coward]

I prefer mine to be solid and throbbing.

Preferably with deep penetration into my specific market share.

Re:Very solid not good enough

[LordK3nn3th]

They just don't perform as well when they are melting down the side of your mobo

Pardon me...

[Short Circuit]

...but why aren't the graphs loading? I can see that they're Flash, but when I right-click on them, it says "Movie not loaded" and "About Flash Player"

If you do manage to see the "movies" ... would anyone mind converting them to GIF or PNG?

Re:Pardon me...

[Anonymous Coward]

They do that so you have to see the ads. Might be annoying to you, but its better for Anandtech.

Re:Pardon me...

[juhaz]

Not that it works... adblock and few handy regexps quickly negates whatever evil trickery they've resorted to.

Got to love blocking flash with patterns.

Re:Pardon me...

[Short Circuit]

Except I can see the ads. Including that animated one off to the right that looks like an animation in space.

Re:Pardon me...

[Anonymous Coward]

Possibly server being bombarded with requests?

Re:Pardon me...

[FuzzyBad-Mofo]

Nifty.. although I'm too lazy to implement such a filter myself. Much easier to use a "Click to play Flash" browser extension. :)

Re:Pardon me...

[CoreDump01]

That's why i love flashblock for Mozilla / Fire*

http://flashblock.mozdev.org/

It lets you select exactly which flash banner / whatever on a site is displayed.
This is a must-have IMO

Re:Pardon me...

[cbiltcliffe]

My question is, why the fsck do you need to use flash to display a static picture?!
What kind of jackass thought that was a good idea? What's wrong with .png's, or .gif's, or heck, even .jpg's?

That's by far the stupidest design decision I've ever seen on a website, and a tech site should certainly know better....

Re:Pardon me...

[canajin56]

They are using ColdFusion. ColdFusion has a nifty little tag that dynamically generates a chart/graph based on input. I think the default if you leave off the format="flash/png/whatever" property, it defaults to Flash, but I could be mistaken (I think that property might be required, so no default)

The flash output option is neat because you can have it so a mouse over displays additional information about whatever bar/pie segment/whatever you are over. It can also make each bar a hyperlink, or perform some other action. But it can do that with PNG formatted graphs as well, using normal HTML image maps.

On the other hand, their website doesn't do those fancy things with their graphs. It just displays them. It might be the case that since graphs are so simple an image, a vector representation such as Flash is smaller than a PNG would be, so they chose it to save bandwidth.

Capsule Review.

[Anonymous Coward]

There's a new consumer CPU out. It's faster than the ones previously made.

Re:Capsule Review.

[Anonymous Coward]

Well, sentence one is given (and that counts as news.) Sentence two is the purpose of the review... which shows that in certain cases, it is faster, and in other cases, it is not. Useful information for people looking to upgrade.

Whatever

[Anonymous Coward]

The two limiting factors in a PC these days (not taht home user should care) are the memory size, and the system bus speed.

Most people won't feel the limits in processor speed.

Re:Whatever

[NanoGator]

"Most people won't feel the limits in processor speed."

Those of us that do 3D rendering for a living would disagree. Then again, to us, faster processor == more money.

It's a pity for Intel/AMD that gamers don't really rely on the processor as much these days. Faster frame rates are a video card away.

Waiting

[unassimilatible]

you'd be better off waiting a couple months for Socket 939."

So if I wait long enough, better, faster stuff will come out?

Re:Waiting

[Anonymous Coward]

See comment here.

Re:Waiting

[Anonymous Coward]

There are times to wait for the next best thing...

AMD's 754 and 940 will be replaced by 939s (I think). If you don't think you will ever want to upgrade this CPU, it doesn't matter what you choose. I think the 939s are dual channel though (754 is single).

BUT... as for current AMD motherboards, PCI Express isn't around and is a _MUST_ at this point. PCI Express will be taking over from this point forward. This means future video cards _may not_ be available for AGP setups. I bet the higher end cards won't be at all.

I haven't upgraded a CPU in any computer I have, but I have upgraded vid cards. I am sure AMD boards will have PCI Express soon. If you are thinking of buying, let this be your reason to hold off.

Re:Waiting

[Anonymous Coward]

So if I wait long enough, better, faster stuff will come out?
no, but it will have more legs/pins.

Re:Waiting

[Imperator]

Yeah, and what's really unusual is that the current stuff will be cheaper if you wait.

Re:Waiting

[Anonymous Coward]

So, how's that parallel universe where 939 is more then 940?

Re:Waiting

[iceperson]

I never upgrade components anymore because it just became too expensive. Anytime I upgraded more than 2 or 3 things I felt compelled to build a machine just to put the old parts in and ended up just spending more and more money. I now have 5 machines at the house and I've supplied machines to most of my family members for cheap. Sure I recovered some of the costs by selling them the old machines but I never recovered all of my expenses and I eventually ran out of family members =)

Re:Waiting

[Anonymous Coward]

>>This means future video cards _may not_ be available for AGP setups. I bet the higher end cards won't be at all.

ATI will be producing both AGP and PCI-Ex versions of their GPUs. Nvidia will use a PCI-Ex to AGP bridge chip that can be added by a manufacturer to their cards so AGP will be supported. It's safe to say that you will see AGP products for at least 6 months to a year before PCI-Ex takes over.

Re:Waiting

[ozbird]

I am sure AMD boards will have PCI Express soon.

Yep, just as soon as the PCI-X video cards are available. Problem is, the video card manufacturers won't produce PCI-X video cards until there are motherboards to plug them into. Catch-22.

Re:Waiting

[Anonymous Coward]

PCI-X and PCI Express are two entirely different things

Re:Waiting

[jo42]

Personally, I'm waiting for Socket 1024(tm).

Re:Waiting

[Anonymous Coward]

> BUT... as for current AMD motherboards, PCI Express isn't around and is a _MUST_ at this point. PCI Express will be taking over from this point forward. This means future video cards _may not_ be available for AGP setups. I bet the higher end cards won't be at all.

You mean the graphics card vendors will refuse to support the many millions of owners of PC systems with AGP ports, and will instead only support tiny fraction of users that have PCI Express? Yeah, that sounds plausible.. most companies are not interested in actually selling cards and making a profit..

In reality, you will be able to find all video cards in an AGP form factor for a long time to come.

wait a sec..

[Anonymous Coward]

i think michaels computers had this processor in his systems for months now, this is old news ...

Re:wait a sec..

[bhtooefr]

http://www.tomshardware.com/column/20040317/images /desktop1.gif

Nope, it's the "AMD FX-51 3400+". Actually, if AMD used PRs on the FX-51, it would probably be 3400+, seeing as the only difference between the FX-51 and the A64 3400+ is that the FX supports dual-channel, but requires ECC. Performance isn't that much different, even...

Still, it's obvious that Michael's Computers is BSing...

Re:

[Anonymous Coward]

I'll get a better beowulf cluster for the price using old pentiums (since I can buy five times as many processors for the money).

Intel changing naming scheme.....

[vwjeff]

Intel is changing their naming scheme soon. You can find the article at:

http://news.com.com/2100-1006-5174895.html

Intel was forced into this due to the many variations of a chip with the same clock speed. It's also a good way for them to explain why their Pentium-m is faster than the Pentium 4-m.

Re:Intel changing naming scheme.....

[trs998]

what do you want to bet their numbers will come out higher than AMD's scheme? remember Intel's complaining about amd's scheme when it came out?

now, a standard numbering scheme, preferably about the same performace per number as AMD's (seems quite good to me) would do. It'd have to be ratified by some independant entity.

Any new processor would be put through this test, say 32-bit MS office automation + 3dmark2003 etc... lots of apps to prevent optimizations and a 64-bit test (XP 64-bit?). Yes, linux will work on anything with the same software etc, but windows speed is what the unwashed masses will want.

each processor would then be issued with a 32-bit and a 64-bit performace figure.

Yes, ok, it'd be wrong because of this, that, and the other I'm sure someone will come up with, BUT, it'll de-confuse the masses. probably.

Re:Intel changing naming scheme.....

[PingPongBoy]

Is there a suite of benchmarks that I can obtain with little cost? I don't trust these performance claims. Computers should demonstrate their power with a wide variety of well known problems.

I also like seeing the MHz go higher because in the end that's what you need to improve performance on a given architecture.

Oh, the humanity--

[Anonymous Coward]

Aiming for funny and getting insightful.

I'm sorry Dave, I can't do that.

[Anonymous Coward]

One of these days, the CPU itself will be telling you that.

Invisible Flash

[Euphonious Coward]

All the graphs looked alike to me: an icon in the middle of a rectangle. Would anybody care to post PNGs?

(Note: I am very happy, otherwise, that flash ads don't clutter my Galeon window. If I could click on any flash that I actually want to see, that might be OK, as long as it didn't execute any actual buggy, insecure Macromedia code. But what good is it, really? Tufte had the word: "chartjunk".)

Osama bin Laden - About to be caught???

[Anonymous Coward]

http://www.foxnews.com/story/0,2933,114497,00.html

Other Reviews

[breakinbearx]

As covered by arstechnica, there are also reviews at [H]ardOCP, Hexus, HotHardware.com, Sudhian, and The Tech Report. AMD's official announcement is here.

A meme is a terrible thing to watch

[Anonymous Coward]

Google shows no hits for "Buzzword Bufoonery".

We have a new trope.

Re:A meme is a terrible thing to watch

[MikeCapone]

No hits?

How do they call that... uh

There's a whole website for that kind of stuff.

Damn, I've forgotten what it's called!

Re:A meme is a terrible thing to watch

[Anonymous Coward]

> Google shows no hits for "Buzzword Bufoonery".

That's because buffoon has two fs.

Re:A meme is a terrible thing to watch

[BJZQ8]

I thought of that after I clicked submit But Google doesn't list "Buzzword Buffoonery" either

Nitpicky Bastard is a person

[Anonymous Coward]

He is author of parent.

Flash for Graphs?!?

[Rob Riggs]

Who the hell uses proprietary Flash(tm) technology to display simple friggin' graphs! What the heck is the purpose of that? PNG, JPEG or GIF isn't good enough?!? Someone needs to hit these guys with a cluestick.

That's just lame.

Re:Flash for Graphs?!?

[Anonymous Coward]

This has been discussed a thousand times. Here goes one more...

Anandtech gets revenue from advertisements. These advertisements are in flash. If you don't have flash enabled, then Anandtech does not get paid for that advertisement. Therefore Anandtech makes sure the information of value is also in flash, to ensure that they are compensated for your viewing of their material.

So please, when you find that cluestick - make sure to give yourself a good whack with it.

Re:Flash for Graphs?!?

[mapmaker]

What the heck is the purpose of that?

The purpose of that is to make you enable Flash when viewing thier page. That their ads are also Flash based is purely a coincidence...

Re:Flash for Graphs?!?

[Anonymous Coward]

Hooray for Flash Click to View!

Re:Flash for Graphs?!?

[beeblebrox87]

The trouble is, this requires that Flash be available for you platform. Many architectures, operating systems, and browsers are not supported by Flash at all. Until a decent, reliable, open-source (i.e. easily recompileable) flash player is available, I will continue to avoid flash-using sites like the plague.

Regarding ad revenue: why can't they just display GIF or JPEG ads? GIF ads can even be animated. What features do the advertisers need that can't be provided by a platform-independent GIF animations?

Re:Flash for Graphs?!?

[Srin Tuar]

>Therefore Anandtech makes sure the information of
>value is also in flash, to ensure that they are
>compensated for your viewing of their material.
>So please, when you find that cluestick - make sure
>to give yourself a good whack with it.

That justification can be (and often is) used for making everything suck.

Pay cable channels with more ads than show (not even counting content-embedded ads), DVDs with non-skippable "previews", DRM, Trusted computing, Windows in general, poorly documented proprietary media formats, nagware, popup ads, websites with hostile flash graphics, etc etc...

When is someone going to come up with a way to make money by making thing better? Where is the free capital market when you need it- or does it just not apply to bit and bytes at all?

The more I see, the more I'm beginning to think copyright was a bad idea from the start...

Re:Flash for Graphs?!?

[Sexual Ass Gerbil]

Fuck that shiznit. I realize people are proud of their products and want to get the word out that these things exist and are available to buy. I like to know what cool toys exist too, but contentless ads that flash and flicker don't do anything to promote a product. All the ads do is proclaim "Hey! We're desparate for your attention and your dollars!" Fuck that. Tell me about your product, damn it! Needless to say, I ignore and right-click block all graphical banner ads, including the ones on Slashdot, but I do read text ads.

Re:Flash for Graphs?!?

[Anonymous Coward]

Many architectures, operating systems, and browsers are not supported by Flash at all.

Of course you probably meant to say "A few architectures, operating systems and browsers are not supported by Flash at all. And those few represent a pittance of the market." C'mon, Flash works on my Sun Ultra 5 running Aurora Linux. If that isn't a niche market, then nothing is.

But go ahead and keep those blinders on. The rest of us will be smarter than you.

Re:Flash for Graphs?!?

[fahrvergnugen]

SWF files are significantly smaller than .gif files for this sort of presentation, and can be generated dynamically from a database of performance numbers with little effort.

An example, using the Aquamark 3d graph under "Directx9 Performance." First, I saved the page. Then I took a screenshot of the web browser, cut out just the graph, and exported it optimized for the web using Fireworks.

The results: .gif: 21,514 bytes .swf: 6,582 bytes

And the .gif file, might I add, didn't look nearly as good.

Now, 15K might not seem like much, but when you multiply it by the considerable amount of traffic Anandtech receives, the savings quickly become significant.

Plus, modem users will appreciate it.

Re:Flash for Graphs?!?

[startled]

I've been using click to view since I first heard about it a few months or so ago. It's just awesome-- there's no more useful extension out there for casual web browsing. You have no excuse not to run it.

Re:Flash for Graphs?!?

[gnu-generation-one]

"Who the hell uses proprietary Flash(tm) technology to display simple friggin' graphs! What the heck is the purpose of that? PNG, JPEG or GIF isn't good enough?!?"

Even Toms Hardware uses images instead of text/stretched-single-pixel graphs to display their results for anything. Means you can't use the browser's search facilities to find items in any of their mega-something-roundup reviews.

I presume Anananandtech is just trying to make people turn off their Flash-blockers so they can display more adverts, but the net result is, as you say, "leave and don't return" to most people.

Re:Flash for Graphs?!?

[HD Webdev]

Now, 15K might not seem like much, but when you multiply it by the considerable amount of traffic Anandtech receives, the savings quickly become significant.

Plus, modem users will appreciate it.

And, Flash scales well if the author sticks with vector for creating the material and doesn't imbed gif's.

Re:Flash for Graphs?!?

[Anonymous Coward]

Who the hell uses proprietary Flash(tm) technology to display simple friggin' graphs! What the heck is the purpose of that?

I assumed that using Flash gave higher resolution graphs when printing. Not having a Flash plugin, I cannot verify this.

Re:Flash for Graphs?!?

[Anonymous Coward]

Another fine example of open source software robbing people of the ability to earn a living.

Re:Flash for Graphs?!?

[geschild]

Hooray for Adblock. No need to click the graphs, the ads just aren't there when you open the page, and never even loaded.

Ymmv, however.

Re:Flash for Graphs?!?

[rnd()]

The more it sucks, then the more of an incentive there is for companies to stop doing it. Eventually, a good balance will be reached. The balance does need to include copyright if you expect people to wake up every day and create it for you (except for rare exceptions, they need to get paid)...

Wait a couple of months?

[fullofangst]

I don't get it. Why bother saying "you'd be best off waiting" for the next chip ? The Athlon FX-53 is a flagship chip. It's the currently fastest chip they do. If you want the highest performance, you would obviously buy it now. If you wait a couple of months then you don't want the highest performance. This is what this chip is for, here and now - the fastest available performance. Yes there will be a faster one in a few months but that just continues ad infinitum. If you lived by the rule of waiting for something faster to come out, you'd die of old age before you actually purchased the damn thing.

Re:Wait a couple of months?

[Phosphor3k]

What he is saying, is that the current socket form factor will be discontinued in a few months. If you want to ensure any sort of future compatibility as far as upgrades are concerned, bite your lip and wait a few months for the new socket.

Re:Wait a couple of months?

[Anonymous Coward]

See comment here.

Re:Wait a couple of months?

[juhaz]

Because the FX-5x line of chips is using a socket type that is doomed in few months. So after you want the fastest ship after few months you end up buying a new motherboard as well. The new socket on the other hand is going to be in use for some time.

Sure, if you've got money buy whole machine every time you want to update the processor (or if you already have a board that's compatible with FX) go for it.

The cycle does continue to ad infinitum, but this is more of a case of deciding at which point to enter the cycle, now it's nearing its end and beginning of another.

Re:Wait a couple of months?

[niko9]

There is one very good reason to wait for 939. It supports unbuffered DDR Ram. RAM that most people are using with their Pentium 4 Dual DDR chipsets. Socket 940 is usable only with Registered ECC DIMMS which are more expensive and a bit harder to get a hold of.

So I assume that many people, like myslef, will take those 2 sticks of DDR 3200 RAM they invested in for their P4/875 setups, and use them in their Socket 939 motherboards.

Re:Wait a couple of months?

[fullofangst]

Re: replies - Your comments about new socket / slot types coming out soon is, of course, entirely relevant. However it's also redundant. A small range of chip speeds are available for a motherboard type, at which point something changes and you need a new motherboard. Hands up who bought a Pentium 4 2.0ghz thinking you would be able to put any future model into your motherboard, only to find Intel went and upped the front side bus speed, making your MB incompatible with a 3.2ghz HyperThreading chip?.

This is my point exactly - this may be the end of the line of chips you can buy for this socket type, but few people upgrade their CPU every couple of iterations. I went from a P4 1.8 to a P4 3.0 - I did not get a 2.0,2.4,2.533,2.8 along the way as there was just no need. I don't plan on upgrading from a 3.0 to anything under 4.0ghz because -it just wont be worth it-

Re:Wait a couple of months?

[Tassach]

Being an early adopter of a Slot-A Athlon, I've often regretted not having waited the extra 9 months for the Socket-A to have come out.

Re:Wait a couple of months?

[owlstead]

Dunno what I would do with a spare, expensive processor without a motherboard. Why not just buy the stupid thing and buy a CPU/motherboard/memory combination later on? You're sure not to only want to update your processor. And with a mobo, processor and memory a new computer is easily born (or upgraded) so it is easier to sell, or to give away to friends.

Well at least that's the theory, i've got 4 computers (not including my PDA) in my 3 by 4 meter room by now :) They're working though. Oh well, one linux server, one main pc, one test pc and a laptop. They're all used.

Re:Wait a couple of months?

[Anonymous Coward]

If you're going to buy the latest and greatest, you might as well upgrade the motherboard and CPU together.

Most people also don't upgrade all that often, and by the time you're upgrading the CPU, you probably want a new motherboard for other reasons, as well.

Re:Wait a couple of months?

[Anonymous Coward]

I was to buy a ZX-80 once, but I heard that the ZX-81 was in the works...and then, the Spectrum was just around the corner. Oh, I'll buy a computer any day now, really really soon...

What's with AMD's name

[superpulpsicle]

Why can't AMD stick to the traditional K6, K7 and keep on adding up. All this Fx and Opteron and 8 million other names are confusing as hell.

Whatever happen to marketing making your purhase decisions easier. It's doing to exact opposite nowadays. Intel ain't doing a whole lot better.

Re:What's with AMD's name

[Anonymous Coward]

Opteron and FX are both socekt 940, both dual channel with registered memory. Opteron for server, FX for enthusiast crowd. A64 are socket 754, single channel, for main stream. Not too hard of a thing to understand. I think AMD's done a pretty good job of seperating the opteron name from the xp and 64 line, where you know the price is backed up with its top notch performance. Besides, K7 was never mentioned like that, it was always the Athlon. Same thing with K8, Athlon 64

Re:What's with AMD's name

[blindbat]

This is even worse with nvidia's graphics cards. The names are numbers which don't really tell you anything about performance. Such as some of the 5000 series cards being worse than the Geforce 4 cards.

Re:What's with AMD's name

[BiggerIsBetter]

Hmmm, what CPU I want? 3Dnow!? SSE? 3000+ or 3000MHz? FX? AMD64? Opteron? HyperTransport? Oh I don't know, just give me that one *points to box on the shelf*.

Two theories here:
1) Baffle them with bullshit. Average punters get confused and just buy whatever the salesperson sells them.
2) Make stuff sound impressive and technological, so uninformed wannabe geeks can buy whatever AMD is making currently, and feel 7331 about it.

Re:What's with AMD's name

[SFBwian]

I just hope that the K9 doesn't bite them in the ass or run like a dog.

You're now permitted to groan at my awful joke.

Re:What's with AMD's name

[tomstdenis]

While that certainly is true it isn't something only AMD does. I mean really what is "NetBurst" or "HyperThreading".

For all intents and purposes SMP can do multiple threads [better I might add] and it's not "Hyper".

You want to know how to buy a CPU? Goto about 10 diff review sites [ignore ones that just copy the back of the box] and form a scatter chart of what they all say. Take the average and go for it.

Best of all, know your needs. If you don't run render/compiler farms chances are you don't need a 350$ CPU. Heck you can get sufficient power for home comptuters (including developers) from a 40$ Duron Morgan core.

Tom

Re:What's with AMD's name

[Slack3r78]

Heck you can get sufficient power for home comptuters (including developers) from a 40$ Duron Morgan core.

Better still is $40 for the new Duron Applebred core. $43 got me a 1.8GHz Applebred chip, which is essentially an Athlon Thoroughbred-B with partial cache.

Re:What's with AMD's name

[tomstdenis]

Point the same. I didn't know they made another duron. I gave up and wasted a helluvalot of money on a Barton 3000+ [partly for fun, partly for the cache].

Tom

Re:What's with AMD's name

[Slack3r78]

I didn't either until a few weeks ago, though they've apparently been floating around for several months now. Almost wish I hadn't heard, that's $40 I wouldn't have spent otherwise ;)

Re:What's with AMD's name

[Anonymous Coward]

The Athlon brand was adopted at least partly because AMD didn't want it to be associated with the K6, which people had bad experiences with. Also consider Intel's successful branding campaign with the Pentium.

I think the FX line is somewhat pointless, but Opteron as a server brand and Athlon as a consumer brand is perfecly reasonable. AMD seems to be sticking to Athlon just like Intel is sticking to Pentium, so it doesn't seem like there's too much confusion.

In fact I like that AMD doesn't use sequential numbering, because that would only be confusing.

Opteron and Athlon 64 share the same architecture...should they share a number? They aren't even motherboard-compatible. Athlon 64 is newer, maybe it should have a higher number...but the latest Opterons are still faster.

For an ordinary consumer, the Athlons so far are Athlon, Athlon XP and Athlon 64; the only thing that might be confusing is that the Athlon 64 has significant capabilities above the Athlon XP...but AMD presumably assumes that people will be running 32-bit software for quite a while (hopefully not as many CPU generations as it took most consumers to start running 32-bit software...), so it probably doesn't matter; soon all the higher PR ratings will be Athlon 64 processors, anyhow.

Duel Opterons

[BrookHarty]

Been wanting to go back to a true Dual system, (my last was a Dual P3-800, My Dual P2-400 is my Linux box) Keeping an eye out on prices for a new modern Dual system compared to a fast AMD FX.

You can pick up a Dual AMD-2800 for about 500 bux for a barebones cpu's+mb+case (also uses PC2100 ram). Opterons for dual systems are ridiculously priced, 248's are about 900 bux each, and motherboard for 300, so about 2500 dollars for a basic barebones system. Dual Xeon 3.2's with 1meg cache are about the same price, but xeon motherboards are less "workstation" friendly, and more expensive. ( PCI-64 slots, etc)

Also with PCI-X gfx cards about to be released, a bunch of new motherboards will come out. And It looks like Socket 940 is going to be phased out later this year for Socket 939, so a FX buy might be a locked in purchase, with no upgrades. Which the Opteron uses 940, so I'm a little confused about the Opteron's upgrade path.

Hoping if I want 6 months, the prices for Opterons will be down enough to build a basic dual system, with PCIExpress, and at least 2+ ghz CPU's. Something that will be fast as an FX in gaming, but also have the dual cpu smoothness feel with power of running virtual machines and crunch numbers well.

The Xeon line is cheaper, maybe some new motherboards might come out and bump it up to the system im thinking about.

Re:Duel Opterons

[Tumbleweed]

1) PCI-Express is what's coming, not PCI-X, which is something older. PCI-Express is mostly being abbreviated as PCI-E, from what I've seen.

2) Socket 940 isn't the one being phased out, Socket 754 is. Socket 940 is for Opterons & Athlon FXs. Athlon 64s are what use Socket 754, and that's the one being ditched for Socket 939. Once Socket 939 is available, Athlon FXs will also be made available in that socket form factor. Socket 939 & the new Athlon FXs will also enable you to use NON-registered memory, which will be less expensive, though you won't be able to use as much of it.

Re:Duel Opterons

[Anonymous Coward]

I too would like to see Opterons duke it out.

Re:Duel Opterons

[BrookHarty]

I do get PCI-X and PCI-Express mixed up, name sounds pretty close.

The roadmap shows 939 for FX AMD 53/59, 940 is being ditched for FX, but not Opteron. So yes, 754 is end of life, and so is 940 for FX.

Why by a 940 FX board when you can only upgrade to single cpu opterons?

Re:Duel Opterons

[trentfoley]

Why do I hear banjo music? ..you got an awful purty chip, boy...

Waiting

[aztektum]

Todays better, faster stuff will be cheaper

FX-53 tested against 16 other processors

[EconolineCrush]

Tech Report's review tests the FX-53 against a total of sixteen other chips. Good reading if you've got a benchmark fetish, too.

Missing one important graph:

[scrytch]

Price. The price difference between some of these chips they're benchmarking puts them in different leagues. The FX-53 is NOT cheap compared to the 3200, but the P4EE makes them both look like chump change. This review looked like the output of a report generator (written by Macromedia I imagine), not a review.

Re:Missing one important graph:

[TrancePhreak]

$855 - Pentium 4 3.2GHz 800MHz 2MB Extreme
$733 - Athlon 64 FX

Please note this is a regular Athlon FX and not the FX 53 mentioned in the article.
This negates your ideals that the P4EE makes the FX seem like chump change. The FX is a highly priced CPU, as are P4EE's.

You're sure?

[eddy]

When they change to socket 939, however, you'll be stuck with what you've got--no upgrade for you!

AMD has indicated that 939 and 940 will coexist for a while, meaning that yes, there should be both 939 and 940 versions available of the same speed grade, probably for a year or so.

Then if that's enought or not, that's up to you. But remember that the rest of the platform evolves too, so if you're going after the top-of-the-line with every upgrade, just replacing the CPU(s) wouldn't be enought anyhow, so a socket-change wouldn't matter.

Re: sig

[irenetheno]

You forgot to add another a() on the end after the closing brace so that the defined function will actually run.

Running precludes looping. :)

Smaller

[ChrisMaple]

I've read that flash uses considerably fewer bytes. Lower cost for anandtech, so they can live longer. Quitchergripin.

Kanavit?

[Moocowsia]

That wounldn't happen to be Kanavit would it? He posted 13000 posts of pro intel crap on the futuremark forums before the mods banned him for spreading disinformation.

... but would you have any grey poupon?

[Kiyooka]

sorry, just had to get that out...

Short version: High price for little return

[Junks Jerzey]

I know, I know, these silly review sites love to have these "longer bar is better" graphics, but let's look at this rationally.

Take the SysMark 2004 benchmark. The commodity priced Northwood 3GHz P4 clocks in at 176. This new Athlon gets a 199. Ooooh, longer bar! But what does it really mean? I means that the Athlon is ELEVEN PERCENT FASTER than the processor that's one notch above the absolute bottom end you can get in a Dell PC (3GHz, the bottom end is 2.8GHz). And the price is over THREE TIMES HIGHER. Is this worth it? Does it make sense?

The answer is no, *unless* you are simply looking at the 64-bit capabilities. If that's the case, then great. Otherwise I don't see why anyone would care about these benchmarks.

Re:Short version: High price for little return

[Anonymous Coward]

yeah well deal with it. it's the 3-inch advantage =)

ps. mine is bigger than yours

Re:Short version: High price for little return

[FuzzyBad-Mofo]

I have trouble understanding what is represented by the SysMark benchmarks. What exactly is "Internet Content Creation"? Anybody? Bueller?

So what if ProcessorX gets 19999 Bungholio marks, show me real world benchmarks!!

Re:Short version: High price for little return

[jbrandon]

I means that the Athlon is ELEVEN PERCENT FASTER than the processor that's one notch above the absolute bottom end you can get in a Dell PC (3GHz, the bottom end is 2.8GHz).

I don't know where you come up with that crap, but Dell sells tons of machines slower than 2.8 GHz.

Re:Short version: High price for little return

[Junks Jerzey]

I don't know where you come up with that crap, but Dell sells tons of machines slower than 2.8 GHz.

You are right. The bottom end 2400 has options for 2.4 and 2.6GHz. Every other desktop starts at 2.8GHz.

Re:Short version: High price for little return

[357_Magnum]

Real world numbers like what? Frames per second ina game? What will that mean to people who don't play computer games? Maybe you would like them to show how many seconds it takes to apply a filter in Photoshop, which won't apply to people who don't use Photoshop. So some number has to be used because people like to hear that one number is bigger than another.

Re:Short version: High price for little return

[Anonymous Coward]

If you read the article, then you'd realise that these same silly review sites also tell you, very clearly - No, this is not worth it, and No, they do not make sense.

Re:Short version: High price for little return

[FuzzyBad-Mofo]

Sure; FPS, Photoshop Filters, Compile Times, IPS, FLOPS. These and other tests like them have their place, and may be easily tested and verified.

However, when SysMark says something like "Internet Content Creation", what does that mean exactly? Rendering a HTML page? Generating an SVG in Flash? Coding a farking webpage? Serving a PHP dynamic page? All of these could meet the above definition, and that's what's wrong with it. Pratically all of SysMarks benchmarks are described by mealy-mouth ambigous terms. Hence my use of "Bungholio marks", which is just as accurate a term as "Internet Content Creation".

I guess the reasoning is something like this.

[Kjella]

This platform is not upgradable. If you can justify the insane premium of buying THE fastest processor, you can probably justify buying another one real soon (otherwise you'd be better off buying something somewhat below the top and keep it longer). And to do this, you need an upgradable platform.

For me, it doesn't really matter. I've found that by the time I want to change CPU, I want to change the mobo, memory and all that anyway. Upgradability only matters if you need to upgrade your CPU much faster than other components.

Personally, I'm torn between a no-fan Hush PC (for the silent bliss) and an Athlon 64 Socket 939 + DDR2 + PCI Express + whatever ATI/nVidia has at the time (for the raw power). Or a Shuttle XPC to be somewhere in between. But in either case, I won't need my current box, that I'm quite sure of...

Kjella

Hey, it loki the loser!

[Anonymous Coward]

Ah, come to troll about Tom's site I see.

Have you thought about creating your own hardware review site, insted of peeing your pants everytime Tomshardware is brought up? I'm sure your monny will let you loan her PC to start. Just don't delete her file...OK big boy.

CPU thoughts

[Tumbleweed]

I wouldn't buy a 940 board for an FX system - 940 will be the thing to get for Opterons, if you want multi-proc, and/or registered memory to get LOTS of memory, for a DB machine or something. It's good to have a choice, as long as both choices are going to be around for awhile. Really, you'll want to target your platform (CPU & mobo choices) to what you want to do with your system.

I guess it's good this socket switch is happening at the same time as the switch to PCI-E, so you can get all this stuff outta the way at once.

Personally, there's no way I'd get a 939 board withOUT PCI-E at the same time - you're just going to have to ditch that soon enough, as well.

Really, I'm more interested in the PPC 970FX & 980, anyway. C'mon - 24.5 Watts for a 2gHz 970FX?! That's pretty amazing. I don't know if the 980 will have an on-die memory controller, but here's hoping.

As for other upcoming CPUs, the Intel Dothan also sounds very promising, as does the possibility of a multi-core version of the Dothan. I just wish Intel would wake up and make the Pentium-Ms for desktop platsforms. I also think the 2Meg L2 of the Dothan is overkill, and makes the chip more expensive than it needs to be. A 1Meg L2 is plenty for Dothan, and would reduce the die size considerably, and thus the cost. Intel's being a little strange on that front, though the Celeron-M is a good move.

I'm still curious as to why Intel and AMD haven't added AltiVec to their x86 processors. :(

Re:CPU thoughts

[Slack3r78]

Because AltiVec is just an Apple brand name for a vector processing unit? What do you think MMX, SSE, and 3D Now! are?

Re:CPU thoughts

[Tumbleweed]

Much less efficient, for one, and not quite the same as AltiVec, or VelocityEngine, or whatever non-brand name you want to use for AltiVec. There's a _reason_ AltiVec machines do so much better at some computing tasks than the x86 crowd.

The thing is, I don't know who owns the AltiVec IP - Motorola, maybe? And I don't know if it's something they (or whomever) licenses or not. One would think AMD, at least, would be openminded enough to check it out, if so. An Athlon 64/FX/Opteron with AltiVec would be a real _beast_. I mean that in a good way, of course.

Quantum

[John Courtland]

The Quantum Computers I've seen use a gas in a pressurized chamber with lasers to "read" and "write" the quantum states of the gas molecules. (Yes pedants, I realize this is far oversimplified, but I'm making a simple observation here) There is your gas processor, I suppose.

Re:Quantum

[Anonymous Coward]

see, I had to mod, then find this...

you're wrong. Quantum computers use NMR spectographs. Lasers are not what are used to do anything to the quantum states. And "gas molecules" is a bit of an odd thing to say, since anything can be a solid, liquid, or gas. Gas molecules? Like O2? CO2? No. Those aren't ideal. Caffiene supposedly is near the top of the list, though.

Re:Quantum

[Anonymous Coward]

You cannot "read" a quantum state in that way, or Heisenburg's cat will defile your lunch.

Re:Quantum

[John Courtland]

GODFUCKING DAMMIT! I put a fucking disclaimer in there for pedants and they still fucking reply!!!! AAAAH! YOU FUCKING MORONS!

Re:Quantum

[Anonymous Coward]

"If you can't say something nice, don't say nothing at all"
--Thumper the Rabbit

You think $500/gallon is expensive

[Anonymous Coward]

That's still cheaper than printer ink per gallon... :-\

Motherboard / Socket 940 fears

[trentfoley]

In January, I built an Athlon64 FX51 system. I debated the merits of getting the 940 vs. waiting for the 939 and decided that I will have to upgrade the motherboard in a year anyway. Why? PCI Express should be readily available, and market tested by then.

Besides, everytime I have actually upgraded a processor, I have also had to buy new memory and motherboard to accompany it. There's no sense putting a shiny new processor in and having it use old, slow memory on an old slow, bus.

However, just last week, I had to replace the mb in my kids' computer - I got a kt600 board in anticipation of upgrading their Athlon XP 1900+ and PC2100 at some point.

How much do you want to bet that by the time I need to upgrade that processor and memory, I will still need to buy a new mobo in order to utilize the new features of whatever processor or memory architecture I use.

BTW, I LOVE MY FX51 MACHINE! I built the whole system for just $3200 ($1150 was the display). For work stuff, I run Gentoo Linux compiled for AMD64, and for games, regular 32 bit Windows XP Pro. Absolutely no complaints. Wolfenstein Enemy Territory shows 60-90fps at 1600x1200 full detail.

Re:Motherboard / Socket 940 fears

[ipjohnson]

I don't know the FX51 is nice but for roughly the price of the chip you could get a nice tyan mb,2 athlon MP 2800 and gig of ECC ram. Now I know it doesn't have the 64bit extentions and its only pc2100 ram but all in all alot of horse power for not alot of money.

Re:Motherboard / Socket 940 fears

[trentfoley]

The dual channel ddr controller included on the FX51 makes a huge difference. I had to fill all 4 slots on my mobo to take advantage of it, but the performance increase is noticeable. PC2100 -- ugh.

BTW, I got my FX51 for $404 at gameve.com. A few hours after I ordered mine, the price was back to $750 or so. I'm a lucky shopper.

Re:Motherboard / Socket 940 fears

[ipjohnson]

As much as the onboard memory control increases speed I still don't think its as fast as 2 athlon 2800s. Not to mention most of us aren't going to get luck and find it almost half off.

Don't get me wrong I think the new 64bit chips are awesome and I'd love to get my hands on 1 or 2 of them but for me right now I'm going to go with 2 2800s.

This is old news

[mattgoldey]

Michael's Computers has had these for 2 years.

Re:This is old news

[Anonymous Coward]

score:-1, redundant

Explain to me...

[vasqzr]

Why does the performance vary so much on the individual test??

570fps vs 525fps in the Quake III test...Intel advantage...8%

2.78 minutes vs 2.48 minutes...11% in the 3D studio test faster with the Intel chip

Then in the compiling test....the AMD is ahead by about 15%!!

The only thing I can think of, is the Intel motherboards have a much faster graphics subsystem and the AMD's are much faster when it comes to disk i/o, and the Intel chips have a better memory i/o system.

Anyone have some insight?

Re:Explain to me...

[Anonymous Coward]

Just because the two CPU families understand the same instruction set doesn't mean that they're at all similar under the hood. Processors are extremely complex, and trying to logically puzzle out why things happen in exactly the way that they do is nearly impossible. That's why benchmarking, as opposed to logical analysis, is the preferred method of testing complex systems.

Re:Explain to me...

[Anonymous Coward]

Different chips have different strengths. 3D performance depends on good floating-point math performance and, in some cases, the presence of SIMD instructions. (SSE, 3DNow, MMX, etc.) Compilation, on the other hand, is all about discrete math. Incidentally, the reason that Macs have always been popular with the creative crowd has a lot to do with the fact that the PowerPC family of chips is very, very good at integer math, which tremendously benefits Photoshop, video compositing applications, etc.

Re:Explain to me...

[edxwelch]

"Anyone have some insight?"
Different architectures. Intel chips have longer instruction pipes that allow them to run faster, but have the penalty of worse branch prediction. Compilers have a lot of branches, which automatically favour AMD chips (same for Office apps). Intel will do better in MP3 encoding, video encoding, etc, where the raw processer speed gives it the edge.

Remember the Ergo Brick?

[BitwizeGHC]

If I recall, the "Brick", a tiny form factor PC from back in the day, dispensed with cooling fans by pressing a bag of Fluorinert right up against the mobo. The stuff had a consumer-level use too.

tests done with 64 bit OS and apps or not?

[usrerco]

Hope this isn't a RTFAC (C=Carefully), but I don't see in this article if they specify the tests were done with a 64 bit version of the OS and 64 bit binaries for the apps, or just a 32 bit versions.

Maybe I'm missing it, but in the "OS" category, they only say "Windows XP Profession SP1", no reference that I can see if it's a 64 bit version of the OS or not. Ditto for the apps.

My understanding is (with the current state of affairs, 03/18/2004) if you run to the store and buy an AMD 64 machine, you'll get a 32 bit version of Windows and apps.

But you only get the benefits of eg. >2^32 ram access IF you run a 64 bit OS and 64 bit apps.

Being able to access >2^32 of ram in an app is really useful for 3D rendering of very large projects (a business I happen to associated with), so if the tests are done with 32 bit OS/Apps, the benchmarks would seem to be not so useful for that purpose.

Considering this is a 64 bit processor being evaluated, it would seem lacking not to mention this.

Most folks in 3D evaling AMD 64's are sticking 64 bit os's on there right away (Suse, Gentoo, Fedora/Yarrow, etc), and doing tests with that.

Am I missing the part where they talk about 32 vs. 64 bit OS in these tests? I would /think/ it would make a big difference benchmark-wise.. correct me if I am wrong.

ps. With all those blinking flashing (*!&@# banner adds, it's often hard to RTFAC. I wonder, do schools now give reading comprehension tests in rooms with flashing lights and spinning graphics to simulate 'real world' scenarios? >;)

Re:tests done with 64 bit OS and apps or not?

[brokenbeaker]

with Anandtech, you can click on "print this article", and get the whole thing in a sinle window with nothing flashing.

anand blows

[Anonymous Coward]

anandtech plugging an amd processor?? im shocked!

his site is so biased it makes me sick.

Cant get your facts straight

[Anonymous Coward]

Its his number 2, not him.

Its Pakistani troops that are closing in on this man, not Americans.

And havent you noticed how GWB didnt exactly put a lot of effort into catching OBL until near the elections. It wouldnt surprise me at all if he has a pretty good idea where he is, but is waiting until days before the elections to do anything.

Someone who is willing to let people die in more terrorist attacks just to win an election makes me sick.

I can beat that.

[pr0ntab]

Same chip, OC'd to 3200. Zalman flower, a tiny bit of compound. Fan direction on the CPU so it directs the air away from the MB toward the 12cm slow-RPM exhaust fan. There is also a similar case fan in the front grill.

My CPU core peaked at 42C when running UT2K3, with distributed.net in the background.

Aluminum cases are a rip-off...

[pr0ntab]

They don't help you in any way help remove heat from the CPU. Aluminum may be a better conductor of heat than steel, but it has a lower heat capacity. This is important because what's on both sides of the aluminum case?
Air.
Which is a poor heat conductor.
You're ultimately limited by the ability of the external air to remove the heat from the outside of the case, which is effectively zero (especially if it isn't moving).
At least the steel case can better retain excess spot heat from the case, whereas the aluminum heats up uniformly, thus rendering it useless for further heat redistribution.

The ability of the case to conduct heat away better is only useful if it is connected to a yet larger heat sink, and unless you replace the rubber/plastic feet with copper pads and use it on the concrete floor in the basement...

It has nothing to do with the chip, and everything to do with thermal management.

Dell PCs don't have this problem because Dell PCs are specifically engineered to remove heat efficiently. Don't you wish you could have a custom air hood for your CPU/northbridge/case?

Stop smoking crack

[Anonymous Coward]

The nForce2 was a new chipset, so OF COURSE it takes some time for Linux to build in compatibility. The reason it worked flawlessly under Windows is because there are Windows drivers. It has nothing to do with the AMD chip itself.

Re:Stop smoking crack

[AKnightCowboy]

The nForce2 was a new chipset, so OF COURSE it takes some time for Linux to build in compatibility. The reason it worked flawlessly under Windows is because there are Windows drivers. It has nothing to do with the AMD chip itself.

I know that, but I bought this machine almost 9 months ago and support is still lacking. I don't dare try to turn on APIC mode or my machine will crash. All I'm saying is be wary of AMD chipsets (specifically Nvidia Nforce) under Linux.. they're very iffy in their compatibility whereas Intel chipsets are very well supported.

Re:Stop smoking crack

[Anonymous Coward]

Older AMD chipsets like VIA and SIS are just as supported as older Intel chipsets. When Intel releases a dramatically different chipset, expect it to take some time for it to work.

Re:Stop smoking crack

[kundor]

No. Intel is not better supported than AMD under linux. Specific chipsets do not a rule make.

Where are the native compiled tests?

[Kynde]

Where are the 64bit processor benchmarks where the tests have been compiled for them?

I mean, given that the x86 64bit decendants have more registers and all, running some stupid Sysmark or Unreal Tournament on top of them is like comparing V4 and V8 engines in such a way that the V8 only gasoline to four of it's cylinders.

What I want to know is the P4 flag ship lined up with the AMD 64bit flag shit on linux with a kernel compiled for 64bit and apps compiled for 64bits.

I have not been able to locate a single such benchmark as of yet. Anyone? Please...

Re:Where are the native compiled tests?

[Anonymous Coward]

http://www.spec.org

you'll find results for p4 flagship (intel compiler) and AMD opteron (both intel compiler, x86 and gcc, x86-64).

two things you'll note: 1) the p4 flagship is faster, even if it's a V4, 2) the opteron is faster with the intel compiler (ie good x86 code trumps GCC's x86-64 code)

Basically, the extra registers buy you a little (about 10%), but x86 has been around so long that compilers and the chip vendors alike have thrown so many little hacks in that they're actually doing a pretty good job in the register starved x86 environment.

Re:Where are the native compiled tests?

[Anonymous Coward]

...and I forgot to mention that anywhere you're doing a lot of pointer chasing, you're basically halving your cache size/bandwidth, and that offsets a lot of the accesses to memory that the extra registers save you from

Re:Where are the native compiled tests?

[Anonymous Coward]

There are some benchmarks (run on FreeBSD, performance should be similar under Linux since they use the same compiler) here.

Re:Where are the native compiled tests?

[Moraelin]

Let me tell you about my attempt about running a 64 bit Gentoo Linux 2004.0 system, with 64 bit apps. To start with a punchline: x86_64 Linux is the biggest practical joke ever. Not a flame, just a non-exaggerated story of my attempt at getting it to work.

Trying to install ATI's drivers? Nope. Sorry, Dave. Can't do that. They're 32 bit only. Can't run a 64 bit X11 with 32 bit drivers, now can you?

So now the whole thing ran on VESA framebuffer graphics. Yippee. So now I have an uber-fast 64 bit CPU, except even painting the mouse cursor takes ages. Oh the fun. The Performance. Or lack thereof.

Oh yeah, better get used to staring into a flickering 60 Hz screen, too. Because you can't set the refresh rate of VESA. Made me want to go buy a TFT monitor. (Then again, it got uninstalled before I got around to that.)

Compiling even the kernel drivers? Well, you know, 75% of those don't even appear in the list when you're configuring the kernel on an x86_64 machine. No AGP for my mobo, for example, if I remember right. (Not that it matters when using the VESA framebuffer, I guess.)

Compiling some cool uber-fast 64 bit apps? Huston, we have a problem. The _vast_ majority of those don't even want to "emerge" (the Gentoo command to download and compile something) for the x86_64. Not OpenOffice Org, not Pingus, not almost anything.

Now, just to be fair: none of that is, strictly speaking, Linux's fault. Can't really blame Linux for the fact that ATI doesn't write 64 bit drivers, or that apps are written by people who think everything is 32 bit.

And also to be fair: Microsoft's 64 bit XP doesn't fare any better in that aspect, either.

But, you know, having an OS without drivers or apps is about as useful as having a pen, but neither ink nor paper. Again, that applies to Microsoft's 64 bit XP just as well.

So, dunno, I'd like to hereby apologize to all the Mac zealots I've picked on before. I'm sorry, guys. Yes, if you really have to have 64 bit Unix system today, forget about Linux. Get a Mac.

Re:Where are the native compiled tests?

[Kynde]

Trying to install ATI's drivers? Nope. Sorry, Dave. Can't do that. They're 32 bit only. Can't run a 64 bit X11 with 32 bit drivers, now can you?

Naturally that stuff can't be up to date just yet. Although nvidia does offer both ia64 and amd64 drivers. Bleeding edge never is well supported in oss world. My point wasn't about that, because by the time I'll be purchasing 64 bit x86 processor, perhaps some time next fall, the support side has gone a long way.

Now, just to be fair: none of that is, strictly speaking, Linux's fault. Can't really blame Linux for the fact that ATI doesn't write 64 bit drivers, or that apps are written by people who think everything is 32 bit.

And also to be fair: Microsoft's 64 bit XP doesn't fare any better in that aspect, either.

Agreed, but we are talking about 64 bit benchmarks here aren't we. I wasn't that interested about the usefullness of 64bit processors in march 2004, I was wondering what the impact of 64bit native compilation along with 64bit kernel can do. Because just by showing some framerates of Unreal Tournamen compiled for 32bits on top of a 32bit windows system didn't quite reflect my future use. (after all, eventhough the support for 64bits may be poor right now, the decision about the purchase will naturally be based on how they _will_ perform).

Re:Where are the native compiled tests?

[Moraelin]

Well, I don't know, I thought that these benchmarks are supposed to be more useful for people buying a CPU now, in Q2 2004, than for people who'll buy one in a year and a half. Or at least _my_ decision to purchase something is based on how it performs now, rather than on how it _might_ perform in some future.

How the Athlon64 will perform in Q3 2005, well, I'll worry about that in Q3 2005. A year and a half is a long time. A lot of things can happen until then. New compilers, new apps, better drivers, a better X11 architecture, people starting to write x86_64 assembly. No idea what.

Basically IMHO at the moment the main use for an Athlon 64 is for running 32 bit apps on the 32 bit OS of your choice. For some it'll be Windows, for some it'll be Linux, but unless you're masochistic it'll be 32 bits. Hence, the benchmarks will, naturally, tend to be in that context.

A Nasty Virus

[Anonymous Coward]

From the article:

Enhanced Virus Protection for the upcoming Windows(R) XP SP2

Finally, a CPU with integrated features to protect you against Windows XP!

No, you listen, jackass

[Anonymous Coward]

Listen "jackass", it's done to save on bandwidth.

how about a review w/out P4EE or 64FX??

[freakmaster]

both of these processors are priced WAY of the pricing curve for the VAST majority of people. They are really server chips in disguise. In both cases they cost twice as much as the chips in the 'standard' lines & they only offer a few percentage points performance increase. Heck for the price of 1 or these 'elite' chips you could buy TWO athlon MP 2800's w/ a super reliable & high quality TYAN motherboard (PCI/PCI-X AND AGP). I'm pretty sure a system like this would crush any single cpu system in a compiler benchmark, that's for sure! That's 5.6 virtual gigahertz!

I suspect that these chips exist entirely for marketing purposes & are there really to have the 'fastest desktop cpu on the market title'. But a much more VALUABLE question to ask is 'what's the best chip for $400, $300, $200, $100?, heck even $50!!'.

every time i look at these benchmarks, i have to cut out the EE's & the FX's & just see how the P4's compare to the althon-64's & XP's. because in my mind these 'elite' chips don't exist. They're toys. there's no remotely reasonable reason to buy them.

cool your innards the old fashioned way...

[-O.ster_66]

have them laying about the tower in heaps dictated by the length of your cables. i lost my side panels about 2 years ago and the last thing i remember was trying to get them on only to find that i had a LOT of my cables wrapping outside the cases support members. since then i've not got a single thing screwed to my case. "my computer" refers to the pile of sparking wires under and around my desk. my fans basically just keep the air from getting stale around my desk :)

Googlewhack!

[MikeCapone]

That's the word I was after...

Proxomitron version of your review

[Quizo69]

"But Tom's review is a little more drawn out"

Seriously, if you aren't filtering the web, I'd say you're not much brighter than those who don't filter email for spam. It's the same crap, just fed through a different medium.

www.proxomitron.info

or

www.privoxy.org

Do yourself a HUGE favour!

Re:what's the point?

[edxwelch]

"Extreme Gaming Edition", "Hyperthreading"
Also they're in danger of running out of superlatives. What do they call the next one? "Super-Extreme Gaming Endition", then "Super-fucking-shit-your-pants-really-fast Gaming Edition"