ZDNet Says AMD Posts Blatantly Deceptive Benchmark

Glasswire writes "George Ou, writing in ZDNet's Real World IT blog, accuses AMD of comparing processors the company will not be shipping for months (2.6GHz Barcelona quad core) with older Intel Xeon quad cores rather than currently shipping ones which would beat the (hypothetical) score AMD claims for the future Barcelona. I guess while even the much slower 2.0GHz Barcelona is due soon AMD didn't think results from the 2.0 would look good enough — even against the slower Xeons they picked. Maybe the right comparison should be either best cpu against best cpu — or compare ones at the same price — and only shipped products."

Comments

Who trusts a vendor's benchmarks anyway?

[nweaver]

Vendor benchmarks are always considered untrustworthy, so I don't see what the big deal is.

Re:Who trusts a vendor's benchmarks anyway?

[PFI_Optix]

If we don't point out every time they use blatantly unfair product comparisons, the amount of disinformation coming out of vendors will only increase. Even though very few people (just the fanboys) place any stock in AMD's or Intel's benchmarks, it's worth pointing out flaws like this to keep them as honest as we possibly can.

Re:Who trusts a vendor's benchmarks anyway?

[yfarren]

The point is, someone is lying.

While it may be the case that marketers generally lie, that is something to be opposed.

When people lie, when people disseminate false information, it harms the public. That people do so a lot simply means that they are hurting the public a lot. To say "Well, everyone harms the public, why is it a big deal that this person is harming the public" is to say it is ok to harm the public.

It isn't. Lying, disseminating false information is harmful. If it is done a lot, that just means there is a lot of harm being done, and should be opposed by the public MORE strongly.

To become blase' about people who lie and mislead simply encourages people to lie and mislead. It means that someone who tells the truth wont actually be listened too, because "well everyone lies". Which makes it more difficult for someone who does tell the truth.

I would suggest you re-examine your values, and whoever modded you up should re-examine their values. Accepting lies as a Fait Accompli, and just assuming everyone lies, as opposed to holding liars accountable for the lies they tell, simply encourages liars, and makes it even harder for someone to tell the truth (which is often more expensive than lying), as they wont be believed anyway.

Re:Who trusts a vendor's benchmarks anyway?

[suv4x4]

Vendor benchmarks are always considered untrustworthy, so I don't see what the big deal is.

Hm, what a nice example of self-fulfilling prophecy.

I prefer another one though: accept Intel and AMD's benchmarks should be accurate to utmost detail, and put up a riot every time we see they aren't.

Re:Who trusts a vendor's benchmarks anyway?

[baggins2001]

No kidding
I don't think the accuracy of the benchmarks is in question here.
The deceptive manor in which the benchmark data was presented is the issue. Which is really a none issue. This is advertisement, anybody who doesn't look critically at data presented by the manufacturere is really gullible.
Anybody who doesn't look critically at the data from a third party is pretty gullible
I really really don't see the problem here

Re:Who trusts a vendor's benchmarks anyway?

[CatsupBoy]

If we don't point out every time they use blatantly unfair product comparisons, the amount of disinformation coming out of vendors will only increase

Or the amount of crap product comparisons will continue to be the same no matter how much its pointed out.

Companies will continue to tout themselves as top dogs, regardless of the facts. And it never ceases to amaze me how far they go to stretch the truth in order to make themselves look good.

How else could salesmen go into a room and pitch their product? Or how can manufacturers sell their AMD products when competitors are pushing Intel? And vice versa? Its capitalism at its best.

Re:Who trusts a vendor's benchmarks anyway?

[Anonymous Coward]

If you read the comments on TFA, apparently the chart was originally posted in April. At the time these results were posted they were believed to be true (albeit based on "estimated" performance.) Same with the supposed WSJ ads. Since then the clock speed of the AMD has come down and the old scores have changed. It seems that AMD's real mistake here was not updating the information on their site.

The skewed numbers for Intel's chip are also because the chart for Intel was compiled with different settings (and possibly a different compiler version altogether.) This information all comes from the comments, so while not verified, they also weren't really disputed in the comments either.

Re:Who trusts a vendor's benchmarks anyway?

[mgoheen]

> Vendor benchmarks are always considered untrustworthy, so I don't see what the big deal is.

That logic gets you into trouble...

Politician promises are always considered untrustworthy, so I don't see what the big deal is.

Auto companies are untrustworthy, so you should expect the brakes to fail.

People are untrustworthy, so if you are robbed, it's your fault for carrying cash.

People are killed every day, so I don't see what the big deal with Iraq is.

etc.

Sheesh...wrong is wrong, no matter who is doing it. If you don't fight it, you're part of the problem.

Re:Who trusts a vendor's benchmarks anyway?

[arth1]

Still, the submission misses the point completely when it want benchmarks of only CPUs that's on the market. The reason for the benchmarks of CPUs that haven't been released yet is so OEMs and retailers know a little more of what to expect, and make plans for ordering (or not ordering) accordingly. If there's no benchmarks of unreleased CPUs, it would not hit the market, and thus wouldn't be benchmarked -- catch 22.

Who the manufacturer compares against is of course up to them, and there's nothing "unfair" about it. It's telling the world that this is the competition they strive to beat. If it's an older CPU, the new CPU is obviously intended as a replacement for these. If I had a large server farm running these Xeons, I'd be most interested to see this benchmark, well before the CPUs actually come out (if they're already out on the market, they will be off the market by the time upper management approves the budget). And remember, AMD and Intel aren't in the game to try to trick you to buy a CPU that won't work well for you -- they want you to return for your CPU needs, over and over again. That's why they publish benchmarks like these, which are relevant, just not to the GP.

Other comparisons both will and do appear once a CPU has hit the market. But for the initial pre-release vendor benchmarks, I'd rather it be the choice of the vendor, so we can see where the market position is going to be.

Move along -- nothing to see here, except for a particularly silly submission.

Re:Who trusts a vendor's benchmarks anyway?

[PFI_Optix]

Or the amount of crap product comparisons will continue to be the same no matter how much its pointed out.

You don't think it can get worse? You don't think it would get worse if there weren't people crying foul at the current comparisons?

You can use legitimate comparisons to tout a product, you don't have to unfairly match them. Look at your average car commercial (fictional example):

Ford's new truck gets better gas mileage than Dodge.
Ford's new truck has a bigger, more powerful engine than Chevy.

They just said it's better than Dodge and Chevy, but in two completely different ways. They do this all the time in marketing. If nothing else, AMD could talk up price points and power efficiency, two things they almost always have over Intel. Skewed benchmarks just make the company look inept and leave knowledgeable consumers feeling like AMD is insulting their intelligence.

Re:Who trusts a vendor's benchmarks anyway?

[Anonymous Coward]

An amusing case and point:

From XGI's 'where to buy' section (apparantly, nobody is selling them!)
A market bottom-dog-pack-runt

Want to deliver the market's most competitive product to your customers? Want to stay ahead from the rest? Please contact us and we will provide you with the most outstanding graphics chipset solution from XGI.


if you don't know, they are a video card manufacturer. The reason I was even on their site, is I remembered they had OSS drivers that they released (apparantly only 2D), but I was looking for something that wasn't nVidia (no 64 bit drivers in BSD), or ATI (horrible experiences), and currently the X3100 chipset (or the X3000 with DVI) is not available from Intel yet...

But back on the topic, you are quite correct, even the bottom of the totem poll will do that kind of phallic waving to suggest they are the top.

Re:Who trusts a vendor's benchmarks anyway?

[Anonymous Coward]

If you read the comments on TFA

You must be new here. Welcome to /.

Re:Who trusts a vendor's benchmarks anyway?

[Applekid]

Thing is, if the new CPU truly is the next greatest thing since shaved ice, they wouldn't need to be deceptive. They'd use fair and unbiased benchmarks and their CPU will float to the top.

The deceptive benchmark is indicative of one thing: the new CPU blows. Otherwise it wouldn't be needed.

What I would be concerned about is whether this was put forth by the marketing teams to cover up a costly R&D backtrack or it was put forth by the lead engineering groups to save their butts from the fire from the executives since they wouldn't know a good CPU if it burned them in the butt*.

* Not a direct attack on executives since, by definition, they're there to RUN a business and not actually develop FOR the business, so, knowing a good CPU is simply out of scope and their only references are what their employees tell them.

Re:Who trusts a vendor's benchmarks anyway?

[arth1]

If someone is lying, could you please care to name a lie?
Cause I can't see a single lie. Self-flattery, yes, and selective truths, yes, but no lies.

If you're in the business (and if you're not, this type of benchmark isn't meant for you), you know very well how to read and interpret the reported benchmarks and marketese. It's the expected format, which is helpful to those who need to know these things, e.g. because they are planning on upgrading a large Xeon farm to faster CPUs at as low cost as possible, or because they're a large OEM who needs to know the market segment this CPU is intended for, so they know both how much to order and how to market it.

Can we all stop this lynch mob now?

Re:Who trusts a vendor's benchmarks anyway?

[arivanov]

Absolutely.

And the fact that the CPU is not going to hit the high street for 6 odd months does not mean that selected engineering samples cannot be clocked to the same frequency. So in fact, the test is most likely run on a real CPU. Even further, if it is shipping in 6 months to stores the engineering samples have to hit OEMs and major manufacturers now so they can verify their designs.

Oh, and by the way, both AMD and Intel do this all the time. Intel was publishing Core benchmarks for 3-6 months ahead of launches. If we dig around their site I bet that we are going to find at least one benchmark for a CPU that is yet to be officially released.

Re:Who trusts a vendor's benchmarks anyway?

[yfarren]

Comparing a product that I (may) produce, 4 months from now, to one that someone else, did produce, 4 months ago, in a rapidly changing market, to imply that "My chips are better than their chips" is lying.

To say "oh, selective truths to imply an untruth" isn't lying, is to play a stupid, and harmful semantic game.

This isn't a lynch mob. It is people pointing out that AMD is lying in their use of the "expected" format, to say something, which isn't true (our chips outperform their chips).

It is people saying "we are tired of people abusing our trust, and using bad numbers to imply things which are false".

To say "well, they aren't saying anything which isn't true, my future chips do outperform their former chips" is disingenuous. It may be, literally true, but the implications of that graph are misleading and therefore a lie.

This isn't helpful to some buyer, as it doens't compare two comparable items. My not yet released chips, are not comparable to their old chips. That you suggest something else, makes me wonder who you are writing for.

Re:Who trusts a vendor's benchmarks anyway?

[Surt]

Actually, scientific research has proven that a little bit of lying is good for the public welfare. The question is really: is this too much lying?

http://scholar.google.com/scholar?hl=en&lr=&safe=o ff&client=firefox-a&q=benefits+lying&btnG=Search

Re:Who trusts a vendor's benchmarks anyway?

[CatsupBoy]

You don't think it can get worse? You don't think it would get worse if there weren't people crying foul at the current comparisons?

No, I dont think it will get much worse because companies will say as much as they can, just shy of getting sued. Whats the point in being number 2 in any market? You have to make the customer think your product is the best purchase out there, no matter what reality is.

You can use legitimate comparisons to tout a product, you don't have to unfairly match them.

Only true if you have something better then the other guy. What if you dont? Then you either make stuff up or go say to your investors "Sorry folks, but we just got beat. Oh by the way, can we have some more money?". Not likely.

Its easy for car companies to say various things are better then the other guy because there are 1000+ different points to consider when buying a car. But were talking about two CPUs that pretty much do the exact same thing. Either it outperforms at a better cost or it doesnt. And if it doesnt who the hell would put money on it? (besides the fanboys).

Skewed benchmarks just make the company look inept and leave knowledgeable consumers feeling like AMD is insulting their intelligence.

Companies only look inept to knowledgeable consumers. If you consider that a dying breed, then their only insulting a small number of people. Their playing the margins, and most consumers probably buy into it.

Re:Who trusts a vendor's benchmarks anyway?

[Anonymous Coward]

You seem to have missed the part where they compared them only to older Intel CPU's, and conveniently ignored the fact that current Xeons could beat them in benchmarks. Fanboy much?

Re:Who trusts a vendor's benchmarks anyway?

[cbreaker]

And, not ONLY all that, all of these enthusiast sites continually post overclocked benchmarks for these CPU's.

They used to do it with the Pentium 4 all the time; You'd see a currently available Athlon versus a currently available Pentium 4 in a bechmark chart, and next to it would be a 60% overclocked P4 that requires special cooling. Yet they'd always say "BUT The OverClocked one BLOWS AMD AWAY!"

Just because this is coming from a manufacturer doesn't make it any less valid, and I don't see why AMD has to go hunting for Intel's latest CPU with the same model number (but a different revision) just to keep things fair OUT of their favor.

Besides, all this SPECint and CPU benchmark crap is worthless anyways, unless all you do with your server is run scientific calculations. In real world SMP applications, such as heavy-use VMware servers or database servers with lots of I/O and RAM, the Opterons will always kick the crap out of the Intel boxes with the Northbridge bottleneck. HyperTransport is the key to actually USING all of those system resources.

Re:Who trusts a vendor's benchmarks anyway?

[arth1]

Comparing a product that I (may) produce, 4 months from now, to one that someone else, did produce, 4 months ago, in a rapidly changing market, to imply that "My chips are better than their chips" is lying.

Really? You have a very strange definition of "lying", then. I think it shows how good a replacement the new CPU would be compared to the older one, but what do I know?

To say "well, they aren't saying anything which isn't true, my future chips do outperform their former chips" is disingenuous. It may be, literally true, but the implications of that graph are misleading and therefore a lie.

How is it misleading? It's a very good indicator on whether the future CPU would be a good replacement for the old CPU, and that is useful information to many -- both large companies and OEMs. The only misleading here seems to be people misleading themselves into thinking the benchmark is for a different purpose than it is.

This isn't helpful to some buyer, as it doens't compare two comparable items. My not yet released chips, are not comparable to their old chips. That you suggest something else, makes me wonder who you are writing for.

Me, myself and I. I refuse to join a lynch mob without thinking things through first.

Re:Who trusts a vendor's benchmarks anyway?

[Qrlx]

Au contraire, you need to re-examine your values. In almost all cases, lying is protected First Amendment speech.

Re:Who trusts a vendor's benchmarks anyway?

[arth1]

You seem to have missed the part where they compared them only to older Intel CPU's, and conveniently ignored the fact that current Xeons could beat them in benchmarks. Fanboy much?

No, I'm far from a fanboy. And no, I didn't miss that part. In fact, I explained why they compare the unreleased CPU to an older Intel CPU, and why that makes sense. Apparently you either skipped this, or didn't understand it, so I guess I have to get a silver spoon and bib for you:

When a CPU hasn't been released yet, what's important to the large scale buyers is how it fits in with their business. If they now sell cheap servers with Xeons, would this be a feasible replacement that could increase sales and profits and/or reduce costs? A comparison with the "old" Xeons is thus very relevant. It also tells exactly which market segment AMD intends this CPU for, which is quite valuable information. Not getting benchmarks until the CPU has been released makes it much harder to make decisions like these. Similarly, there's little value in having a CPU compared to the "state-of-the-art", when you wouldn't purchase it as a state-of-the-art replacement anyhow.

It may be of interest to the casual hobbyist, but that's not the big market.

Re:Who trusts a vendor's benchmarks anyway?

[Anonymous Coward]

Nobody should trust a manufacturer's benchmark since they will skew the results they want you to see. I used to work at an large software company and I seen this many times where they will tweak the machine get the best result but this machine is not a "standard" machine that can be used normally. I would like to see several third party companies do test on several configurations so you have near "real world" benchmarks.

Re:Who trusts a vendor's benchmarks anyway?

[adisakp]

Cause I can't see a single lie

You can imply a lot of things that are completely untrue without saying a single lie.

Re:Who trusts a vendor's benchmarks anyway?

[Idbar]

If we all believed in those marketing strategies... we'd all had Apple computers with Windows Vista on them.

Computer jokes aside, I'm particularly afraid of those commercials about depression and how you should talk to your doctor to prescribe you medicines for it, I think it's pretty scary that they are trying to convince you about your moods and get money out of it.

Re:Who trusts a vendor's benchmarks anyway?

[Anonymous Coward]

More fool you, then!

Re:Who trusts a vendor's benchmarks anyway?

[heinousjay]

wrong is wrong, no matter who is doing it

Wrong is subjective, depending on who is interpreting it. To state otherwise is to be the cause of the problem.

Re:Who trusts a vendor's benchmarks anyway?

[arodland]

Er no, quite the opposite. To state that is to be the cause of the problem. Notice that you can't even make your point without falling into your own trap. Your argument against a categorical morality is that believing in one is categorically wrong. :)

Re:Who trusts a vendor's benchmarks anyway?

[Anonymous Coward]

Wow, what is that you're wearing? Eau de AMDFANBOY?

Re:Who trusts a vendor's benchmarks anyway?

[logicnazi]

Actually the test results themselves say they were the result of internal simulations. Still, assuming they were done honestly, internal simulations are probably more accurate than test silicon.

The real story here is not that "AMD LIED." Parent comments are right that AMD did not make any false statements. They were, however, misleading but I would normally let that slide for advertising.

The story is that AMD slammed intel for being deceptive and turned around and did it themselves.

Re:Who trusts a vendor's benchmarks anyway?

[cicatrix1]

Glad to see you've got your auto-reply-with-standard jokes perl script working. Congratulations!

Re:Who trusts a vendor's benchmarks anyway?

[Anonymous Coward]

"In real world SMP applications, such as heavy-use VMware servers or database servers with lots of I/O and RAM, the Opterons will always kick the crap out of the Intel boxes with the Northbridge bottleneck."

Gross exaggeration. Xeon boxes these days are really good as database servers.

Re:Who trusts a vendor's benchmarks anyway?

[Anonymous Coward]

You really, enjoy using, random commas, don't you?

Re:Who trusts a vendor's benchmarks anyway?

[BlueTrin]

IMHO, first amendment is mostly about protecting political ideas, bullying your neighbour, calling your ex-girlfriend's friends to say lies and tell them how much she was a sicko are hardly things that are protected by the first amendment.

I doubt also that the first amendment was written to protect the marketing departments from being sued for lying in order to increase their sales ...

Re:Who trusts a vendor's benchmarks anyway?

[sfarmstrong]

People are untrustworthy, so if you are robbed, it's your fault for carrying cash.

People expect benchmarks to have "sales puff" in them, so they know to greet the comparisons with a grain of salt.

People expect city streets to have churlish thugs in them, so they know to greet the robbers with a grain of assault. :)

People are killed every day, so I don't see what the big deal with Iraq is.

Now that's just silly. If I know that I'm in a spot with inflated death totals, I just need to take some extra lives with me. Death is just another marketing scam from the Wintel conspiracy.

I can smell the desperation

[Urusai]

Core 2 is smoking AMD and they are panicking. Do they even have a real next gen architecture, aside from bizarre (albeit intriguing) CPU/GPU hybrids?

Re:I can smell the desperation

[rstarg]

I hope that AMD can get their act together - I would hate to see Intel as the only game in town.

Re:I can smell the desperation

[moore.dustin]

Only if they deserve to be there. Having AMD around just to it around is pointless. Intel is competing and dominating right now. If AMD cannot survive because of it, so be it. they deserve to go. Of course, I am not advocating for a monopoly, but if AMD cannot compete, I am all for them failing.

Re:I can smell the desperation

[Zantetsuken]

Yes, and it is this very "Barcelona" K10 and "Phenom" - which unlike Intel's "Quad-core" which is 2 dual-core processors on one 2 socket motherboard, AMD's approach will actually have 4 cores on one die. This means AMD doesn't have the same self-imposed bottleneck that Intel put on themselves by using 2x socket (the bus speed between the 2 is the bottleneck).

Re:I can smell the desperation

[Anonymous Coward]

It really isn't, you know, at least not at the high-end - there, AMD's hypertransport bus makes it far more attractive. Core 2s are okay for gamerz and desktops, pretty crappy in a climate simulator or whatever.

Re:I can smell the desperation

[Phu5ion]

It takes time for either side to release a new arch. How many years was AMD smoking Intel before they came out with the Core 2 series? AMD fanboyism aside, the Core 2 is a very great platform, but you have to thank AMD for forcing Intel to step it up, otherwise we would be paying twice as much today for a chip that performs about as well as the old P3. Competition is good and I, for one, can't wait to see what AMD comes up with next.

Re:I can smell the desperation

[Anonymous Coward]

Only game in town? What about VIA?

Re:I can smell the desperation

[RingDev]

You should head over to the wikipedia and read the history of AMD. It's an interesting read, and it might shed some light on the whole AMD vs Intel deal.

-Rick

Re:I can smell the desperation

[Anonymous Coward]

VIA doesn't matter. It's CPUs are aimed at a different market. Guess whose CPUs they use for their performance epia line? I'll give you a hint, it begins with an I.

Re:I can smell the desperation

[Anonymous Coward]

I think you're mixing something up, there...

Re:I can smell the desperation

[LWATCDR]

Well for one I don't think Intel is dominating right now.
In the 8 or more core server market AMD is still the better choice.
On the desktop AMDs X2 3800s are a great choice for about 95% of the market.
Where Intel really is dominating is the Laptop market. That is an extremely important market that AMD just can not seem to do well in.
As to wanting AMD to fail. Good greif I hope not. If Intel didn't have AMD we would still have P4s and be paying through the noise for them!

Re:I can smell the desperation

[moore.dustin]

Oh I agree, I was just saying that if a company cannot compete it does not deserve to be around. I do not foresee AMD going anywhere anytime soon and we all benefit from that. Was merely making the case that a company does not deserve to be around just to be around.

Re:I can smell the desperation

[GuyverDH]

Only to the bit-bucket... Too many processor bugs that can't be fixed via bios/firmware. Not saying that AMD doesn't have there's as well, but rushing hardware out only costs more than it makes in the long run if bugs like those creep into the product line.

Re:I can smell the desperation

[petermgreen]

hmm do intel have any quad core multiprocessor capable chips in production or do you still need to go to four sockets (read: very expensive case and motherboard) to get 8 cores with an AMD system?

Re:I can smell the desperation

[toddestan]

Core 2 is smoking AMD and they are panicking. Do they even have a real next gen architecture, aside from bizarre (albeit intriguing) CPU/GPU hybrids?

Why would they be panicking? For years, they were a very strong #2, completely dominating the budget CPU segment. The reason they were #1 for a while has more to do with how weak the P4 line was more than anything else. They still have the budget market, and unlike years past they now have OEMs like Dell on board and a pretty decent server CPU. AMD seems to be in a pretty good position overall.

Re:I can smell the desperation

[banesong]

Actually, yes they do. I currently posses a HP server that has a 4 core Intel processor with a second available slot. They do have the ability to go eight way with two slots. ~T

Re:I can smell the desperation

[LWATCDR]

Nope but then 8+ core systems are not cheap systems. You really don't need one to read your email, play games, or surf the web.
Okay if you are running Vista and outlook then maybe you do :)

An AMD 8 core system with four duel core chips will beat an Intel 8 core system with two quad core cpus.
When AMD ships quad core chips things could get very interesting.
Simple truth is that AMDs new X2 3800s really seem to be a great CPU for 95% of the market. It is cheap, fast enough, and very low power.

I choose AMD for the price...

[Spy der Mann]

i'm really not that interested in benchmarks. Besides my personal position is that AMD are the "good guys" and Intel are the "bad guys" because of their monopolistic practices.

It's kinda hard when you see your "heroes" do bad things, and I feel tempted to give excuses. In any case, the news won't make me trade my 3800+ dual core Athlon 64 for an intel Core 2 duo of the same speed and have to pay twice the price.

Re:I choose AMD for the price...

[bealzabobs_youruncle]

Or course it won't cause you to "trade my 3800+ dual core Athlon 64 for an intel Core 2 duo of the same speed and have to pay twice the price" becuase none of those circumstances are possible. I like AMD too, but they got owned this round and that is just the way it is. AMD is just as capable of evil and more guilty of whining, brand loyalty is for suckers in regards to performance desktop computing; buy the fastest gear you can get at the moment of purchase.

Re:I choose AMD for the price...

[QuantumPion]

AMD and Intel are CPU manufacturers, not sports teams. Buy the product that is the best performing at the lowest price.

Re:I choose AMD for the price...

[locokamil]

I've currently got a couple of 3800+ based dual cores doing server duty and media recording, and a 5200+ for my main desktop rig. I'd spring for Core2's... except for the fact that they cost twice as much and don't add much value to my home network. Basically, AMD beats Intel in terms of value, but loses the "brute force" race.

Re:I choose AMD for the price...

[Jeff DeMaagd]

In any case, the news won't make me trade my 3800+ dual core Athlon 64 for an intel Core 2 duo of the same speed and have to pay twice the price.

You shouldn't replace something you already have with something that isn't any better. Who is suggesting you do that? That would be a pretty stupid thing to do. Someone that has the Core 2 Duo would also be stupid to replace the system for an Athlon 64 of the same speed, both ways, it's spending money to get no improvement.

Re:I choose AMD for the price...

[timeOday]

I mostly agree, but customers big enough to influece market dynamics, like Dell or the US Government, should think about how awful it's going to be buying a computer in a few years if AMD falls out of the race. Personally I'm a little worried about it.

Re:I choose AMD for the price...

[Zantetsuken]

OTOH, perhaps an Athlon x64 3800 x2 is overkill in speed for whatever purposes he uses it for, meaning he actually doesn't have any need for a faster Intel Core 2 Duo?

Also, AMD may not quite be the absolute fastest, but last I looked on Newegg (around last week) they certainly were the cheapest. And I don't mean only the processor was cheapest - each time I spec out parts for a motherboard/cpu/RAM/etc (that is, using a minimum standard of manufacture and processor class being Athlon/whatever Intel's equivalent is called now) using an AMD socket motherboard is AFAIK every time cheaper than an Intel solution...

Re:I choose AMD for the price...

[freedumb2000]

That is true but should all be still hoping for a comeback of AMD. No one really wants the early 90s back where Intel ruled the block with no competitor, but high prices and complete control of if and when raise x86 CPU speeds to the next iteration. There is a reason why Intel still fights them tooth and nail. There where others before AMD but it's the only and first successful competitor in the x86 space for Intel. Imagine if there where only ATI or Nvida.

Re:I choose AMD for the price...

[Anonymous Coward]

AMD and Intel are CPU manufacturers, not sports teams.

And why is it that the argument is suddenly justified if you're talking about sports teams? It's all a bunch of follow-the-leader rah-rah-rah-ing as far as I'm concerned.

Re:I choose AMD for the price...

[morgan_greywolf]

Indeed. A quick look at Pricewatch will confirm this. In terms of bang for the buck, AMD is much better. An Athlon 64 x2 3800 at 2.0 Ghz mobo combo will set you back ~$115-135, while the equivalent core 2 duo system (IMHO), a e4300 will set you back ~$155-175, at the time of this writing.

Re:I choose AMD for the price...

[Rakishi]

The e4300 seems closer performance wise to a 4600+ than a 3800+. The 4600+ is almost the same price as the e4300.

So while AMD has cheaper processors that doesn't mean they are a better bang for the buck. To be honest if you have no use for the performance then just buy a used system of some sort.

Re:I choose AMD for the price...

[Rashkae]

Of course it makes sense not to replace perfectly working, top of the line systems.. but where do you get this figure that Intel costs twice as much? Intel CPU's now cost *less* than equivalent performing AMD CPU's (Although, the figures kind of even out with all of AMD's recent price cuts.)

Re:I choose AMD for the price...

[darkwhite]

I'm worried too, but AMD is not a startup. It's just as old as Intel, has diversified numerous times in its history, and has been at war with Intel for longer than you think. It's been through much worse times and is very used to playing catch-up. I'm sure its current financial situation is nowhere near the worst in its history, either.

I question their management in the wake of the ATI purchase, but I'm not the one with 35 years of competing against Intel under my belt.

Re:I choose AMD for the price...

[Chandon Seldon]

Intel CPU's now cost *less* than equivalent performing AMD CPU's

The best price/performance at the moment is the AMD 3600. From there, the price performance ratios get progressively worse as performance increases. Other processors that aren't blatantly obsolete include: The AMD 5000, Intel 4300, AMD 5600, and a bunch of Intel processors costing over $200.

Re:I choose AMD for the price...

[Chandon Seldon]

Sorry... "obsolete" wasn't the right word. I was looking for "superseded by better deals".

Re:I choose AMD for the price...

[hurfy]

Wish i could have upgraded my computers this year in the same way i built them, but i could only afford to rebuild one :(

Years ago when i built my ubercool mod boxes i could not decide on Intel vs AMD...so i built one of each :) Turns out that an Athlon 2400 was indeed a tad faster (at gaming anyway) than its Intel counterpart (P4 2.5) I am guessing it would be a wash if the P4 had HT like the newer ones. So at least their old naming scheme was pretty close.

Glad i got 4 years+ out of those boxes with only 1 GPU change and an extra memory chip. I had no need for hardcore systems at the moment so i snagged a super cheap Core Duo left over from work :)

Re:I choose AMD for the price...

[danpsmith]

It's kinda hard when you see your "heroes" do bad things, and I feel tempted to give excuses. In any case, the news won't make me trade my 3800+ dual core Athlon 64 for an intel Core 2 duo of the same speed and have to pay twice the price.

I'd kind of like to know where you are pricing processors. I bought a low end core 2 duo processor that beat the AMD processor on performance at the same price in a Dell computer. Maybe if you are making your own box, but whatever. The prices aren't that different. In some cases AMD is more expensive. As a side note, anyone who includes in their list of heroes any corporation seriously needs to get out of the house more often. None of these corporations are worth idealizing.

Re:I choose AMD for the price...

[ozbird]

Buy the product that is the best performing at the lowest price.

If you are purely a consumer, yes. If you want to ensure the continuation of the competition that lead to the current cheap Core 2 Duo vs. Athlon 64 X2 battle, you need to factor that in to the equation too. Otherwise, we'll be back to the days of paying through the nose for overpriced Pentium x shite.

Re:I choose AMD for the price...

[Matimus]

In the early 90's Intel competed against AMD, Cyrix and Via (maybe others too). If anything, it is worse now. As far as I know, there is only AMD, Intel. Via has an offering, but it doesn't seriously compete with the others in any of the major market segments. People don't remember anyone but Intel because their "Intel Inside" campaign was so wildly successful.

Re:I choose AMD for the price...

[CCW]

Yep, Dell is smart to be sourcing from both now. Any large company ought to spend at least 10% on the currently unfavored processor, since they save far more than 10% by having a competitor exist. Similarly, they ought to spend about 10% on the unfavored software component. Having a "Linux Desktop Development Team" and a "PostgreSQL implementation group" easily pays for itself come license renewal time, provided the competitor is plausible.

The lack of support from AMD for free software has made me switch to being an intel customer for my next build though. The price/performance of the alternatives isn't different enough to be compelling, but the lack of drivers for ATI chipsets and the good drivers for Intels chipsets make it a no brainer as to which is better. If you want to run Linux, want to try out linux, want to resell your machine to somebody who might want to use Linux - you're way better off with intel today. AMD could fix that easily, but they don't seem to be interested. Intel has a whole team of people churning out free source code for people who want to use their hardware - I'm going to support that with my wallet.

Re:I choose AMD for the price...

[tcc3]

Ok so when the athlon and opterons were eating Intel chips for breakfact and exacting prices to match the story was "Amd has faster chips!"

Now that Intel got their act together and are cleaning AMDs clock its "They are cheaper / a better value / more bang per buck."

AMD dropped their prices because of the performance differential.

They both make great stuff these days and are pushing each other to higher standards. Buy what you want or need but lets not pretend alot of this isnt just blatant fanboyism.

Re:I choose AMD for the price...

[Bilestoad]

You may have the prices right, but you have no clue of what "equivalent" means in this case. An e4300 shits all over the Athlon 64 x2 3800.

Re:I choose AMD for the price...

[Blackhalo]

The interesting or sad thing, depending on one's perspective is that for companies who buy processors, and systems in quantities, is that they run their own benchmarks using their own applications before they buy in bulk. So no matter what the marketing spin the chip vendors run, they must perform before they can sell.

Re:I choose AMD for the price...

[Joe The Dragon]

you can also get better chipsets with amd then with intel at the same price.
Mid to high end AMD SLI Chip sets are the same price as low to mid end intle chipsets.

Re:I choose AMD for the price...

[evilviper]

AMD and Intel are CPU manufacturers, not sports teams. Buy the product that is the best performing at the lowest price.

CPUs are not homogeneous products. One is not just the same as the other.

AMD has better 64-bit support, lower power consumption, and far less likelihood of major microcode bugs.

Intel... happens to have slightly faster CPUs at the moment, and is hemorrhaging money to maintain prices competitive with AMD.

Re:I choose AMD for the price...

[lucas teh geek]

Intel... happens to have slightly faster CPUs at the moment, and is hemorrhaging money to maintain prices competitive with AMD.

and i should give a damn that a company I buy a product from is hemorrhaging money why exactly?

Re:I choose AMD for the price...

[steelsmack]

The e4300 seems closer performance wise to a 4600+ than a 3800+. The 4600+ is almost the same price as the e4300.

So while AMD has cheaper processors that doesn't mean they are a better bang for the buck. To be honest if you have no use for the performance then just buy a used system of some sort. So what Intel processor can you buy - with motherboard - that will be faster than that 3800+ at that price? You have to spend MORE. It was his fault for claiming "equivalent", but it's better at that price than anything Intel can offer at that price. If you want to spend more than that, Intel all the way, that price of lower, and unless you are either stupid, or an Intel fanboy, you should get an AMD product.

I am shocked!

[Anonymous Coward]

A company trying to make itself look as good as possible?! Next thing you'll know, Windows Vista isn't the most secure operating system ever made.

It's not the deceptive benchmarks that bother me

[realmolo]

What REALLY upsets me is the fact that the writers at ZDNet actually get *paid* to regurgitate data they likely found on some other website via Google.

What a great job.

Not the architecture..

[Junta]

In the broader sense of an 'architecture' in my mind, AMDs has a more advanced one than Intel (the integrated and hypertransport IO/multi-processor strategy).

Intel does, however, have a faster processor design than AMD's released product. If Barcelona levels the field in terms of instructions per clock, then the ball is back to Intel's court to at least meet AMDs memory/SMP/IO architecture or offset that deficiency with another leap in the processor technology. I hear Intel's roadmap eventually brings in a more AMD-like architecture in 2008, so that will probably be the time where the two vendors have the most easily comparable technology.

Re:Not the architecture..

[Iam9376]

In the broader sense of an 'architecture' in my mind, AMDs has a more advanced one than Intel (the integrated and hypertransport IO/multi-processor strategy).

Then it seems your mind needs an update. Intel's Core 2 architecture is significantly better than AMD's current or past (and seemingly future) architectures.

Putting all the fanboy drivel aside for a moment;
Intel's processors are faster without using more transistors, indications that the architecture is more optimized and makes better use of the available transistors.
Intel's processors scale vastly better than AMD's offerings both current and future.
Also consider, the die shrink to 65nm for AMD produced little to no benefits in speed and scalability (read: you couldn't over clock them very much)
Also, if anyone remembers, the Pentium M (which the Core 2 is based off) was benchmarked a few years ago against the AMD 64bit desktop processors and spanked them, no not in all cases or by any significant margin, but the fact a low power laptop processor (32bit) matched a 64bit mid-range/hi-end processor from AMD; that should indicate the advantages of the architecture.

Just because Intel does not currently have the memory controller on board, as well as the use of the older FSB does not make the Intel architecture any less advanced, the proof is in the puddin`, check any benchmark that puts current purchaseable processors and see how wins.

You're right about Intel, they will be releasing CSI (common system interconnect) for their processors in `08, if CSI does for Intel what even half of HT did for AMD, they may be in very serious trouble.

Sorry, but the rest of your post is moot.

Re:Not the architecture..

[TheRaven64]

Intel's Core 2 architecture is significantly better than AMD's current or past (and seemingly future) architectures In some ways, yes. The micro-op fusion stuff is incredibly shiny. They took some good branch prediction logic from NetBurst, and have a lot of neat tricks internally, particularly in the cache controller. On the edge of the CPU, AMD have the lead. They have a better interconnect (they are going to lose this lead soon, once Intel get CSI out of the door), and they have more intelligent memory controllers, which give them the edge in virtualisation and a few other things.

It's not entirely fair to say Intel is ahead of AMD architecturally. Both architectures have their strengths and weaknesses. At the moment, Intel are getting better overall performance (which means performance per Watt these days), but their architecture does have a few issues.

Re:Not the architecture..

[Ambassador Kosh]

I definitely don't agree that the intel systems scale vastly better. Most of the 4+ way benchmarks I have see with 8 or more cores go to amd pretty handily, The more memory the benchmarks need to use the worse off it gets for intel. So for desktops and very small servers where IO is not very important Intel is currently ahead in pure performance. If you need to setup an 8 core db server with 32GB of ram I would definitely go with opterons.

AMD is definitely not losing on the higher end server stuff, they are losing on the gaming desktops though since the Core 2 is a faster chip. For business work you pretty much never need something very fast. Probably the 3600+ is overkill for just about any business task and it currently as the best value of any chip I know of.

Re:Not the architecture..

[Iam9376]

I definitely don't agree that the intel systems scale vastly better.

Sorry I should have clarified, I meant scaled in the sense of clock frequency (overclocking).

Re:Not the architecture..

[Ambassador Kosh]

Ah for what I do I have never overclocked. I run all items at their spec settings and think of scaling as working with much larger datasets and things like that.

Re:Not the architecture..

[Anonymous Coward]

the proof is in the puddin`,

No, no no no no no no no no no no no no no no no stop doing this.

The proof of the pudding is in the eating.

OK? argh.

Now if I can only get people to stop saying 'at the end of the day'...

Re:Not the architecture..

[Vancorps]

You mean a processor that does more and offers rock solid performance versus and overclocked rig that needs rebooting every few hours?

Parent has an odd notion of what scaling is, although technically that would be considering scaling up when the general industry trend is scaling out. My database and web servers are Opteron based for good reason, exactly the ones you specified. Seems a shame AMD not being able to keep up on the lower end of things but they really aren't challenged by Intel in the mid-range server market. Would be nice to see some larger AMD setups, like 8-way 16 core, or 16-way 32 core. Architecturally speaking if that hardware were released it would only widen the lead at the top end and Intel really has some serious problems with scaling out.

Re:Not the architecture..

[mako1138]

Putting all the fanboy drivel aside for a moment;

Chill out, man. The Parent didn't have any fanboyism at all, but your post is full of it.

Intel's processors are faster without using more transistors, indications that the architecture is more optimized and makes better use of the available transistors.

The Core2 has almost 2x the transistor count of the Athlon X2, due to the huge cache. Or do you mean logic transistors? I can't find numbers for those. (Not that using more transistors would necessarily speed things up. In pipelining, perhaps, but we see how well that approach worked in the P4. I'd hesitate to pass judgment on an architecture, solely based on its transistor count and clock speed.)

Also consider, the die shrink to 65nm for AMD produced little to no benefits in speed and scalability (read: you couldn't over clock them very much)

It improved power consumption, though. And yields, I'd assume.

Also, if anyone remembers, the Pentium M (which the Core 2 is based off) was benchmarked a few years ago against the AMD 64bit desktop processors and spanked them, no not in all cases or by any significant margin, but the fact a low power laptop processor (32bit) matched a 64bit mid-range/hi-end processor from AMD; that should indicate the advantages of the architecture.

If we're talking about Yonah, it was slightly slower than the X2, clock for clock (which I'd hardly consider a "spanking"). True, Yonah had no on-die memory controller, and had worse cache latency compared to its predecessor, which makes for a pretty impressive showing.

As for the future, it looks like Barcelona will have killer FP, presuming it gets released soon.

Anyway, it's obvious that the Core2's features (4-issue, macro-ops fusion, 1-cycle SSE, etc) provide the current performance leadership on the desktop, but the FSB/memory interface is weak. Which is why Intel is going the same direction AMD has gone in those areas.

Re:Not the architecture..

[Bert64]

Intel processors certainly do not scale better than AMD...
AMD has a dedicated interconnect between processors, and a seperate memory bus for each processor, so adding an additional processor effectively doubles your memory bandwidth with a NUMA aware OS.
Intel on the other hand, effectively halves it's per processor memory bandwidth each time the processor count is doubled. The shared bus has to be shared for everything, inter-processor communication, memory access and IO device access. Also, their quad core chips are effectively just a pair of dual cores, so aside from needing less physical sockets they have no real advantage over AMD's dual cores... Infact, a pair of dual cores has a memory bandwidth advantage over a single intel quad core.

As for transistor count, do remember that current AMD processors do more than current Intel processors... They have the memory controller, hypertransport controller and some other logic on cpu whereas Intel requires this to be on the motherboard.

Die shrink - lets not forget the die shrink to 90nm for intel initially resulted in p4 chips which were even hotter and slower than their 130nm counterparts. It takes time to perfect a new production process.

As for Core2 being better than current AMD offerings, sure for small systems it does, but it still doesnt scale, so they have yet to trounce AMD across the board. Lets not forget that a couple of years ago, it was AMD who had a significantly better architecture and actually was trouncing Intel across the board.

If it weren't for AMD, Intel would still be pushing the P4 and the old "MHz is king" mantra. The chips would cost much more than they do now, we wouldn't have core2 at all.
Intel became complacent while they had no serious competition, so they got trounced by AMD. Then, with the P4 being such a joke AMD became complacent instead while Intel pulled their fingers out. We need AMD and Intel competing, otherwise progress and advancement will slow to a crawl.

Re:Not the architecture..

[petermgreen]

For business work you pretty much never need something very fast.
If all you use your PC for is paper pushing and as a glorified terminal then I agree.

Otoh if you actually use your PC to do actual computing work whether its compiling software, synthesizing vhdl/verilog for fpgas, converting media, running simulations etc then more speed can be very important to you.

Re:Not the architecture..

[imgod2u]

Well, to be fair, that's really saying that AMD's system architecture is better. I know it's a fine line but the interconnecting architecture (in the case of AMD, it's point-to-point crosslinks and dedicated pockets of memory) is what dominates performance in the "scaling" sense. As we've seen from SGI's Altix (using the Itaniums, which uses the same shared bus that the Core 2 uses), there is nothing stopping a system designer from making a NUMA design from dedicated busses. And performance is incredible.

The Irony Is Amazing

[Anonymous Coward]

So poor liddle Intel, the company that is the undisputed king of bogus benchmarking with their infamous SPEC compiler bullshit and a sickening array of other crap, is crying foul over someone else playing hardball?

Suck it up bitches. Don't dish it out if you can't take it right back Intel.

Re:The Irony Is Amazing

[Iam9376]

Are you too stupid to read past the headlines? Intel is no where mentioned accusing AMD of bogus benchmarks. ZDNet accused them, moron, even the fucking header says that.

Get a life fanboy, not every anti-AMD post is a result of Intel bitching.

(and yes, i know this is slashdot, but come on..)

Re:The Irony Is Amazing

[Anonymous Coward]

Yep, this had nothing to do with the Intel PR/Marketing types making calls/sending off emails...

Dipshit.

This is surprising?

[Etrias]

George Ou has long been an Intel/Windows whipping boy. He's not far off of writing the article that says that AMD has "Seal clubbing days" and internal seminars on "Making your grandmother cry".

Re:This is surprising?

[edremy]

I saw a lot of comments about that on the talkback section

Perhaps you'd like to actually address the complaint? Seemed pretty solid to me- Intel has used the best available, hard-to-cheat-on benchmark out there (SPEC) and gotten results. AMD is posting old results for Intel, results for AMD processors that don't exist yet and ignoring the best possible Intel products. Yes, it's advertising, but it's pretty crappy advertising, bordering on the deliberately deceptive. I'm a longtime fan of AMD- my home machines are AMD, I own stock in AMD, but crap like this makes me think about selling. If AMD is this desperate, they are in serious crapola

Re:This is surprising?

[njfuzzy]

A whipping boy is someone who gets an undue amount of negative attention. Are you trying to say that George Ou gets a lot of crap from Intel/Windows? I think you mean he is a schill, or maybe a hired gun.

Re:This is surprising?

[Anonymous Coward]

thank you for noting that.

i find it increasingly more difficult to communicate these days, despite having an iq of 157.

being surrounded by morons, who freely abuse the english language, one starts to doubt their own sanity...

Bad comparision but.

[ThisIsWhyImHot]

I agree that this is a poor comparision made to make AMD look a lot better then it is. But the fact is that these new cpu's coming out from AMD look like they're going to be the best on the market.

Re:Bad comparision but.

[geekoid]

Facts? what facts?

Re:Bad comparision but.

[99BottlesOfBeerInMyF]

But the fact is that these new cpu's coming out from AMD look like they're going to be the best on the market.

The best for what purpose? I evaluate chips based upon speed, price, power consumption, reliability, and "bonus features" with my needs in mind. Looking at the numbers, AMD is not winning the high end for speed, even with their next round of updates compared to Intel's current offerings. Power consumptions seems like a wash, or even a little bit in Intel's favor as well. I don't see any "bonus features" I care about in the revisions likely to ship in the next 6 months. So that pretty much leaves price, which of course will be constantly changing over time as chips are released.

Now I'm not counting AMD out for the next time I build a box, but unless that box is a really cheap server or low end pseudo-dumb terminal I don't see any reason to think AMD winning my evaluation is likely. They seem to have been well and truly outgunned and it looks like it will stay that way for a year or more based upon fab processes and roadmaps. I certainly hope they can turn it around and keep the competition fierce, but I just don't see where someone can objectively rate AMDs next round of processors as the "best."

Re:It's not the deceptive benchmarks that bother m

[eln]

Welcome to tech journalism. All you have to do is publish companies' press releases. For "in depth" articles, you visit their offices and have the PR guys talk to you all day. For product reviews, you repeat the companies' benchmarks and then turn on your demo unit to take some screenshots (if you can't find screenshots on the manufacturer's website, that is).

As someone who once worked for a company producing a product that had major hardware issues (as well as some fairly significant software bugs) yet still got rave reviews from tech sites, I can tell you this is pretty much how it works.

trust?

[SolusSD]

large publicly held corporations are usually only as honest as they have to be to hide their dishonesty

Yeah, But Ou Loudly Beat's Intel's Drum

[Anonymous Coward]

Being a subscriber to various TechRepublic newsletters I can say that ol' Georgie seems to have a very pro-Intel attitude, perhaps even to the point of "Fanboy".

That being said, yes, these are vendor benches, which we all know are a scam. At the same time, the anti-AMD guy shouldn't be blowing the whistle and crying 'foul'; it makes him look like a whiner.

What a load of lying malarkey

[(negative video)]

If AMD was comparing one architecture to another, they MUST do so based on identical core clock to memory clock ratios.

So what are the ratios in question, ZDNet? <pull string> "Math is hard."

Then the ZDNet jerkoff has the gall to complain that AMD didn't use the latest SPEC.org numbers. Well, duh. RUNNING benchmarks means just that: running them. You get the actual machines you want to compare, scrupulously make all the software as identical as possible, and let 'em rip. You DO NOT just grab random numbers generated by random software off a random website, no matter how impressive the numbers claim to be.

Re:What a load of lying malarkey

[Rakshasa Taisab]

By get[ting] the actual machines you want to compare, scrupulously make all the software as identical as possible, and let 'em rip, you mean building a standard performance testing tool-set and letting an impartial non-profit organization oversee the tests to ensure fairness?

That sounds like a really good idea... Why don't someone make one, and call it SPEO or something?

Re:What a load of lying malarkey

[edremy]

Then the ZDNet jerkoff has the gall to complain that AMD didn't use the latest SPEC.org numbers. Well, duh. RUNNING benchmarks means just that: running them. You get the actual machines you want to compare, scrupulously make all the software as identical as possible, and let 'em rip. You DO NOT just grab random numbers generated by random software off a random website, no matter how impressive the numbers claim to be.

You don't understand how SPEC works, do you? It's *not* a benchmark you just download and run and post the results everywhere. For a result to appear on SPEC's site, you have to buy the (expensive) suite, document everything about the machine, have SPEC vet the results and pay $500 a result to even appear in the tables. Core clock ratios and the rest matter not at all- all that matters is the final score on a set of tasks designed to mimic difficult real world problems.

In the world of scientific computing, SPEC numbers matter a lot when looking at machines, and vendors spend a lot of time trying to get to the top of the tables, but since the numbers are vetted by the organization, it's a *lot* harder to cheat. (Cheating tends to be more "Intel tuned their compiler to give good results on this part of the suite, but nobody else can use that optimization")

AMD is trying to pull a fast one here. As an AMD fan, I'm annoyed

Re:What a load of lying malarkey

[Anonymous Coward]

So it's like EEMBC, big deal. I spent a while at EEMBC meetings. I've ported and run all that stuff on on MIPS and ARM, multiple compilers, instuction set options, optimization settings, etc...

Expensive benchmarks for sure, strictly for the marketing guys.

Performance and I/O being adequate for the application, I have always given a much heavier weighting to tools, part cost and availability than I ever did to benchmarks.

I'm Totally Shocked!

[smackenzie]

So, what you are trying to tell me is that some company called AMD is posting benchmarks using processors that won't ship for a while (ahem, Sun / Sony), probably using carefully selected benchmarks (ahem, Apple / Motorola / IBM / Sony), and probably bragging about certain carefully selected synthetic results (ahem, Apple / Sony / IBM / Motorola) in carefully selected applications (ahem, ENTIRE FAB INDUSTRY).

I only left Intel out because I'm typing this on a Core 2 and I'm scared that if I point out the numerous times they have done something similar then my computer will crap out on me.

Now, having said this, can we all admit that AMD seems to have lost quite a bit of their edge recently?

Re:I'm Totally Shocked!

[Kohath]

AMD may have lost their edge a little. But they used to look good by comparison with Intel, and now they look bad by that same comparison. Improvements at Intel are the major factor.

Re:I'm Totally Shocked!

[Ash-Fox]

AMD may have lost their edge a little.

How did they lose just 'a little' edge?

But they used to look good by comparison with Intel, and now they look bad by that same comparison.

Why is that the case?

Improvements at Intel are the major factor.

What improvements are you talking about?

Deceptive Benchmarks?

[grasshoppa]

Can you say Oxymoron? I knew you could.

Re:Deceptive Benchmarks?

[gstoddart]

Can you say Oxymoron? I knew you could.

In this case, we want redundant (which I'm sure I'll get modded for this), not oxymoron.

An oxymoron would be something like "government intelligence" or "a little pregnant". Now, "Honest benchmark" when given by the vendor might be an oxymoron, "deceptive benchmark" is just a needless qualifier. :-P

Some links to brush up on here and here

Cheers

Re:Deceptive Benchmarks?

[Anonymous Coward]

Indeed. A benchmark only has value if all the details are known and that 'value' is questionable in the first place. Management graphs do nothing for me. The article is a total waste. Its not even clear if these tests were done with a real 64bit OS.

If you have to run Windows, Intel will probably do better. On the other hand, to do what a computer is made to do -- crunch numbers, its AMD any day and in my experience has been so for the last few years. It really does matter if it 'super computing' or 'just mousing around'.

Re:Deceptive Benchmarks?

[niceone]

No, "Deceptive Benchmarks" is the opposite of an oxymoron - an nonoxymoron or noxymoron or maybe even oxysmartdude.

Re:Deceptive Benchmarks?

[jd]

A single benchmark, no matter how good or bad, gives you no meaningful information. It is neither honest nor deceptive, although it can be correct or incorrect. To get total coverage of a system with N characteristics, you must make at least N(N+1)/2 measurements of the system, isolating every combination in turn. (Characteristics not being measured need to be held to a fixed, known value.) Because profiles are more useful than isolated points, you really need at least three times as many measurements - three points being the minimum needed to define a curve.

In practice, this is not a viable approach. There are far too many characteristics, which would lead to there being a gigantic number of combinations. What's more, different systems will have different numbers of characteristics, making it very difficult to do a direct comparison. You can't just ignore the differences, as they may reflect more efficient ways to do something, which will lead to an inaccurate comparison.

Benchmarking with any kind of meaningful accuracy requires considerable skill in determining an efficient number of subsets of characteristics and the number of samples per characteristic you'll need to get a good idea of the shape. It is not something you do in an afternoon, except in a trivial case or unless you're not intending to be honest or accurate.

Let's take a typical CPU. How many variables are of interest? Well, you've how fast the various I/O lines operate (instructions, data, any additional busses), caching speed for the different levels of cache, the speed of the different units as a whole (ALU, FPU, and so on), the speed of the visible registers, the speed of the internal registers, the speed of the pipeline, the performance of the cache dropping algorithm, cache strain when using multiple cores and/or when hyperthreading, performance degradation with temperature (error rates, clock errors, etc), heat generated by instructions, the latency when tranfering between units, etc.

300 benchmarks should be sufficient to describe the processor, but who is going to publish that many stats? Or read them? I imagine that a well-chosen 50 should be sufficient and still be publishable.

Re:Deceptive Benchmarks?

[Ogive17]

Not an oxymoron, it's just repetative.

Re:Deceptive Benchmarks?

[grasshoppa]

Indeed. I is sthmart, honest!

Some days it's best if I'm not given sharp objects, I swear.

Vista and its Blot makes Benchmarks Moot...

[Anonymous Coward]

Look for the average Joe/Jain running IE, Outlook, Word, and Excel, the hairs breath of performance difference is indistinguishable from Intel or AMD to that of the bloated Vista OS.

The Only Benchmark that Matters for Most folks

[N8F8]

Is the price/performance ratio. Intel is currently ahead in the CPU performance top-end but AMD is ahead in the price and GPU areas. Thinking of past history I would expect a "Coup de grâce" from AMD in the near future. They seem to be some clever folks over there at AMD.

Re:The Only Benchmark that Matters for Most folks

[geekoid]

Yep, nobody can copy intel in more clever way then the people at AMD.

"Coup de grâce" ... I think you mean "Core de grâce" ;)

Re:The Only Benchmark that Matters for Most folks

[Anonymous Coward]

It's awefully hard to beat a competitor on both performance per watt and performance per dollar for multiple years running by copying them. If anyone could be blamed for copying at this point in time, it would be Intel, who pulled a 180 and started copying from AMD after the severe error in judgment that is the P4 architecture.

Let's all scream FIRE!

[nick_davison]

From the comments on the original article:

The graphs are from a several months old marketing promo. Suddenly there's really no story.

Claim: AMD listed a product they don't intend to release.
Truth: AMD listed a product they intended to release at the time but subsequently withdrew.

Claim: AMD deliberately used out of date Intel scores.
Truth: AMD used the most current Intel scores available at the time. Improved scores came from an improved compiler - which may well change AMD's scores too. Either way, it wasn't available at the time of writing.

Claim: AMD ignored the most recent Intel processor releases.
Truth: Those Intel processors weren't released at the time of writing and no benchmarks existed.

Journalistically, this is about on a par with finding footage from the 50's saying we'd all be driving flying cars by the year 2000 and boldly asserting there's clearly a government conspiracy to hide the technology from the people to protect big oil.

Bold claims are one thing. Making them on the back of badly researching where the information came from is a great way to look like an idiot.

Re:Let's all scream FIRE!

[Trojan35]

And this is why Blogs are a bad source of good journalism. Luckily, most news sources have editors to weed submissions like these out.

Re:Let's all scream FIRE!

[neersign]

from the summary (i refuse to read ZDNet articles):

(2.6GHz Barcelona quad core) with older Intel Xeon quad cores rather than currently shipping ones which would beat the (hypothetical) score AMD claims for the future Barcelona. I guess while even the much slower 2.0GHz Barcelona is due soon AMD didn't think results from the 2.0 would look good enough - even against the slower Xeons they picked. Maybe the right comparison should be either best cpu against best cpu - or compare ones at the same price -- and only shipped products."

I don't understand how the Xeon 5355 is "older than currently shipping ones". I'm not a server processor guru, but when I go to Intel's site and click on server processors, I only see Quad Core 3200 series and Quad Core 5300 Series. Of the 5300 series, the 5355 is the top of the line processor as shown here. So, it looks to me like AMD picked the top of the line, currently shipping processor from Intel.

And while I'd rather see benchmarks on the 2.0ghz Barcelona since they are going to release it first, I do appreciate seeing numbers on a clock-to-clock basis. This shows me that per clock, AMD's Barcelona is doing more work, so even at 2.0ghz I can expect it to do more work than an Intel quad core at 2.0ghz. But I think it's clear that AMD chose to run their Barcelona at 2.6ghz to match the clock speed of the currently shipping, top of the line Intel processor.

While any benchmarks released directly by the manufacturer need to be taken with a grain of salt, I think any story written by some one who works for ZDNet needs even more scrutiny.

Re:Let's all scream FIRE!

[Paradise Pete]

Journalistically, this is about on a par with finding footage from the 50's saying we'd all be driving flying cars by the year 2000 and boldly asserting there's clearly a government conspiracy to hide the technology from the people to protect big oil.

Sure, just keep covering that up. Very clever of you to "casually" tie that in to this silly AMD article just to throw us off.

Re:Let's all scream FIRE!

[Wavicle]

Truth: AMD listed a product they intended to release at the time but subsequently withdrew.

This article references AMD's CURRENT MARKETING page on Barcelona performance.

Go to www.amd.com
->Processors
->Multi-core
->Products
->Barcelona
->Performance
(You may have to select language in there somewhere)

I don't see how calling AMD out on this is in any way inappropriate because they continue to use it.

Truth: AMD used the most current Intel scores available at the time. Improved scores came from an improved compiler - which may well change AMD's scores too. Either way, it wasn't available at the time of writing.

Are you trying to imply that AMD has had no reason to update their current performance page on Barcelona? I wonder why.

Truth: Those Intel processors weren't released at the time of writing and no benchmarks existed.

Has AMD fired its marketing department? I'm going to guess that the reason they have not updated their current literature is because the news isn't good.

Re:Let's all scream FIRE!

[GaryPatterson]

Look at the article's author. It's not so hard to believe George Ou would make such claims.

After the whole OS X 'hacked in sixty seconds' fiasco, it appears he's still a gutter journalist.

Still, to his credit he at least has a document to show everyone this time. That's a big step up for him.

Re:Let's all scream FIRE!

[Anonymous Coward]

No, AMD did not choose to run their processor at 2.6GHz - they chose to imagine running it at that speed. They wanted to ship at that speed but they have failed to execute. Although they may have it at 2.6GHz in a lab somewhere they have no hope of delivering at that speed this year. Intel OTOH has 3.0GHz Quad Core processors shipping, right now (and has had for several months) in Apple's Mac Pro machines. If you can buy it - it's shipping. If you can't - it's vapor - and that's a fair description of even the 2.0GHz Barcelona part right now.

You like clock-to-clock comparisons? OK, well you might like to imagine a P4 running at 10GHz. Or a PowerPC at 100GHz. Or a Z80 clocked even higher. It would be equally valid.

Re:Let's all scream FIRE!

[Bozdune]

Oh, yeah, I'm just SO impressed by "news sources" other than blogs. MOG. Forbes. Fox. As far as I can see, the only thing editors are doing, if editors even exist any more, is to fire anyone trying to do serious journalism.

MUST? ( was Re:What a load of lying malarkey)

[Anonymous Coward]

I can compare eels to escalators if I want to. AMD just got called on a less than forthcoming comparison. Big deal. Corporate behavior as usual.

The Real Issue IS..

[JohnnyOpcode]

Do you want Intel Core 2 Duo/Quad microcode bugs or do you want AMD Quad Barcelona (maybe with less or no bugs)! This chipwar is making me hold-off PC/Workstation purchases solely on this issue.

Re:The Real Issue IS..

[Glasswire]

...or do you want AMD Quad Barcelona (maybe with less or no bugs)
Well, statistically, all other things being equal (which I'm sure you would not agree they are) it's much more likely that a new processor like Barcelona, which is in it's early steppings will have more errata than a processor architecture that has been shipping for a year and has gone through several steppings. Be careful of the comparisons you ask for.

Comparison points

[Todd Knarr]

I'm interested in a side-by-side comparison at three points:

1. Best against best. How do the current top-of-the-line CPUs from each company compare.
2. Similar price points. If I'm willing to spend $X on a CPU, which company gives me the most performance for my money?
3. Similar clock speeds. This is more a techie thing, gives me an idea of which company's wringing the most from each clock cycle in their chips. With current tech it's not a really reliable guide to which CPU to buy, but it gives me an idea of how their tech stacks up.

Re:Comparison points

[geekoid]

Price points based on what? dollar to Hz? Dollar to power consumed? Straight Dollar to dollar?

"This is more a techie thing, gives me an idea of which company's wringing the most from each clock cycle in their chips. "

This means almost nothing with todays chips. A well designed 2.5Ghz would be a better chip than a poorly designed 3GHz. what about duel/quad/infintium core chips?

Nobody is going to release a chip that isn't pushing comparable cycles anymore.

Re:Comparison points

[wolfemi1]

I would like to add 4. Similar performance per watt consumed.

Re:Comparison points

[Todd Knarr]

Price points based on what? dollar to Hz? Dollar to power consumed? Straight Dollar to dollar?

Straight dollars. Essentially, if I spend $300, how much performance will that buy me from each company?

A well designed 2.5Ghz would be a better chip than a poorly designed 3GHz. what about duel/quad/infintium core chips?

Which is why I noted in that item that it's not a reliable guide for buying chips. There's a lot of things that affect performance anymore, FSB speed, FSB architecture, memory subsystem, cache, core interconnect, it gets slightly insane. But if you're a techie and like laying all that out on the table and analyzing it, it gets to be fairly interesting. I just wouldn't advise deciding on which chip to buy based on clock-speed-based comparisons (or any other single aspect, for that matter).

Re:Comparison points

[Bert64]

Mips per dollar
Mips per watt

Processor X uses 100 watts, produces 3000 MIPS and costs $500
Processor Y uses 50 watts, produces 2800 MIPS and costs $200

etc..
Or use mflops if that floats your boat.

Re:Comparison points

[Blackhalo]

Mmm, heat is a bitch, where is the power consumption in your metrics?

Re:Comparison points

[Chandon Seldon]

Mips and Mflops are not useful benchmarks at all on modern processors. That measurement would be completely unaffected by any of the design decisions that effect actual code - stuff like cache and IO bandwidth just wouldn't matter at all. To use a bad car analogy, that's like asking what the sum of the theoretical maximum energy output per cycle of the cylinders - sure, it's a performance related metric, but not a useful one.

Re:Comparison points

[Todd Knarr]

In my case ventilation and cooling. :)

Seriously, I mainly work with desktop tower systems. Power consumption's an issue mainly when it comes to the monthly power bill, and the delta from one computer burning an extra 150 watts is rarely a killer there. Cooling mainly comes into play when deciding what sort of CPU cooler to install and what fans I'm going to put where in the case. Or maybe whether I'm going to stick to air or go to a water cooling system. It'd be a different matter if I was designing high-density rack-mount systems or laptops, or embedded or appliance systems where there's little or no ventilation or fan noise is intolerable, but for an ordinary tower system the big question is just "Can I get a big enough power reduction to make a >5% difference in the power bill while still hitting my target performance?".

Re:Comparison points

[steelsmack]

1. Best against best. How do the current top-of-the-line CPUs from each company compare. Best against best? Intel is best, pretty easily. Unless you are looking at multiple socket (8+) situations, otherwise Intel's Core 2 Duo stuff is the best.

2. Similar price points. If I'm willing to spend $X on a CPU, which company gives me the most performance for my money? My feeling with this is pretty straight forward. If you want to spend less on a CPU than the cheapest Core 2 Duo ($120ish range), you are MUCH better off with an AMD, they have something cheaper and faster than anything Intel has once you get under the Core2 bottom end stuff. Anything in the higher $$ range, and you're best off going with Intel.

3. Similar clock speeds. This is more a techie thing, gives me an idea of which company's wringing the most from each clock cycle in their chips. With current tech it's not a really reliable guide to which CPU to buy, but it gives me an idea of how their tech stacks up. I believe AMD is down somewhere in the 10-20% range as far as performance per clock, so again Intel wins. I'm an AMD fanboy, my dad works there, but right now Intel rests at the top, hopefully Barcelona will change all of that, but I'm not holding my breath (and neither is he).

A Bit Tricky

[JamesRose]

When you are shipping a chip in two motnhs, its pretty hard to do a bench mark, comparing your chip to a chip from a competitor which they wont let you have FOR TWO MONTHS! I mean, they can't very well compare their chip to a chip that hasn't been released, so they did the next best thing and compared it to the chips that are around, anyone who actually looks at bench marks seriously will take note, look at what it is comparing to, and of course consider how much further intel are likely to advance in the next 2 months, take the information they give and work with it, they aren't just going to present it in an unbiased way because they are trying to turn a profit.

Comparison By Price

[xero314]

Maybe AMD used older Intel processors in the comparison because they wanted to compare processors of equal price. I don't know if that is the truth, but it certainly could be, which would then make the benchmarks realistic. Personally I don't care how fast the latest Intel Processor is unless it's performance per dollar is comparable to the latest AMD chips. This is part of why Apple moved away from IBM, which recently showed the fastest Micro Processors currently in existence, because even if IBM were to out perform Intel it would be at a much higher cost.

How about 4 sockets result?

[Anonymous Coward]

I can't find any Benchmark for 4 sockets machine here, so what's the point of this article?

I used to work at a very BIG IDC 2 years ago, the problems that made me sick at that time were noise and power consuming, especially power issue, you had to put less and less machines in a rack if you used a Intel CPU based ones. And this was the reason why I suggested my boss to switch from Intel to AMD and took 4 sockets machines instead of 2 sockets. Today AMD CPU based 4 sockets machines are still very popular at this IDC.

WSJ ads are from april

[nniillss]

According to this post, the comparisons, in particular the WSJ ads, are from april and at least the Intel numbers were correct at this time.

Re:WSJ ads are from april

[Glasswire]

And, of course the numbers for the AMD processor (which has not been released then or now) were just as unreal in April as they are today.

Re:WSJ ads are from april

[Glasswire]

Actually, did I understand you to say that AMD should only have compared processors which were available in April? I guess that means you can't compare processors which still aren't available in July either and there would be NO Barcelonas in the comparison at all.

Mission accomplished

[Anonymous Coward]

http://www.vnunet.com/vnunet/news/2193610/amd-prom ises-honest-barcelona

Thanks to the work of Ou, AMD will be promptly take these down.

Perhaps

[Colin Smith]

They are benchmarking against Xeons because they are going to price them at the older Xeon's level rather than the newer faster faster ones...

Yea, and....

[AetherBurner]

ZDNet introducing Pot to Kettle....Physician heal thyself. In reality, I prefer AMD over Intel any day. Everyone knows that benchmarks are rigged. Is this going to tilt my sale one way or another? Not at all. I have not bought a computer where I got to choose the make of processor in several years and I probably won't buy anything until my desktop draws 8's and Ace's. I have a feeling that websites that offer comparisons are biased as well. In God we trust, all others pay cash......

I think

[SlashDev]

companies SHOULD compare manufactured products compared to other manufactured products.

a video from AMD talking about comparing bench

[Anonymous Coward]

http://multicore.amd.com/us-en/benchmarks/
LMAO wow.

MOD PARENT DOWN

[Anonymous Coward]

He repeatedly states that Intel quad core uses two sockets, which is not the case at all.

Re:MOD PARENT DOWN

[Zantetsuken]

Quoth the all-knowing Wikipedia (third paragraph):

The Kentsfield's layout was strategically similar to that of Pentium D, as also the Kentsfield comprised two separate silicon pieces, called dies, (dual-core each) in one package constituting its quad-core[25] (i.e. Kentsfield's quad-core - or rather "double dual-core" - was a result of "gluing" two dual-cores together, so all Core 2's dual- and quad-core CPUs share the Core architecture otherwise impossible). Hence, the max. power consumption (TDP) of the Kensfield (QX6800 - 135 watts[26], QX6700 - 110 W[27], Q6600 - 105 W[28]) was approx. double of its similarly clocked Core 2 Duo counterpart. For example, the QX6700 consisted of two E6700 chips connected together by a 1066 MT/s FSB on one MCM, resulting in lower costs but less bandwidth to the northbridge. The Kentsfield was one socket processor sitting in a LGA775 socket, as well as Core 2 Duo (AMD Quad FX consisted of two dual-core processors in two separate sockets on one motherboard with a 2 x 125 W[29]= 250 W TDP).

Particularly note:

(i.e. Kentsfield's quad-core - or rather "double dual-core" - was a result of "gluing" two dual-cores together...

and

For example, the QX6700 consisted of two E6700 chips connected together by a 1066 MT/s FSB on one MCM, resulting in lower costs but less bandwidth to the northbridge. The Kentsfield was one socket processor sitting in a LGA775 socket, as well as Core 2 Duo

Now I suppose perhaps I read that wrong, but it sounds an awful lot like 2 sockets, or at the very least 2 processors as it says "glued together" instead of making a "true quad-core" - however wrong I was, doesn't that still say that Intel's quad-core solution here is inferior?

Re:MOD PARENT DOWN

[movdqa]

Intel's quad-core is one socket. AMD's FX quad-core is two-sockets (and boatloads of electricity). AMD has been touting native quad-core for quite some time but Intel's approach has big advantages in manufacturing the chip. If you run into defects on the wafer, you have to toss one big die instead of a die about half the size. Using pairs, you have the flexibility to pair core-pairs that use little power, or pair higher-power cores with lower-power cores. And you have the flexibility to respond easily to market demands for dual-core vs quad-core. So when you're talking about actually bringing something to market that sells, makes a profit and makes customers happy, Intel's approach is one that's actually been working. They claim to have sold 1 million quad-core chips so far. What's the score for AMD? Intel delivers Nehalem in the second-half of 2008. What would be incredibly funny is if they demo'd Nehalem this fall. Can you imagine the Architectural advantages of Core 2 Duo with the SSE4 and 45 nm enhancements of Penryn combined with cache coherency and an integrated memory controller? Intel has said that Nehalem is doing very well.

Re:MOD PARENT DOWN

[NeoThermic]

Everything you quoted points to the fact that Kentsfield is a single chip:

"For example, the QX6700 consisted of two E6700 chips connected together by a 1066 MT/s FSB on one MCM"

MCM is Multi-Chip Module, or basically many chips in a single unit.

"The Kentsfield was one socket processor sitting in a LGA775 socket"

"Kentsfield comprised two separate silicon pieces, called dies, (dual-core each) in one package constituting its quad-core"

Two cores, one package, one LGA775 socket. I make that Pimms o'Clock.

NeoThermic

You think AMD cares about Slashdot?

[Joce640k]

You think AMD cares about Slashdot's opinion? They're aiming at the PHBs.

If only...

[brian0918]

If only everything posted on the internet could be continually updated to reflect modern expectations.

It all makes sense now.

[Anonymous Coward]

Your ideas intrigue me and I wish to subscribe to your newsletter.

Xeon X5365 - not yet available ?

[this great guy]

The ZDNet blogger claims that Intel's XEON X5365 3.0 GHz quad-core CPU shipped back in April. However I can't find it in the current Intel pricelist, neither at any of the popular online retailers. Only a bunch of hardware websites have been able to review this processor. It looks like its not yet available to the general public. Am I missing something ?

Re:Xeon X5365 - not yet available ?

[Anonymous Coward]

I'm not 100% sure, but what about this

http://www.apple.com/macpro/specs.html

Re:Xeon X5365 - not yet available ?

[Smorkin' Labbit]

Apple uses those in the Mac Pro models, going up to 3.0 GHz, and they've been selling them for months. So possibly not available for everyone to buy, but they do exist and Intel has been delivering them for some time.

Deceptive Article

[Anonymous Coward]

As can bee seen from the press release, the benchmark was released in April.
http://www.amd.com/us-en/Corporate/VirtualPressRoo m/0,,51_104_543~117115,00.html

That said, it is simple wrong to accuse AMD of comparing their parts "with older Intel Xeon quad cores rather than currently shipping ones."

Wherefore art thou FactCheck?

[macraig]

Unfortunately for us, FactCheck limits its interest to politics, although... this sorta posturing might even qualify as politics. Maybe they'll register the domain FactCheck.BIZ as well and start a new branch?

Time to Panic???

[Big Smirk]

I understand the author is outraged but over what?

AMD posted (or is still posting I couldn't find a date) based on old data? Or was the data accurate at the time they created the document.

More outrageous would be the fact that somehow Intel got nearly a 50% boost in speed since April. Unclear from either the article or AMD's website on how this minor miracle was done.

I mean, suppose someone was selling a sports car and publish in April 2007 that the car tops out at 78mph. Then a reviewer bitches that competitors are using that number and not the 'latest' numbers which show 102 mph.

1) "It's fictitious since AMD doesn't have a 2.6 GHz Barcelona quad-core CPU"
Yeah it says "Estimated" in big bold letters.
2) Using numbers from April 2007.... shocker. Use the new improved test (that of course was tweaked on Intel) what is the new numbers for AMD chips? Does this change AMD's "Estimation" spec.org seems to be slashdotted right now.
3) So AMD compared its 2.6Ghz to Intel's 2.6 Ghz and the author is complaining they didn't include the 3.0Ghz???? What? AMD is being 'selective'? Yes dammit! They are claiming that clock for clock they are 'Estimating' they will be faster.

BFD! Get a life Intel fanboy.

To take the author's quote:
"this is just outrageous"
Yes, (64.9-60.4)/60.4 = .074 That's 7.4% advantage not 14.7% as quoted in the article. Outrageous using old numbers!!!

Re:Time to Panic???

[Anonymous Coward]

AMD will be shipping the 2ghz chip this September and people will overclock them to 2.6, so I don't find this article excessively misleading.

Benchmarks haven't compared clock for clock since Pentium III and the first Athlons. I recently purchased a pair of Xeon CPUs and I don't recall any 53xx series Xeons that were 3.0ghz so it does look like a fairer comparison to use the 5355. It doesn't make sense to compare them by clock due to the differences in architecture. Price point and performance are the only valid comparisons these days.

correction

[petermgreen]

replace the first intel in that comment with amd. I know intel have a quad core xeon.

The best way to settle this benchmark thing

[Anonymous Coward]

The best way to settle this whole benchmark thing about which processor is faster is simple: Each company should give me a really loaded system that incorporates their latest and greatest processor. I'll use them for a couple of years, and then report my unbiased impressions right here on Slashdot. What could be fairer?

SPEC isnt gospel

[asm2750]

I've used SPEC before, honestly, its nothing great. You would get more concise results if you just take two physical machines and ran them against each other, with their compilers optimized. Also this is a first actual quad core processor from AMD, you cant expect it to be absolutely perfect, look at the crap Intel made when EM64T first came out, they corrected it when the next revision was made. As for AMD and their fusion line, its an inevitable architecture, maybe not for high end desktop systems but, definitely for laptop and embedded applications. One gripe I have about processors today is that we are still talking about clock speed for performance. That shouldn't matter anymore honestly, these days it should be all about instruction parallelism, and architecture optimization.

Opteron is fine, just overpriced

[billcopc]

I'd like to remind everyone that AMD started out selling cheap slow CPUs in the upgrade market. Remember that K6 400mhz your uncle used to have ? Well that K6 cost a whole lot less than any Intel processor at the time, and it breathed new life into old boards one or two generations behind. Then one day AMD released the Athlon, took the performance crown and didn't really know how to play their role. Their marketing was shit, and their pricing wasn't so good anymore. They had tons of experience being the underdog, but zero skill as a leader.

Now Intel has come back on top, but AMD doesn't want to go back to being #2. Instead of putting their efforts toward a new, faster architecture like Intel did, AMD is resting on their laurels, releasing outdated underwhelming puke way too late in the game. I'm sitting here with an AMD x2, I've had it for about two years and I've been running it slightly overclocked since I got it. Well my 2 year old chip is still faster than their fastest CPU today. I would love to buy a new CPU that's 30-40% faster, but they don't make one. Even my buddy's brand new Intel E6600 is faster than what I have, and he didn't pay all that much for it. That's why I'm getting an Intel Q6600 in a few weeks, when the prices drop again. AMD still won't be anywhere near releasing their first quad core processor.

AMD needs to shut up and take their place. They're really good at selling slower, inexpensive processors for the everyman. They need to stop lying to themselves and accept the fact that they just can't cut the mustard when it comes to high-end, which is fine because the big money is in the OEM market, where every dollar counts. If AMD can produce a decently fast and affordable chip, and hire a goddamned business director to get some partnerships going, they could make a ton of money. Just don't pretend the Athlon is a performance king, because we all know it's a lie and the only fool is AMD.

ZDNet? Why even bother with the story?

[mp3phish]

You are reading this from ZDNet?

As if ZDNet is any more trustworthy than the vendor's own benchmarks in the first place?

Get back to me when you start hearing this from scientists criticizing AMD who are actually doing their own real world benchmarks. ZDNet is to computers as Faux News is to politics.

Intel + AMD = mediocre

[manowar821]

I'm pretty pissed off about this whole CPU company deal lately. I don't want to give my money to either company.

I wish AMD would surprise us all and release a 3.5ghz quad core Phenom this week, or something. That would deserve our attention. Neither company has "WOWed" me since AMD released their X2 line. It's just disappointing, is all.. They're supposed to be competitive, but I don't see any busting of their asses to get sales. Intel hasn't impressed me since the p4EE. Even that was only okay...

Now there is all this talk about AMD not releasing anything worthwhile until Q1 2008. That is disturbing. D: