Intel Loses Market Share to AMD

diverge_s wrote to mention an article examining Intel's market share loss to AMD in the fourth quarter of 2005. From the article: "Sales of Intel-based desktop PCs fell 22.3 percent during the fourth quarter, according to Current Analysis. As a result, sales of AMD-based desktops took the lead during the pivotal fourth-quarter holiday shopping season. AMD chips were found in 52.5 percent of desktop PCs sold in U.S. retail stores during that period."

Comments

VIIV

[F_Scentura]

Their new push for quality engineering over marketing fluff will surely give them the lead again!

Re:VIIV

[Speare]

How about that move to change the motto from "Intel Inside!" to "Intel Aside!"

Re:VIIV

[canuck57]

Their new push for quality engineering over marketing fluff will surely give them the lead again!

I am still ticked at my PERL mobo w. P4 HT 2.4GHz that died just out of warrenty.

If Intel want's back, cheaper, faster, cooler and more reliable come to mind. AMD has this over Intel at the moment and I have a 1.2GHz AMD that keeps on ticking.... so naturally one of those dual core AMD 64 X2 systems is on my list.

Re:VIIV

[GmAz]

I personally still have an AMD 1Ghz Thunderbird still ticking. I don't use it much because I have my AMD64 system, but it runs like the day I got it.

Re:VIIV

[JackAtCepstral]

I have a 1.2GHz TBird, as well. I too, have an AMD64 system that's the workhorse now, but that TBird never gave me a single problem.

Re:VIIV

[blair1q]

What's with the reliance on anecdotal evidence?

Re:VIIV

[deesine]

It's personal anecdotal evidence, that's the difference.

Re:VIIV

[ncc74656]

I personally still have an AMD 1Ghz Thunderbird still ticking. I don't use it much because I have my AMD64 system, but it runs like the day I got it.

A machine like that makes a great home web/mail server. Mine got pressed into that type of service when the P!!! dually I had been using went tango-uniform. (The dually was a pair of 500-MHz Katmais on an N440BX. Some power-regulation components on the motherboard literally burned up.)

The Athlon's on a Biostar M7MIA, which is one of the few Socket A motherboards that still had an ISA slot. I bought it about five years ago. It's been running alfter.us for about two years now. It still runs fairly well, but now that there's a way to run Gentoo on the Linksys NSLU2, I might switch to that at some point.

Re:VIIV

[Anonymous Coward]

My 1 Ghz AMD just died, and was working great up until recently. A few years ago I had a Pentium I die on me, and my 3+ year old P4 still works great. I also have old P I's and early AMD's still running. What's your point?

Re:VIIV

[Narishma]

Yeah and I still have my Intel 80286 running like the day I got it.

Re:VIIV

[Anonymous Coward]

My AMD Turion processor died within 1 week of the first time I booted the machine.

Re:VIIV

[Guppy06]

AMD has shown you can have both! My Athlon 64 3400+ is running at 2400 MHz!

Re:VIIV

[F_Scentura]

Heh, but that's primarily matching Intel's marketing pap, not exceeding :)

Re:VIIV

[snuf23]

I personally have a Pentium 233MHz, a Pentium 3 850MHz, a Celeron 2GHz, a Pentium 4 2.4GHz and a Pentium 4 3.2GHz which are all still ticking.
So does that mean I should get an Intel next time?
I think I'll take a look at the options when I need to buy and then decide which manufacturer to go with.

Re:VIIV

[Fred_A]

Well, I used to have a 486-50 (AMD) /I think/, and an Athlon 800 which both died. Both times the CPU apparently burned out.

Ok, so they ran 7/24 for their 7 or 8 year lifespan, but you can't call that quality !

I'm currently running an amd64, but if it dies before 2015, I swear I'll never buy another of those crappy AMD CPUs.

Re:VIIV

[Anonymous Coward]

My AMD Turion processor died within 1 week of the first time I booted the machine.

Um, did you, like, use the thermal paste?

Re:VIIV

[Anonymous Coward]

Were you overclocking them?

Re:VIIV

[carl0ski]

I have a duron 1ghz
near 5 years old morgan core


I did have an Athlon XP 1700+ running @1.66mhz

it ran extremely cool compared to comparable Pentium 4's ive seen
but the XP died at a 600mhz over clock
(2.1ghz) it ran for hours but once i shut it down in never booted up again.

i now have an Athlon 64 X2
setting it up, the fan is just so quiet i couldnt telling if it was even running or not


note it chews through folding project 5x as fast as my old XP

Re:VIIV

[Jeff DeMaagd]

I sold my dad a 2x 500MHz Xeon for a desktop a couple years ago. It still runs non-stop and is pretty darn quiet too, the system only has two fans, both are 12cm, so there's no stupid high pitch.

From the tags, it appears it was assembled in 1997. I think it's too bad that the chip makers moved away from the processor cards, though I understand that it is a cost cutting move.

Re:VIIV

[Bing Tsher E]

You probably didn't overclock any of them. Nor did you select the fans and case based on the latest blinking LED technology.

Re:VIIV

[Anonymous Coward]

I sold my dad a 2x 500MHz Xeon for a desktop a couple years ago. It still runs non-stop and is pretty darn quiet too, the system only has two fans, both are 12cm, so there's no stupid high pitch.

That was the old days, I even remember when a Z80 didn't even have a heat sink let alone a fan.

Today, one of those Intel processors has more heat coming out of than an easy bake oven so the fans have to run higher.

Re:VIIV

[snuf23]

The one I overclocked was my Celeron 300a. Which as far I as know still works. It hasn't actually been in a motherboard since I upgraded to the P3 850Mhz.
I find actual CPU failure to rarely be the cause of system failure. At work I don't think I've seen a CPU failure in 5 years. Usually the motherboard blows up. But yeah, we don't overclock at work either.

Re:VIIV

[blair1q]

Makes it subjective anecdotal evidence, all the more reason it's less credible.

Re:VIIV

[Fred_A]

Were you overclocking them?

No, I never overclocked any of my CPUs, they just died. Maybe they ate fast food when I wasn't looking, I don't know.

Godd quality and low prices work :)

[gasmonso]

AMD just proves that regardless of your advertising budget, it all comes down to good performance and good price. I don't think I have ever seen an AMD commercial, whereas Intel was all over the TV. Dell has finally taken notice and will start widespread use of AMD chips soon. Thanks for the giving Intel some competition AMD!

http://religiousfreaks.com/

Re:Godd quality and low prices work :)

[Firehed]

That's not true in most situations. Almost all products come down primarily to effective marketing. In fact I'm honestly surprised it's true here, but it's also true that if there's one thing people are sick of, it's a slow computer.

Re:Godd quality and low prices work :)

[Anonymous Coward]

And what does this have to do religiousfreaks.com? If you want to use a sig, put in the proper field on your preferences page.

Re:Godd quality and low prices work :)

[PlanetX 00]

I agree that AMD ships a great product, but let us not forget the following:
1. Intel still has a commanding lead on overall desktop processor sales
2. Intel is a silicon fabrication machine (great yields, great process, large volume). What they lack in cutting edge CPU features they make up in fabrication
3. Intel = server CPU. This is a very high margin area that AMD continues to fail to penetrate
4. Intel has more money than the know what to do with. This allows them to keep on redesigning their chips to keep up with AMD (see Pentium M)

This cat and mouse game will continue as follows:
AMD innovates, Intel falls behind spends a boatload of money to catch up. Intel's great fabrication team steps up to the plate shrinks the process size, increases yield, increases margin, and they make another boat load of money.

Until AMD can match Intel on fabrication and chip set development they will remain the little fish in the pond :-( As for me I'll still try to cut through the marketing fluff and buy my CPUs based on the best price/performance.

Re:Godd quality and low prices work :)

[paitre]

3. Intel = server CPU. This is a very high margin area that AMD continues to fail to penetrate

If you are referring strictly to the high-volume, sub-$25k/machine market, you're only kinda (barely) right.
If you are referring to any other segment of the server market that Intel and AMD both play in (ie. 4-way, and >$25k), you're wrong. Dell sold a whopping -4- machines that cost greater than $25k last quarter. -4-. Clusters are accounted for as lots of little machines, and while Intel has greater share there due to volume, AMD's presence in the cluster market is anything but insignificant. Saying that AMD is failling to penetrate the server market would have been true two years ago. It's not been true for a while, now (cf. Q32004 AMD's share in servers was 8% - for a company that effectively had -0- prior to Opteron, that's significant).

AMD's current share of the overall x86 server market is some 16% now. Calling 16% insignificant is a stretch, at best. This is particularly true in light of the near -40%- share that AMD has in the 4-way market. Of course, that's Q3's numbers. Not Q4's (which were announced yesterday - .40/share before the one-time Spansion spin-off charge).
Judging by Intel's -miss- of their market estimates, and AMD's blowing theres away, I'd say that their server numbers are up, yet again.

Re:Godd quality and low prices work :)

[C0vardeAn0nim0]

"3. Intel = server CPU. This is a very high margin area that AMD continues to fail to penetrate"

you probably stoped following the server market, then.

since the release of their first 64 bit chips a while ago, AMD jumped from a virtual 0% penetration on the server market to more than 11%. except for Intel's best buddy *COUGH*michaeldell*COUGH* the other big player all have Opteron oferings. not to mention the small mom-and-pop shops that now have the oportunity to compete in the 16-way server market with boards like Thunder K8QW (S4881) which supports 8 (with aditional daughter card) dual core opterons.

AMD is inching it's way against intel in server, desktop and mobile with their great offerings. since i moved from 8-bit machines to PCs i've alway been an AMD user, and i'll continue to be for the foreseable future.

Re:Godd quality and low prices work :)

[dysfunct]

I especially like AMD's strategy.

They started off being cheaper than Intel at comparable quality and pricing to achieve decent brand recognition and market penetration. Then they ignored the clock speed war for some time while Intel was still only focused on clock speed and couldn't change their production easily. Using this advantage AMD beat Intel to 64bit processors, quality and pricing AND continued the clock speed war naming their processors so that they appear to be clock rates ( AMD 64 3200+ etc. ).

I actually don't understand how until now Intel could sustain all of AMD's pressure except for marketing and lobbying, since for some time now AMD processors are faster, cheaper, support 32 and 64 bit natively ( and gain performance by their 64bit design even in native 32bit mode ) AND are way cooler, more silent and energy efficient than most - not all I must add - Intel processors.

No, I'm not a drooling fanboy. Until about two years ago I was a pentium customer. I'd switch any time if Intel managed to produce a processor that's cheap, performant and does not require ex[t|p]ensive cooling to achieve said performance under stress load because it's already close to melting ( 70C, 158F ) when idle.

Re:Godd quality and low prices work :)

[lcsjk]

"Intel is a silicon fabrication machine (great yields, great process, large volume)."

I have always thought that Intel has a high yield, quality process. However, their prices have always been higher than AMD, Cyrix, and any of the other competitors over the years. If their yield was good, then there was no reason for them to be higher priced. I always felt that they were gouging the customers, so I quit designing them into my systems. Since they are still higher priced than AMD for competing parts, I wonder if their yield has always been not so good.

Re:Godd quality and low prices work :)

[toadlife]

"I have always thought that Intel has a high yield, quality process. However, their prices have always been higher than AMD, Cyrix, and any of the other competitors over the years."

That's because with Intel you're paying for the chip, plus a portion of the Blue dudes' salaries.

Beige boxes?

[Anonymous Coward]

I wonder whether AMD's success is an indication that PC's are well into their commodities phase and so el-cheapo models at Best Buy are (more than) sufficient for people's use? Intel's in the pricier boxes, so they stand/fall with those vendors.

Re:Beige boxes?

[HermanAB]

Well, especially, when the pricier models are worse than the cheaper ones... "Style, is the ability to distinguish quality, without looking at the price tag."

Re:Beige boxes?

[Lonewolf666]

Not all people are gamers who really need the performance.
For other people, el-cheapo models at Best Buy are (more than) sufficient. Except maybe on longevity, where things like lousy fans and faulty capacitors still are a problem.

Re:Beige boxes?

[GmAz]

Not really. Bestbuy, CompUSA, Circuit City all have salemen working the floor. I was once a salesman for CompUSA and did rather good at it. 90% of people buy what the salesman tell them to buy. You get people uneducated in the field of computers and look to the salesmen as the "pros". I sold quite a few $4000 systems because I could read the customer. If that person had money, I could sell him on the high end machine no problem. The line "In computers, the old saying 'you get what you pay for' auctually applies". Talk about Hook Line and Sinker. But, you also made money there based on the service plans you sold and was based primarily on a percent of gross sold vs. gross service plan. People wanted to see big numbers on the front of the computer and a low price tag. AMD machines reigned. Espically since I set them up with the AMD machine right next to its Intel counterpart. Lower priced machines made me more money as it does every salesman. I think salesmen have figured this out and that is why AMD took that large market share. And guess what, when more AMD machines were pushed out, people started realizing that they were as good as the covited Intel machines. Word spread and customers started looking into AMD and not just saying that they only thing they wanted was Intel because its the only thing they know about computers. Gotta hand it to Intel for the A+ marketing strategy.

Re:Beige boxes?

[Anonymous Coward]

I don't think so. The cheap machines sodl in supermarkets and department stores are mostly Intel-based, Celeron if the buyer is a bit unlucky.

Re:Beige boxes?

[Philosinfinity]

Really? Are you claiming that a Celly system is less than the low market AMD system? I think that the price structure between comparable speed systems are comparable across platforms. The insinuation that AMD units are the cheapo ones may have been true throughout the K6-2 and original Athlon models, however I really think that Intel and AMD have processors in systems that range all across the price spectrum, thus competing with each other. However, an argument could be put in place that since Dell is the largest seller of Desktop PCs, and Dell is Intel Only... these numbers could easily be put into some question.

Re:Beige boxes?

[Anonymous Coward]

"I was once a salesman for CompUSA... I sold quite a few $4000 systems because I could read the customer."

... and that's why you should (if you insist on buying from a retail store) go to BestBuy or some other store where sales people don't get commissions.

Consumer vs Corporate?

[Anonymous Coward]

Of course, it's always been my understanding that Intel is dominant in corporate computing, where no small number of third party corporate applications are only "certified" to work on Intel processors and the use of AMD processors endangers your ability to take advantage of your pricy support contract.

Re:Consumer vs Corporate?

[HangingChad]

Of course, it's always been my understanding that Intel is dominant in corporate computing...

If Intel is holding on to dominance in any market segment it's more likely to be the result of their business relationship with a company like Dell, which has been propping Intel up for the last two years while AMD ate away the rest of their market.

AMD makes a great product at a competitive price. What happened to Intel will happen to every other company that starts thinking they have a right to exist. Intel sometimes acts like they're a government agency.

Re:Consumer vs Corporate?

[diegocgteleline.es]

It's not just Intel. Until recently, +95% of the x86 server market did belong to Intel - that means that mobo/chipset/other products makers have their product offerings oriented to Intel. AMD made a great chip, but moving the whole x86 server market from one CPU vendor to another is not something that will happen from one day to another - Intel can't influde that, lots of companies are just not switching to another CPU vendor just because they've had good sales in 1 year. There're no good server-oriented chipsets for AMD chips except the ones from AMD for example (please don't bother mentioning that company that makes graphics cards and started making chipsets yesterday); however you can find chipsets from serverworks or IBM for Intel CPUs. I'm not suprised Intel keeps selling more than AMD, AMD will have to _keep_ the lead a few years if they want to get support from the industry.

Re:Consumer vs Corporate?

[corngrower]

Intel gave as one reason of lower than expected revenues was their inability to supply enough chipsets for their processors. They also were predicting slower growth the next quarter or two. AMD's revenue was also limited by supply problems, but in their case, the demand drove up the price of their processors. They also expect good growth in processor sales next quarter.

It will be interesting to see how things unroll the next half year, with AMD's new Dresden fab coming online, and new chips being offered. I'll be interested to see how much more inroads AMD will be making into the server, desktop, and laptop markets. How will Intel's Xeon hold up against the Opteron? Will AMD get more competitive in the market for laptop processors?

Right now I see Intel having the better fabrication technology, and the ability to produce more chips than AMD. Will Intel catch up and surpass AMD on the chip design front? Maybe the question should be 'when' and not 'if'.

And finally, can we expect to see a 5 GHz processor this year?

Re:Consumer vs Corporate?

[winwar]

"If Intel is holding on to dominance in any market segment it's more likely to be the result of their business relationship with a company like Dell, which has been propping Intel up for the last two years while AMD ate away the rest of their market."

I suggest you find a window and look outside. What you see is reality. And AMD hasn't remotely impacted Intel's market. Hell, an awesome market share for AMD is 20%. Which they have trouble keeping (they MIGHT be near or slightly above that mark now).

Intel keeps its market share for multiple reasons. The most important is production capability. The rest include advertising, competive pricing and products, and the willingness to share roadmaps to partners (and generally follow them). I'm sure there are others.

Re:Consumer vs Corporate?

[corngrower]

Hell, an awesome market share for AMD is 20%.

I believe I read that AMD's market share of x86 processors is 21% as of late. It will probably increase to 22 or 23% this year.

Intel keeps its market share for multiple reasons. The most important is production capability. The rest include advertising, competive pricing and products, and the willingness to share roadmaps to partners (and generally follow them).

Can't disagree with that. AMD is not going to increase market share without increasing their production capacity. Plus they're going to have to do a better job with their laptop processors.

This news is old

[Anonymous Coward]

As soon as they started selling their 64bit stuff I'd get rid of my noisy power and nerves sucking Intel heater and my geek life was definately worth living again.

There are many Intel-fanboys left because of power and buzzwords. But even the dude of the ASUS hotline told me to go for AMD since Intel CPU's crash because else they would burn my house down.

There. Job done. Nothing for you to see.

Marketing misstep?

[PornMaster]

Anyone looked into the possible marketing misstep by Intel stopping marketing their processors by clock speed?

Re:Marketing misstep?

[Sique]

It was a 'misstep' they had to take with going away from the Netburst architecture anyway. The Pentium M and successors all have much lower clock rates with still retaining comparable performance. For low power devices the high clock rates were hell.

Re:Marketing misstep?

[Anonymous Coward]

I think it is important to point out Intel's move away from processors as the main source of their business. A lot of resources have been moved to processors and chipsets for non-PC computers. One might expect some loss in market share while trying to refocus. It doesn't mean Intel doesn't care about PC processors, obviously it is a lot of revenue, but the market is not as profitable as before.

Re:Marketing misstep?

[ls -la]

the market is not as profitable as before

Perhaps because they have competition now?

Re:Marketing misstep?

[Guysmiley777]

The "megahertz race" was a monster of their own creation. The Pentium 4 was a misstep, changing the design to allow higher clock rates with less processing per cycle.

Would you rather have an engine that puts out redlines at 6,000 or 12,000 RPM? I forgot to mention, the 6,000 RPM motor is a 5 liter V8, the 12,000 RPM motor is a 60 CC weedwacker motor.

Re:Marketing misstep?

[HarvardAce]

Would you rather have an engine that puts out redlines at 6,000 or 12,000 RPM? I forgot to mention, the 6,000 RPM motor is a 5 liter V8, the 12,000 RPM motor is a 60 CC weedwacker motor.

I'd rather have the 12,000RPM motor in my weedwacker. I just imagine the 5 liter V8 would be a little cumbersome trying to get the areas around my fence.

Re:Marketing misstep?

[Guysmiley777]

LOL, another analogy down in flames. What if they both weigh the same and the 12,000 RPM moter runs hotter and burns more gas? :)

Re:Marketing misstep?

[greed]

Maybe these guys (caution, may contain Flash) can turn their attention to a suitable V8-powered weedwacker; if you dare them to do it now, they might have one ready for spring.

Re:Marketing misstep?

[toofast]

the 12,000 RPM motor is a 60 CC weedwacker motor.

You be wackin' some big weeds with that 60cc weedwacker motor.

Re:Marketing misstep?

[servognome]

The "megahertz race" was a monster of their own creation

Both Intel and AMD created the MHz race (remember 1Ghz being hit by AMD first). Ultimately Intel won the race with Northwood, as AMD's marketshare fell from 20% to 10%. The problem for Intel, they kept in the race too long with Prescott, when the laws of physics finally caught up with their marketing will.

Re:Marketing misstep?

[kermitthefrog917]

Clock speed improvements are at a standstill. People aren't going to understand what a new processor is capable of because they will see it has the same clockspeed as an older POS. By not marketing the clockspeed, they bypass the embarrassment of the overall CPU industry in failing to speed these processors up.

meh...

[DigDuality]

I'm not really a fanboy on either side of this Chevy/Ford arguement. They both support Trusted Computing which makes me wish there was another option out there.

Re:meh...

[Urusai]

Don't forget http://www.via.com.tw/en/products/processors/c7/ and http://www.transmeta.com/efficeon/index.html. Or maybe you would be better off forgetting them, I don't know.

Re:meh...

[syphax]

They both support Trusted Computing which makes me wish there was another option out there.

I don't see MOS Technology on the list...

Here's the CPU for you!

Re:meh...

[Kjella]

I'm not really a fanboy on either side of this Chevy/Ford arguement. They both support Trusted Computing which makes me wish there was another option out there.

Microsoft basicly went out and said you'd need TCPA to run Vista. Given the OS market, The only one who could have refused that without being cast into obscurity would be Intel. And Intel/AMD both want the "Media Center" concept which sells their CPUs, I don't blame them. Your third-party candidate would have about as much power as in US elections. If we assumed that people actually cared (they don't), then he'd probably be outcompeted by the last generation of DRM-free machines sold on eBay, and file for chapter 11 soon after.

Re:meh...

[Nutria]

MOS Technology

Gag, spit.

The Z80 is obviously superior to a toy like the 6502.

Re:meh...

[DigDuality]

thanks, i'll give a deep look into each.

Re:meh...

[Anonymous Coward]

They both support Trusted Computing which makes me wish there was another option out there.

Trusted computing isn't necessarily a BAD thing. However, I havent found a vendor that I would trust - yet.

Re:meh...

[Nahooda]

AFAIK, Transmeta was the first x86 processor maker to support Trusted Computing, so it's a bad choice:

http://en.wikipedia.org/wiki/Trusted_Platform_Modu le

According to an unofficial member list, VIA is a member of the Trusted Computing Group.

http://www.againsttcpa.com/tcpa-members.html

Acoording to the offical list, it's not, so I'm a bit unsure about VIA.

https://www.trustedcomputinggroup.org/about/member s/

Regards,

Dennis B. Schramm

Re:meh...

[SilverspurG]

Due to name recognition of transmeta I checked out the Efficeon. What's wrong with it? It looks like it has potential.

Re:meh...

[DigDuality]

It is a bad thing when they don't allow you to have the choice on what to trust and what not to, and make that decision before hand.

Not to mention this, and mark my words, will be used as a battle against open source software. It's the perfect way in which a user is locked into a software set that works on a particular hardware set. That's no only a danger to open source, but a danger to a free market that thrives on competition. This is big business dictating to people, plain and simple.

Re:meh...

[xrobertcmx]

Transmeta just got out of the processor market and sold off the line as far as I recall.

Re:meh...

[jafuser]

[LAFKON] - A movie about Trusted Computing

And to think...

[Concern]

This stock was at $5 not so long ago... :D

Re:And to think...

[Anonymous Coward]

When? 1992? I'd say that was so long ago.

I bought at $13 and it split a few times and I recently sold it all.

Re:And to think...

[Concern]

Ah... 2 years ago?

Just wait 'til the Intel-based Macs come out

[Jim in Buffalo]

When the Intel-based Macs hit the market, Intel processors will be found in 52.6 percent of desktop PCs, so there!

Re:Just wait 'til the Intel-based Macs come out

[danpsmith]

I don't understand this logic. Why would more people buy a Macintosh just because it has a different CPU in it? I understand that there exists the possibility to dual-boot Windows once someone figures out how, but what does this mean to the average consumer?

Whether it has an intel in it or not, it's still a Macintosh running MacOS, so why would the average user want it more?

Re:Just wait 'til the Intel-based Macs come out

[Anonymous Coward]

In the words of Foghorn Leghorn, it's a joke son.

Re:Just wait 'til the Intel-based Macs come out

[P3NIS_CLEAVER]

And once the transition is complete, expect to see Apple move to AMD. They are playing Intel like a fiddle.

Re:Just wait 'til the Intel-based Macs come out

[Browncoat]

I think a lot of average consumers will be more willing to buy a mac with an Intel chip because it's about brand recognition. The more unfamiliar technology there is, the less people are going to want to learn it. If they recognize something, they might be more inclined to take a second look.

Re:Just wait 'til the Intel-based Macs come out

[Anonymous Coward]

hahahaha.. i doubt 5.26% of desktops sold are apple.. your an idiot..

Re:Just wait 'til the Intel-based Macs come out

[Jim in Buffalo]

It wasn't logic, per se, it was a joke. I expressed that the number would notch up by a tenth of a percent in Intel's favor because of Apple's transition, making a gag out of Apple's notoriously low market share. If anything, the Intel-based Macs should impact Intel's market share far more than that. I know that I'm planning to buy an Intel-based Mac this year as soon as I can, so I'll have an Intel-based computer on my desk for the first time ever, for whatever that's worth (ain't much, I know).

Re:Just wait 'til the Intel-based Macs come out

[Anonymous Coward]

Didn't this same article get posted last quarter? This is retail sales. AMD has been making gains but these sales are not helping their bottom line much at all. Honestly, how much can AMD make when somebody buys a $400 PC at best buy? What percentage of PC's are purchased at retail stores? If I buy a Mac, it will be online and therefore will not count in the stats anyway.

Re:Just wait 'til the Intel-based Macs come out

[ls -la]

I agree. When I was looking at laptops for college, it took a good two months of research to convince my mom that an AMD64 would run anything an Intel could.

Re:Just wait 'til the Intel-based Macs come out

[antek9]

for whatever that's worth

It will keep the house warm in winter, if that's nothing I don't know what is...

Re:Just wait 'til the Intel-based Macs come out

[justsomebody]

I think a lot of average consumers will be more willing to buy a mac with an Intel chip because it's about brand recognition. The more unfamiliar technology there is, the less people are going to want to learn it. If they recognize something, they might be more inclined to take a second look.

I call that BS

1. People could run Windows on their Macs before. Why didn't they jumped on that bandwagon? MS did make NT for PPC. Project failed.
2. OSX might be attractive to some, but still less attractive than cheap to average consumer.
3. Wine won't really help to run Windows software, I was hoping for that option for a long time and lost my interest nothing else. Wine just leaps behind MS, nothing else. Always some compatibility problems.
4. People will still hate to dualboot or run OS in a window. They could all run Linux, Windows, FreeBSD for a long time now if they would want it
5. OSX will still be just another world than Windows
6. Coders will still write Windows only software. Especially if they know you can install Windows on your Mac. They'll just inform you, that you have to install Windows. Meaning Apple computer will be just more expensive for Windows retail
7. Put another OS on same arch, it is still unfamiliar technology. Put same OS on two different architectures, it suddenly becomes familiar technology.

Re:Just wait 'til the Intel-based Macs come out

[nova20]

I don't understand this logic. Why would more people buy a Macintosh just because it has a different CPU in it?

Yes, the poster was joking, but I think he has a point.

First off, I think Macs are getting more and more popular -- OS X kicks ass, and Apple's getting recognized because of the iPod.

I also think that a lot of mac users get frustrated because all the popular hardware is for windows and they can't use it.

Thus, when micro$loth does finally get off it's high horse and write a 32-bit OS that supports the machine's alternative to a BIOS (or the Core Duo 64-bit comes out), mac users will rejoice... and more people will see that it's the best of both worlds -- windoze for all the software you just can't get on a mac, and the nice, friendly mac interface where you can do everything else that you normally do.

Re:Just wait 'til the Intel-based Macs come out

[Bing Tsher E]

You could never run Windows NT on a PPC Macintosh. You could run NT on a PPC Reference Platform machine, commonly known as a PREP system. There are some PREP systems (if I am not mistaken) that were targeted as Mac clones that could run NT. But when I ran Windows NT PPC, it was on one of the light-duty RS/6000 machines that was a PREP architecture machine. It had integrated S3 Trio 64 graphics, PCI and ISA slots, IDE drives... it was essentially a PC-clone with a different processor. It ran either AIX or NT (and NetBSD, of course, but everything eventually will be capable of running NetBSD). I have one of the Power Computing Mac clones, which I believe is also a PREP system (it will run MacOS, but also has PS/2 keyboard and mouse, PCI slots, IDE drives, etc. etc.,) and (maybe) capable of running NT, but I haven't tried. That machine, actually, I got cheap at an auction and haven't done anything at all with yet.

Does anyone even know what chip they have?

[ZachPruckowski]

I don't mean this in a negative way, but what percentage of the computer buying public even knows about AMD? I mean, it seems to me that the average person couldn't tell you what chip is in his computer. I mean, the answer I usually get to that question is "Dell" or "HP". So basically, what I'm saying is that it may not be AMD chips that are doing well, but the particular brands they're in?

Re:Does anyone even know what chip they have?

[jhutch2000]

There is some truth to that, but also, in my experience, those same people didn't pick the computer on their own. The asked their local computer geek, who DID know the difference, what they should buy and then did so. I know I'm personally responsible for computer purchasing advice for around 20 computers in the last couple years. And *I* know the difference, even if most of the people getting computer did not.

Re:Does anyone even know what chip they have?

[Arthur B.]

This is actually a Good Thing. The less people care about the chip in their computer - as long as it works - the better for competition. The people who care are either -very informed, a small minority -sensitive to marketting (I wanna this pentium thing, it will speed my internet connection, they say it in this commercial)

Re:Does anyone even know what chip they have?

[MobyDisk]

...the answer I usually get to that question is "Dell" or "HP."

Even worse, most people I know say "Microsoft." Who else here gets the mom and pop calls that start with "I'm having a problem with my Microsoft. Can you help?"

Re:Does anyone even know what chip they have?

[Dun Malg]

Who else here gets the mom and pop calls that start with "I'm having a problem with my Microsoft. Can you help?"

Another favorite is "Adobe won't load".
"Adobe WHAT? Adobe is a company, not a product."

"You know, Adobe. For reading PDFs."
or (slightly less common but still strongly represented)
"You know, Adobe. For doing stuff with pictures."

Re:Does anyone even know what chip they have?

[Anonymous Coward]

Basically, everything you said, applies to Intel as well. For the love of Bob, people who are bombarded with Intel advertisements every freakin' commercial break, can't tell you what the hell an 'Intel' is.

The general 'dumb' market is covered by the run of the mill PC manufacturers, which do pay attention to quality and performance of chips. And also cost.

The hobbyist market is damned well informed.

And the IT professional market? AMD is wiping the floor with Intel. One word - Opteron. Intel has no answer to the Opteron. Oh, how about another two words? Dual core. Intel's offerings are pathetic in comparison. Not to mention, AMD already has eight core processors planned.

(Sidenote. I want one. An eight-way, eight-core Opteron system. Oh, dear god.)

So, in short, the latter two markets can't be claimed as chip ignorance. Nor can the first - and AMD is certainly gaining ground there. Victory goes to AMD for this round.

Re:Does anyone even know what chip they have?

[lcsjk]

"I'm having a problem with my Microsoft. Can you help?"

So far, that seems to be the main problem I have with my computers also!

Re:Does anyone even know what chip they have?

[Anonymous Coward]

Yeah, but in all fairness, "Reader" is a stupid name and makes people think it's a modifier of "Adobe." Most people call Photoshop "Photoshop," but the remainder probably refer to their cars by the make as well.

Re:Does anyone even know what chip they have?

[Dun Malg]

Yeah, but in all fairness, "Reader" is a stupid name and makes people think it's a modifier of "Adobe."

True, but my experience is from back in the olden days when it was still called Acrobat. stupid name too. What the hell does acrobatics have to do with documents?

Is AMD profitable?

[keester]

That's the biggest question in my mind. Market share is important, but will AMD be able to sustain whatever growth they have accomplished? So, within the last few years, they've opened up new fabrication plants, and probably they have more room for growth. Still, it will be interesting to see their earnings (revenue and profit).

Re:Is AMD profitable?

[beavis88]

Barely, but yes -

http://finance.yahoo.com/q?s=amd

Re:Is AMD profitable?

[Brian Stretch]

AMD reported earnings yesterday. They earned $0.45 per share, far above "analyst" estimates. They are very profitable and becoming moreso as they take big chunks out of Intel's hide.

Intel slowed AMD's run to profitability last year when they shifted production from chipsets to flash memory (they use the same fabs). Dumping flash below cost hurt AMD more than it hurt Intel... at the time. Then Intel ran short of chipsets. Oops!

Intel is rapidly undergoing death-by-management. Never let marketing weasles run a tech company. I haven't bought an Intel CPU since the Pentium 3, the last good desktop CPU Intel made.

Re:Is AMD profitable?

[sasquatch21]

http://finance.yahoo.com/q/is?s=AMD&annual

Re:Is AMD profitable?

[Stalyn]

"For the three months ended Dec. 25, AMD earned $95.6 million, or 21 cents per share, on sales of $1.84 billion. In the fourth quarter of 2004, it lost $30 million, or 8 cents per share, on sales of $1.26 billion."

"For all of 2005, AMD earned $165.5 million, or 40 cents per share, on sales of $5.85 billion. That compares with a 2004 profit of $91.2 million, 25 cents per share, on sales of $5 billion."

So AMD earned more money in the recent 4th quarter than all of 2004. And a 125.6 million increase for 4th quarter earnings from last year. No wonder AMD stocks are up so much today.

Re:Is AMD profitable?

[Anonymous Coward]

AMD stock seems a bit strange to me; anyone know why the stock can maintain price with a P/E of 545? Average for 1950-2000 is 16.4, Intel is 16.98. I do not understand what about AMD in particular justifies this extreme overvaluation and priced-in expectation of tremendous growth (more than the market even has for Google, which as P/E of 99.03).

Sure, AMD is making a lot of smart decisions Intel is not; but how fast can it regain market share, and at what point will AMD's earnings really justify its stock price? No real math behind this other than historical statistics, but 545 just feels about an order of magnitude to high.

Re:Is AMD profitable?

[ChrisMaple]

AMD's price/sales is about 2.5, below Intel's 3.5. As AMD improves its profit margin and pays off its relatively high debt, P/E will fall to a more reasonable range. It is this expectation of future profits that is driving AMD's price.

Re:Is AMD profitable?

[pjp6259]

Wow.. According to that, AMD's P/E is over 500! Is that possibly correct?

Re:So

[th1ckasabr1ck]

What are you talking about? AMD chips are very resonably priced - Certainly moreso than the alternative.

Re:Point of interest

[jhutch2000]

You aren't looking at comparable chips, then. At similar performance marks, the AMD chips are cheaper than their Intel counterparts.

Naming

[gamerluke]

I miss the days when a processor name included the speed of the processor. It is possible that intel is losing market share because AMD (and Intel to a certain extent) don't show their speeds in the names and so it harder to compare systems. I know that AMD gives you more for less Ghz, but not everyone knows that and Intel has a preceived advantage in speed they can't fully exploit.

Re:Naming

[RailGunner]

Intel has a preceived advantage in speed they can't fully exploit.

The only scenario where Intel wins is if the operation fits entirely in the CPU cache. Otherwise, Intel processors spend an awful lot of time waiting for data.

AMD, on the other hand, is far more efficient in data access and transport, and that's why, despite running at a lower clock rate, AMD wipes the floor with Intel in benchmarks these days.

Given the benchmarks that show AMD consistently outperforming Intel, how does Intel have a "perceived advantage in speed"?

Even the Athlon XP CPU's were faster than the P4 counterparts. Compare an Athlon XP 1900+ (runs at 1.667 GHz) to a similarly clocked P4, and the Athlon wipes the floor with the Intel CPU.

AMD's architecture and chip design are, quite simply, better than anything Intel is offering.

Re:Naming

[vux984]

I do to. But only because in those days the speed told you something useful. That hasn't been true for several years now. 2 GHz out of one model is the same as 3GHz out of another which implies that 2.5GHz out of the 2nd model is slower than 2GHz out of the first.

At this stage it would be like asking car makers to put horsepower ratings in the name of each model. Consider a Lotus Elise 190 vs a Honda Prelude 190 vs a Ford Mustang 300 vs a Dodge Ram 235 vs a Porsche 911S 355 -- did the engine rating really add anything useful?

I mean the Elise is like half the the 911S, and the lowest rated, yet its easily the 2nd, possibly even the fastest off the mark - meanwhile the 911 at 355 is just not built to haul your yacht home but the much weaker RAM will do it handily.

I mean yeah the number has meaning, and 'more is better'...but without the context of the whole package it doesn't tell you anything useful. There's no overridingly practical use that should make it part of the name of the product. It should be an available spec sure... but not the product name.

Re:Naming

[StewedSquirrel]

It would be absolutely irrelevant to try to compare a 2.6GHz Athlon 64 against a 2.6GHz Pentium 4.

The 2.6GHz Athlon 64 roundly beats even the 3.5GHz Pentium 4 by a substantial margin in "real world" applications. For example, on the "Multimedia Winstone" real world benchmark, the 2.6GHz Athlon 64 FX60 beats the Pentium "Extreme Edition" at 3.5GHz (both a dual-core and are top of the line chips in their respective lineups). The benchmark shows that the Athlon is fully 20% faster than the Pentium 4 despite having a clock speed that's 35% slower.

Clock speed is meaningless to CPU performance today in the days of speculative operations and Out of Order Execution units and parallel pipeline processing, etc.

I'm not sure what you mean that "Intel has a perceived advantage in speed they can't fully exploit", because as we know, the P4 was designed to sacrifice performance to try to reach high clock speeds and that approach has failed to scale as well as Intel expected. AMD designed the Athlon to increase the work done per cycle and this approach has worked.

Intel has followed suit and based their next line of desktop processor development on the Pentium-M chip which was designed independently of the P4 (it is architecturally closer to the Pentium 3) because they realized the folly of their approach with their current chips.

Squirrel

Re:Naming

[tomstdenis]

This isn't exactly correct. While externally AMD has a good lead on Intel with the integrated low latency memory controllers and dedicated HT links [e.g. think MP].

Internally however, AMD is behind. Intel has 2 cycle caches ... at >3Ghz whereas AMD has 2 [and 3] cycle caches at 2Ghz. I'm talking L1 specifically... So internally Intel can move data much quicker than AMD could ever hope to.

However, AMD has more resources dedicated to execution. The problem with Intel is hardly to "get data to the core" it's "get the core to actually do some work!". There are a lot of pipeline hazards that AMD processors don't have and the AMD processors execute a lot of opcodes faster.

For instance, really trivial but comes up in crypto a lot, "adc" which is add with carry. It takes one cycle on the AMD cores and 1.5 on the intel. "mul" takes 2/6 [2/8 for 64-bit mode] cycles [the slash is throughput/latency] whereas Intels are 10-15.

Then there is the fact you have THREE ALU pipes. So if all is going well you can have an IPC upwards of 3 [the decoder can only issue three opcodes per cycle] but in theory you could retire more [think FPU]. While the IPC over an entire program is likely to be less than 2 certain hotspots could easily get over 2.

Bignum multipliers like those in TomsFastMath [iirc] routinely hit over 1.5 for IPC, I don't have the exact values handy but they're certainly pretty fast.

Intel cores, specially for ALU operations just lack IPC. So the clock may be 1.5x faster but the IPC can be 2x to 3x lower.

Tom

Re:Point of interest

[Albanach]

You don't tell us what you're comparing - it's like saying I can buy a Ford for 15k or a Merc for 30k therefore Merc are uncompetitive.

pricewatch.com Says the slowest Sempron being produced is the 2200+ and you can have it delivered for $57. For $60 you can get a 2.2Ghz Celeron which is no match for AMD's processor. The 2.2Ghz P4 costs $79 delivered, $22 more than the AMD offering.

The reason all those AMD chips appeared before Christmas was because they are so competitive at the lower end. When you match that with their server options AMD are wiping the floor with Intel at almost every level.

Eh..

[Kranfer]

Reading this brings up bad memories of working at Best Buy During college. However, while I am happy to see a smaller company like AMD make a dent in the PC market, I have to say, I would still never buy an AMD based system. I have had AMD based systems in the past... Issues that I have had included overheating and missing bench marks, so I switched to Intel and have never had an issue with my P4 processor. And now that Intel is being put into the new Macs, I think we will see Intel take the lead again this year with the AMD users grumbling that they are in the minority of new PCs again. Either way I personally don't care, But I would prefer an Intel processor over AMD just because of sheer quality.

Re:Eh..

[colinrichardday]

And what is this "sheer quality"? I have two AMD machines and a P IV, and the former haven't troubled me too much (at least not the chip set).

Re:Eh..

[Guysmiley777]

It's a perfect example of the amount of bamboozling a good marketing campaign can generate. Textbook example, actually.

Intel goes outside

[digitaldc]

"The new slogan is supposed to signify Intel's shift away from focusing "inside" and starting to look at platforms and solutions for the end users."

(From an earlier discussion and article.)

Now I am beginning to understand why Intel has made the decision to start focusing elsewhere.

How's the laptop market doing?

[Fearan]

With the decreasing market share of desktops in the consumer computer market, I'm interested in knowing how AMD is doing in the laptop sector and total overall processors sold in comparison with Intel. Most people I know wouldn't consider anything other than Centrino for some reason that I don't understand (marketing?) Furthermore, how will Apple's new MacBook and other Intel offerings affect AMD's apparent marketshare takeover?

Re:How's the laptop market doing?

[VisiX]

Most people I know wouldn't consider anything other than Centrino for some reason

I know exactly why this is because I worked at Best Buy for three months. When centrino was introduced there were commercials everywhere. These commercials, intentional or otherwise, made it seem like centrino was the only way to connect wirelessly saying things like "Centrino technology is a huge advancement in wireless networking for people on the go". This statement was true, as centrino allowed for longer battery life and generally laptops using centrino were lower weight, allowing for longer use and easier transport. This was not conveyed by the advertisements though, as the vast majority of people coming in to buy a laptop thought that centrino and wireless were the same thing.

Re:How's the laptop market doing?

[helmespc]

I have an AMD laptop.... it runs much to hot to do anything more than surfing the net and such.... sometime writing code... thats my biggest problem with AMDs in general... they run HOT..... so much so that I have actually considered heading to the dark side....

Re:How's the laptop market doing?

[jetxee]

Laptop market is growing. And when I want to change my laptop, I shall definitely look for lighter/slimmer/longer-running/quieter piece of hardware, and performance/64bit will matter less.

So, whether I buy new MacBook or Intel Somethino (tm), it is likely to be Intel.

And I strongly suspect, that every laptop chip is today more profitable, than 5 desktops-for-sale.

Re:How's the laptop market doing?

[MarkVVV]

For God's sake, stop calling it a "centrino"...it's a fscking Pentium M!!!

Centrino is the chipset used in those notebooks.

Re:How's the laptop market doing?

[aschlemm]

I don't really know if it's the marketing by Intel but the Pentium M is very good in a laptop. It runs cool and the battery life is quite good IMHO. When I look around the office every laptop I see is a Pentium M based one. At one time I had a laptop with a 3.06GhZ P4 and it ran so hot I went through 3 mobos in 2 years. I was real happy when I finally got a laptop with a Pentium M in it instead as my laptop is no longer cooking itself in its own juices.

Re:How's the laptop market doing?

[Kamots]

Your heat comments would make sense a few years ago, but Intel and AMD have kinda swapped places in the heat arena. Intel is the one that's putting a space heater in your machine now.

Although, what's going to happen as the switch to multi cores continues... who knows.

Re:How's the laptop market doing?

[Cyphertube]

And what kind of AMD laptop do you have? What processor do you have stuck in there? I know that a lot of people buy the wrong processor when they want something to run cool. That's one reason why the Pentium M keeps going well.

On the 64-bit side of things, most retail companies continue to sell the Turion ML series, for which the processor costs $5 less that the MT, but uses another 10w of power. When it comes to running cool, this makes a huge difference. Hence, why my wife's laptop was built into an MSI notebook with a MT chip. Much cooler. Makes my old Pentium Celeron seem exceedingly hot, and runs about the same as my Pentium M from work. Moreover, processing-wise, her machine kicks my work machine's ass.

Re:How's the laptop market doing?

[ls -la]

When I look around the office every laptop I see is a Pentium M based one

Are your laptops purchased by the company? Companies tend to buy all their laptops in bulk and as similar to any they all ready have as possible (it makes support easier).

Re:How's the laptop market doing?

[colinrichardday]

Don't know about longer lasting, but my Athlon-equipped Averatec 3150H is lighter (4.3 vs 5.6 pounds), about as slim. As well as a good deal less expensive.

Re:How's the laptop market doing?

[wx327]

Centrino is the chipset used in those notebooks.

Close...

From Intel:

The technology represented by the Intel Centrino mobile technology brand combines the Intel® Pentium® M processor, the Intel® 855 Chipset Family and the Intel® PRO/Wireless 2100 Network Connection. All components were optimized, validated and tested to work together with mobility in mind.

Re:How's the laptop market doing?

[aschlemm]

Yes and we even have a mix of companies in our office right now since our current project has a number of contractors that have their own laptops (not supplied by us that are mostly from Dell from what I've seen).

Re:How's the laptop market doing?

[jetxee]

Well, that probably means, that AMD has also a chance on laptop market. If it is as good as Intel or better, I shall not mind buying it.

Re:Point of interest

[Daath]

But TFA also says that AMD has >52% of the desktop CPU market.
I don't know where you shop, but where I shop, AMD is waaay cheaper than Intel, and always have been. You get a lot more bang for your buck with AMD, especially if you are a gamer!

Affordable 64-bit

[zlogic]

When I recently checked OEM prices for CPUs, it was interesting that a 64-bit Sempron costs only slightly more than a Celeron with roughly the same perfomance.
And Intel doesn't have any reasonably priced 64-bit chips...
It's also interesting what x86 imacs would be like if Apple chose 64-bit CPUs, because, well, they had 64-bit G5 before.

Re:Affordable 64-bit

[Plocmstart]

I think it's Intels refusal to bring 64-bit CPUs into the consumer market. Intel has previously stated that a home user doesn't need 64-bit while AMD's marketing scheme is riding on the whole idea that the larger the number the better it must be (and in some cases it is). When it comes to most non-computer savvy purchasers that is what works. Even though Intel isn't going to disappear, they are going to continue to have good competition until they realize this.

Re:Affordable 64-bit

[jedidiah]

AMD probably just thinks that it will cost less to have dual production lines. If they don't have to worry about engineering or supplying any ia32's that's more resources that could be devoted to other things.

Re:Affordable 64-bit

[Mostly a lurker]

Just as a matter of interest, which Celeron do you believe has roughly the same performance as which 64-bit Sempron, and for what general type of computing?

Re:Affordable 64-bit

[jandrese]

I put a "64" bit semptron in a machine I built for another person (who was on a very limited budget) and it was my understanding that the chip was based on a 64 bit chip design, but was intentionally neutered down to 32 bits in hardware. I mean it would be great if she actually has a 64 bit machine, but I doubt she'd get much use out of the extra address space anyway. Frankly, people buying Semptrons probably aren't going to be doing lots of things that would benefit from a 64 bit chip anyway. I'm sure in 2-3 years there will be lots of things that run considerably better on a 64 bit architecture, but currently you have to be running huge databases or doing scientific computing to see much benefit.

Re:Affordable 64-bit

[the quick brown fox]

That was true when the Sempron first launched, but last summer they added 64-bit-enabled Semprons to the lineup (I think to counter 64-bit-enabled Celerons). The Semprons with 64-bit capability enabled are generally marketed with "64" in the product title, like this.

Re:Affordable 64-bit

[Guysmiley777]

Of course as it stands right now "64-bit" is a marketing gimmick. It doesn't increase gaming or desktop performance at ALL. If you're running a SQL server and need 16GB of RAM or something, then yeah, a 64-bit system is what you need. But Quake 4 will NOT run faster with a 64 bit CPU.

I'm a huge AMD fan (hell I have 4 AMD PCs in my house), but there seems to be a perception that 64 automatically is better. Well, IF your OS supports it AND your peripherals support it AND your drivers support it AND the app you're running supports it you MIGHT get performance gains, depending on what you're doing.

Re:Affordable 64-bit

[jamesborr]

This isn't technically correct. I would argue if if want more then 3GB, then you need a 64 bit CPU, a 64 bit OS AND, a >32 bit BIOS/motherboard architecture. Else, you can forget about actually using any of that RAM above the 3GB boundary, as the system board will use that memory address space for various controllers and such (i.e. drive controllers, video controllers, etc.

Re:Affordable 64-bit

[corngrower]

I'm a huge AMD fan...

I thought it was the Intel chips that needed huge fans these days.

Seriously, I have 3 intel based machines. All were surplused. I think I paid a grand total of $65 for them, less than the cost of the AMD processor in my newest machine.

Re:Affordable 64-bit

[LWATCDR]

"Of course as it stands right now "64-bit" is a marketing gimmick. It doesn't increase gaming or desktop performance at ALL. "
That depends on if you are using 64-bit programs. On non-brain-dead chips like the PPC and Sparc 64bit code is only a benefit for programs that need to use more than 4 gigs of ram or do a lot of 64bit int math. The X86 is registered starved so the extra registers you get in 64-bit mode do help performance on every program that is compiled for 64bit mode.
It is kind of like way back when one of the columnists for Byte told people to not waste their money buying 16mhz 386sx machines because you could buy a 20 mhz 286 for a little less money that run dos just as fast. He did say that the 386 ran Windows better but who used Windows?

El cheapo?

[C10H14N2]

Are you KIDDING?

AMD is successful because from day one they've been in the business of making better products, not cheaper products. That they happen to be cheaper in some cases is just a sign that they have a successfully diverse product line.

Re:El cheapo?

[Twid]

I'm calling you out on the "from day one" statement. The AMD K5 was not exactly the pinnacle of performance, features, or price competitiveness. AMD is doing well now, no argument there. I'd like to see an article that compares total chip chipments worldwide, though, rather than say limited statements like "52% of all retail desktop sales, in the USA, in the 4th quarter".

In related news, my pants were the leading distribution method for iPod nanos, in the USA, in California, in my house, yesterday.

Re:El cheapo?

[jandrese]

Yeah, the original AMD chips were generally thought of as "not quite as good as Cyrix". Anybody who cared one whiff about quality back then went Intel. It wasn't until the K6 line that AMD started to really position themselves as a quality chipmaker, and it wasn't really until the Athlon line that they pulled themselves out of the pit of the Computer-Show Beige Box hawked by some greasy fat guy crowd.

Re:El cheapo?

[dougmc]

from day one they've been in the business of making better products, not cheaper products.

Eh? I don't know about `day 1', but it wasn't that long ago that AMD was lagging behind Intel in terms of performance, power consumption (though that wasn't such a concern back then) and such. For example, the K5 was intended to compete against the Pentium chips, but the Pentium Pro came out almost immediately after the K5 did and it blew the K5 away. The K6's came closer to beating the Intel offerings, but even then, the Intel chips had a small performance lead, and the fact that 3Dnow never took off further hurt the K6 chips. Back then, people bought AMD because it was cheaper, not because it was better.

Going back even further, the AMD 8086, 80286, 80386 and Am486 chips generally were just clones of the Intel offerings -- with similar performance, but coming out some time later at a lower price.

But things have changed. AMD has finally caught up to and passed Intel in many respects, and I suspect that the reason that Intel is still selling so many chips is more due to interia than anything else.

Re:El cheapo?

[AzsxQuii]

Hmmm. Ineteresting that no one mentioned the Intel ID fiasco. When P3 started comming out with serial numbers on the chips which allow internet connected computers to be tracked (Later Mobos had feature set which allowed this to be turned off). This swayed me from future intel purchases. My first AMD was the K6-2 and back then the bang for buck was high. Also, people should realise that in developing countries people dont care about brand, cost goes a long way and AMD will win out on this share of the market.

Re:El cheapo?

[SilverspurG]

I had a Cyrix 166. They actually had a chance at one point.

Re:El cheapo?

[Anonymous Coward]

um, day one? No, maybe now, but certainly not day one.

My early AMD experiences

AMD 486sx/2 : Wow, its like a whole new defination of suck.
AMD 386-40mhz (i had heatsink and fan on it even tho not required) : FRIED
AMD K6: Had that wonderful, opps, sorry if we address more then 16mb of ram there are errors
AMD K6-2: Dunno if i can blame AMD, but warcraft AND myth wouldn't run on this processor :(

Also i'd like to point out that early AMD procs always ran hotter then their intel counterparts, until the 5v pentium 60/66 was introduced.

Re:El cheapo?

[Anonymous Coward]

I don't think so.. Back in the early 90's when a 386 was king of the processors, AMD's 40MHz chip was always faster than Intels DX33 chips. I've been an AMD fan ever since. When I worked on my first Cyrix desktop, I remember thinking how it was a slow piece of crap. Then I found out they had such skimpy caches, no built-in FPU, etc. From my perspective Cyrix was NEVER better than AMD when comparing any desktop chip they made.

Re:El cheapo?

[Anonymous Coward]

"52% of all retail desktop sales, in the USA

Shesh - we get one of these aricles every 2 weeks. Retail solely in the US is no longer the end of the story folks. In other news, UNIX is on 100% of all developer machines in my basement garage.

Get the AMD/Unix fanboys off the article selection committee around here and do some real research.

Re:El cheapo?

[Beatbyte]

In related news, my pants were the leading distribution method for iPod nanos, in the USA, in California, in my house, yesterday.


coming out of any /.'ers pants, it's got to be nano!

Re:El cheapo?

[jonadab]

> In related news, my pants were the leading distribution
> method for iPod nanos, in the USA, in California, in my
> house, yesterday.

Ah, but my research shows that iPod nanos are on the way out, because the results I've just collated indicate that xmms is now the leading method of listening to music for middle-aged college-educated male computer geeks between the ages of 30 and 32 using FreeBSD, in small-town USA, in Ohio, at my house, for the last several days running. Also, according to the same study, the leading music writer now is apparently J.S. Bach, with the performer market shared among several major players, notably Wendy Carlos, Karl Munchinger, and the Stuttgart Chamber Orchestra. Meanwhile, AMD market share hovers around 50% adoption in the same demographic.

Re:El cheapo?

[Twid]

zing! :D

Re:El cheapo?

[Anonymous Coward]

Ohh..but it played Quake so poorly. I remember the dark days....Intel truly made the better chip back then.

Re:El cheapo?

[vsprintf]

In other news, UNIX is on 100% of all developer machines in my basement garage.

Isn't it difficult to get your car out of there?

Re:El cheapo?

[Anonymous Coward]

Hehehehe.. chipments.

Re:El cheapo?

[binary paladin]

Let me first state that I have been using AMD chips since my K5-133 and had a K6, K6-2, K6-III and multiple itterations of the Athlon.

"The K6's came closer to beating the Intel offerings, but even then, the Intel chips had a small performance lead..."

Small? I guess if you're talking about office applications. (In fact I believe in the realm of the integer, K6 had the lead.) The P-Pro/PII/PIII were the undisputed kings of the floating point in the age when 3D gaming was becoming "the thing." Even when programs utilized 3D Now! the PII/IIIs usually came out on top. The Athlons even had issues with the PIII evenly clocked in the higher ranges (1GHz comes to mind) until the Athlons moved away from the slot and put the cache on die (since if I remember, the cache on the higher clocked models was running as low as 1/3 core speed).

It wasn't until the P4 vs AMD64 war rampted up that AMD had an obvious advantage in most respects. I guess I'm taking your "not day 1" thing a little further! I talked way more than I needed to.

Re:El cheapo?

[Anonymous Coward]

Sony.
They once made an improved X86 clone, and now come out with the cell. They blew it big time - they could have been #1 processor maker if they brought out MP3 players/ipods. Dunno why AMD does not quietly ask IBM or Sony to assist.

Re:El cheapo?

[bored]

Your confused, or you may have been running a 486. The 6x86 line from cyrix was _VERY_ competitive with anything on the market. There only problem was one of perception, its like intel paid people to write bad reviews or something. When cyrix released the 6x86 it was suppost to work in every motherboard on the market. The only problem is that it required more current than the pentium's because it was a fab generation behind and a good percentage of the motherboards were unstable with the extra current draw. This caused a motherboard compatability list to come out, which everyone promplty ignored, then M$ disabled the L2 cache on cyrix for windows NT because the only cyrix they had (all of M$ had 1 cyrix) couldn't run it without crashing. Turns out that it was running on an unsupported motherboard which was getting bus reflections when it was doing burst transfers. The final blow was that everyone started using quake as a benchmark. While the cyrix 6x86-133 was labeled as a PR166 it ran more like a P90 in quake. This is not unusual even today, I can find benchmarks that prove the P4 is way faster than the P ratings on the AMD's. Its just a case of which benchmark you trust. Turns out that ID software had optimized quake for the pipeline arch of the pentium and didn't have an alternate code path for the cyrix which had a very strong math coprocessor but it didn't do back to back multiplies like the Pentium did. So some selective benchmarking pretty much was the final blow for a very good product. I ran linux on my 6x86 long after I purchased an MMX pentium to play unreal. The 6x86-166 was far more responsive running X and assorted other crap than the MMX-233 I had was. At that time AMD was draging its feet getting the K6 out.

BTW: I am old enough to remember the 286 and granted AMD's 286-20 and their 386-40 were reasonable processors, but they didn't compete with intel on the high end. By the time the fast (>16Mhz) 286's were out intel was selling 386's at the same clock rate and the fast 386's were competing with 486's.

Re:So

[Anonymous Coward]

Considering theyre already cheaper than intel, what you using now, a dorito?

Someone PLEASE mod this troll down...

[Anonymous Coward]

before a ton of noobs flame him...

Two factors that are important!

[ami-in-hamburg]

1) Most people that buy PCs have no idea what processor it has inside. They are generally motivated by price. If two PCs with the same performance and options are sitting next to each other at Best Buy, and one costs $150 less than the one next to it, which one will people buy? No brainer!

2) The speed of most newer computers is so ridiculously fast compared to just a couple of years ago that the processor just doesn't matter to the average PC buyer. Most people want to read their email, surf the web and store their digital photos. They are not running CAD or compiling code or editing full feature films. Even the lower end PCs available in stores these days can perform the average tasks just as well as the high end system. Therefore, again, the purchase experience will be motivated by price!

Re:Two factors that are important!

[Zerbs]

You've hit it right on the nose. The fact is that CPU performance has outpaced the software performance demand for typical home users and business users. When I help someone buy a new computer, unless it's for really high end stuff, I generally tell them to buy one with a slower processor, and use that extra money to get more memory, better graphics card, etc.

Re:Two factors that are important!

[Anonymous Coward]

From my experience working at Best Buy, definently more than half knew the difference between AMD and Intel. Well, more accurately they knew there was a difference at least. Even old people know about clock speeds now, since it's been pounded into everyone's heads for the past decade. And now just as soon as everyone asks what the "megahertz is" of a computer, it almost ceases to be an issue. Anyway, most people DO know there are multiple brands, and more often than not ask what the differences are. It's too bad that most BB employees have an overly-simplified explanation fed to them by their supervisor who knows even less about computers than they do. I love the atmosphere a total lack of product training in the retail store has bred. /rant

Anyway, that was my experience, not saying it wouldn't be different elsewhere though.

Re:Two factors that are important!

[MtViewGuy]

The speed of most newer computers is so ridiculously fast compared to just a couple of years ago that the processor just doesn't matter to the average PC buyer. Most people want to read their email, surf the web and store their digital photos.

That's true if you're running mostly business apps or surfing the Internet. But with people editing multimedia files downloaded from digital still cameras and/or MiniDV/MicroDV camcorders, such file editing does require quite a lot of CPU power. For example, if you want to do things like color, brightness and contrast adjustments, not to mention removing things like scratches and dust from old images scanned into the computer using a program like Paint Shop Pro, you definitely want a faster computer because these image-editing programs do use a lot of CPU time to process the image.

Re:Two factors that are important!

[porkThreeWays]

You hit the nail right on the head. For most PC users, hardware speed doesn't matter that much anymore. It's all so damn fast you have to be doing something special to need more speed. Something more than email and web browsing anyway. The big culprit for slow computers is garbage on computers. And it's sad that you can buy a 4,000 dollar computer at best buy and in a week it be slower than a lean P III 1 gigahertz box. We get fast computers all the time of people who don't realize it's not the hardware, it's the 50 running apps from an OEM HP, Dell, whatever, install.

But still DRM

[Harmonious Botch]

Last I heard, AMD was still working on DRM just like Intel. So what if they have 50+% of market. I don't see a substantial difference.

intel rise

[the_tsi]

And they won it all back when every reader of slashdot bought a macbook pro and/or intel imac after being brainwashed with Apple stories for four consecutive days.

Shift in importance from hardware to software

[Caspian]

It may be that a part of the reason for this change is the shift in importance from hardware to software.

It seems to me like more and more, people simply do not care what the hardware is so much as they care about what the software is. A few years ago, clueless consumers were demanding the "Pentium" brand (not even knowing what that word really meant); now, they simply ask "Does it have 'Microsoft XP'?" The answer, of course, is always "yes", so they ask no further.

Now that Mac OS X runs on both PPC and x86 machines and Windows XP on both x86 and x86-64, I think we are moving towards an era where the software matters more than the hardware (at least, from the perspective of Joe User).

Re:Shift in importance from hardware to software

[Anonymous Coward]

The shift from hardware to software happened when the IBM PC clone market came to be. We already shifted from hardware to software. I think what is happening now is that we are starting a shift from software to ?? {services? back to hardware?}. Linux, is starting to make that happen. Just look at what Sun is trying to do? Are they a hardware company (AMD, Sparc hardware)? software (free Solaris)? What value do they bring to the table? What about Google? Software is becoming the new commodity -- at least the OS layer is. OS vendors are packing in as much features into their OS as they can to keep them afloat so that they can continue to have an advantage over their market.

Apple just happens to be in a really good position now as they already know this. The Mac OS is already free; it's call Darwin. What's not free is Apple's windows manager and their application API. Remember, Next STEP? Now the foundation of Mac OS X. Well, NeXT came out with a product called OPEN STEP which allowed your to compile your NEXT STEP applications on other OS like Solaris, Windows, etc. Well, Apple still has that option.

Re:Shift in importance from hardware to software

[Shivetya]

If that is the case then OS/X is in even more trouble as the hardware pricing is totally controlled by Apple and price is driving the PC market.

Hardware is what Apple used to celebrate as being their important difference, now that they are on Intel platforms they have to rely on design. Trouble is the public wasn't keen on buying their design before and since the price didn't change neither will the buying public's opinion.

Re:Shift in importance from hardware to software

[level_headed_midwest]

That more or less happened about 15+ years ago when about everybody had 80386s and 80486s and could run 32-bit x86 code. As of a few years ago, you could get x86_64 chips, but they can all run 32-bit code natively.

Re:Shift in importance from hardware to software

[Bob 4knee]

...A few years ago, clueless consumers were demanding the "Pentium" brand (not even knowing what that word really meant)...

I never did figure out what the word Pentium meant. Could you enlighten me?

Re:Shift in importance from hardware to software

[AK Marc]

I never did figure out what the word Pentium meant. Could you enlighten me?

After a ruling where Intel could not trademark processor classes by number (386, 486, 586), Pentium was invented as a word to call forth the image of "5" (pent-) to imply the 586 class processor, but use a word that could be trademarked so that they could reduce confusion with competitors. So it is a translation of "586" into a trademarkable word. Though, it did evolve later, as the didn't go with Hexium, Septium, or such for later itterations, but opted to go with PII, PIII, P4...

AMD to the rescue ;-)

[Lonewolf666]

Now that Apple goes x86, the step from Intel to AMD might be easier than from IBMs Power chips to something else.
So if Intel gets too aggressive on Apple, we might see Apple computers with AMD CPUs :-)

Speed?

[Nazmun]

Err... you mean mhz? Going away from the netburst architecture intel would shoot themselves on the foot. A 2.0 ghz pentium-m can outperform a 3.2ghz p4 in many cases or come very close in others.

Consumers would think a 2.2 ghz p-m type core to be inferior to a 3.0ghz p4. Thats not a correct indication of "speed" either. Hell you can't simply just say the word speed anymore. Theres so many things to consider.

Intel plays it smart.

[oilisgood]

I like how Intel casts AMD as "competition." Intel has a market cap of $153,860.10(Mil) AMD's market cap is $13,277.70(Mil). Last years gross profit for Intel was $23,049.0(Mil).

So... Intel's gross profit for about 7 months would be ALL of AMD's complete market capitalization. How can you call that competition? What a great way to dominate a market, while keeping the feds off your back!

Re:Intel plays it smart.

[Guysmiley777]

Umm, where did you get your numbers? Intel's market cap is $136.56B and AMD's is $15.06B. Intel is a mammoth company, but the issue here is their current product line is measurably weaker than AMD's.

With regards to competition - I want to build a PC. I can build an Intel based box or an AMD based box. How is that not competition? Do you think consumers think "Wellll, the AMD CPU is faster, cooler and cheaper, but boy their market cap just isn't impressive enough."?

Re:Intel plays it smart.

[Anonymous Coward]

Umm, where did you get your numbers? Intel's market cap is $136.56B and AMD's is $15.06B. Intel is a mammoth company, but the issue here is their current product line is measurably weaker than AMD's.

He didn't take into account yesterday's earnings reports ahd to the stock prices. ;)

Re:Intel plays it smart.

[oilisgood]

It is hard to have today's blip in numbers count in the evaluation. (And for a quarterly report at that) I am not arguing that Intel has better processors than AMD. I am saying that Intel makes more money and that is really all Intel cares about. AMD can make a better processor and I bet Intel could care less. I would think that for Intel, as long as they dominate market share and do not have the feds on their back it is all fine with them.

Re:Intel plays it smart.

[Deskpoet]

You obviously missed the poster's point.

Intel has so many fabs that they number them in the 20s, with new ones coming on-line (and doing 65nm or less and 300mm at each instance) all the time. Q4 saw Intel gross 10.2 BILLION, and still disappoint. Do you honestly believe that if there wasn't anti-trust Intel wouldn't *BURY AMD ALIVE* with its production capacity, market partnerships, etc.?

When the next depression hits the US--not that far away, really--Intel will be able to sit out YEARS of underproduction while still slashing AMDs margins with cheaper chips made in state-of-the-art manufacturing facilities THAT THEY OWN THEMSELVES.

People want to see the giant fall, but the reality is that Intel will be standing tall at the end. AMD exists to ensure Intel's existence. While it may be an embaressment to Intel to not have the best product line, the numbers are what count, and Intel has plenty of that.

Re:Intel plays it smart.

[snarkh]

Intel's marketshare has just dropped by 2-3% and they are certainly extremely worried about it. Intel's huge market cap needs correspondingly huge profits
and Intel's profits came below expectations because of AMD.

Cache...

[brunes69]

Yeah but a Sempron 2200+ will stomp all over a 2.2 Celeron. It has way more cache ( 128k/256k in the Sempron vs 8k/128K L2 in the Celeron) and also a generally better pipeline. You can't judge a CPU on MHZ alone.

Re:Cache...

[Anonymous Coward]

That's exactly what your parent said...

Re:Cache...

[Anonymous Coward]

Who the hell are you arguing with???

Re:Cache...

[lubricated]

>>Yeah but a Sempron 2200+ will stomp all over a 2.2 Celeron. It has way more cache ( 128k/256k in the Sempron vs 8k/128K L2 in the Celeron) and also a generally better pipeline. You can't judge a CPU on MHZ alone.

Yeah good thing that the MHZ for the sempron was not even mentioned. The 2200+ is just a label amd gives it and has little to do with MHz

Re:Cache...

[freidog]

Actualy the socket A Semrpon's are nothing special. They're stripped down versions of the old AthlonXPs with a new system that rates them against Celerons (So AMD can put bigger numbers on slower chips).
Now, the first Netburst based Celerons, the 400mhz FSB / 128KiB L2 parts, are some of the worst chips intel produced since the cacheless Celeron 300s...

A more appropriate comparison of budget chips today would be the S754 Sempron 2500+ - 3100+ against the Celeron D 2.53 - 3.06. They stack up fairly comperabely in overall performance (Sempron wins for games, Celeron wins for multimedia), and prices are almost identical 63-80 for AMD, 66-80 for Intel.

AMD still has the price advantage against many P4s, but in the budget world it's a much closer race.

Re:Cache...

[Hal_Porter]

I have an Athlon because it run the kind of applications I run faster than an equivalent Pentium 4. But I have a problem with these model numbers. Mhz was a kind of Gold Standard for chips. You basically bought your favourite (micro)architecture in the highest speed grade you could afford. No one could cheat on this, since Mhz was something you could measure with a scope, at least in practice. Then AMD had an architecture which was quick but harder to crank up to higher clock rates, so they moved to model numbers. It didn't matter though, because the processor performance was linked to Intel Mhz. They underpriced the chips too, to make them more attractive.

But now both AMD and Intel use model numbers, and they are quickly becoming more abstract. In an economic sense, we've lost the gold standard completely, there is now no way to tell how fast a chip is by looking at the label. Isn't there a danger of inflation in the long run, so AMD will sell the Athlon X4+ 64+ 96500+ vs the Intel Core Duo Pro Plus 840002, and both of them will run about the same speed as one of todays high end chips because both Intel and AMD will have fallen completely off the Moore's law curve?

And soon of course, there won't be any desktop architectures other than x86-64, and high clock rates / per core performance is being de emphasized. Could this be the coming of a plateau, where CPU performance increases gradually fall to some lower rate?

Re:Cache...

[kesuki]

yes, but i can have a venice core 3000+ in s754 for $126 intel doesn't even have an entry in the 512k cache space... they go from 256k straight to 1 MB, and the prescott core there in the same perfomance level is almost $50 more.

For a buget performer/overclocker a socket 754 venice is nice, because you can overclock it by at least 400 mhz with quality air cooling alone.

so yeah, although the socket 939s are on a line with the intel pricing for that level or performance, amd has some nicely priced socket 754's.

Re:Cache...

[GnuAge]

Actually, the Venice core is s939, not s754. And an A64 3000+ Venice core starts at about $149 on Pricewatch. That said, a Venice s939 3000+ is apparently a good choice for overclockers.

Re:Point of interest

[GmAz]

I used to work at CompUSA and was in charge of the desktop department so I saw every new machine come in and personally set up the demos and the price tags. I saw numerous HP, Gateway, Compaq and Emachine models come out with AMD64 processors around 3000+ to 3700+. The basically identical Intel model always cost at least $250 more. And since we, the salesman, had pretty much free reign as long as we sold computers put whatever we wanted on the computers. I loved doing benchmarks and the AMD always came out ahead. Only a couple times did intel beat out AMD and it was usually the new Prescott cores, though not new anymore, until AMD came out with their new cores. I don't consider myself a fanboy of one particular manufacturer, I am a fanboy of cost vs. performance. For the past four years, AMD has won me over.

RETAIL sales..

[sadr]

Note that this only applies to retail sales.

It does not include total sales, where AMDs market share is significantly lower. e.g. this report excludes Dell entirely. Overall, they're somewhere around 25% of total shipments.

AMD is taking marketshare away from Intel, but they are still a much smaller player.

Re:RETAIL sales..

[Anonymous Coward]

Exactly what I thought, this is hardly a minor point.

"(Sales by online retailers were excluded from the market share analysis as well)."
(FTA)

May change soon with Dell threatening to sell AMD processor based solutions... but I'll believe it when I see it.

Elephant in the room is Dell

[DrSbaitso]

AMD chips were found in 52.5 percent of desktop PCs sold in U.S. retail stores during that period."

Of course, Dell doesn't sell many of its computers in retail stores, it is the largest manufacturer in the US, and it doesn't use AMD chips (yet). So the quoted statistic is misleading at best. Still, more competition is always a good thing.

Re:Elephant in the room is Dell

[Dashcolon]

"Current Analysis' market share numbers measure U.S. retail sales only, and therefore exclude figures from Dell, which uses its Web site to sell directly to consumers. Dell, the top PC vendor in the U.S., exclusively uses Intel's processors in its PCs. (Sales by online retailers were excluded from the market share analysis as well)." That was mentioned in the article

Re:Elephant in the room is Dell

[StewedSquirrel]

Misleading in regards to total market share, yes, but it has a very valud purpose.

To me, it says "where AMD and Intel chips are BOTH sold, AMD is the solid preference amongst consumers"

That's valid no matter how you slice it.

Stewed

Re:Elephant in the room is Dell

[HarvardAce]

To me, it says "where AMD and Intel chips are BOTH sold, AMD is the solid preference amongst consumers"

Please explain how 52.5% is a "solid preference." Nowhere in the article does it mention Intel's share, but assuming it's around 45% (have to remember Apple!), I don't see a 7.5% swing being a "solid preference."

Re:Elephant in the room is Dell

[Jeff DeMaagd]

It doesn't cover mail order, web order or sales to businesses. I bet mail and web order might be leaning toward AMD soon. Sales for business use is a different animal, many businessess are more conservative than makes sense to be.

RETAIL stores

[dostert]

Notice that this statistic is only for RETAIL stores. No mention as to any online sales. I'm not sure... but I'm thinking this article shows AMD in a much better light. I think if you factor in online sales (say ... all of Dells PCs sold) that AMD is still lacking by quite a bit. I'm a big AMD fan. All of my computers have AMD chips, but as a stockholder, I'm getting a bit worried. AMDs big benefit, at least to me, has been the "cool and quiet" benefit. I get a faster processor, using less power, that runs cooler. Now I've been reading that Intel has finally caught on and is basing all of their new processors off the Pentium M ... which I'm pretty sure outdoes the Athlons in terms of power/heat to computing power ratio. I'm excited to see what AMD comes out in terms of power/performance ration after they debut the new socket M2 processors. Should be an interesting couple of years in the chip area.

Re:RETAIL stores

[Guysmiley777]

Yes, but if online sales also include 'naked' chips I'm willing to bet far more AMD processors get sold to homebuilders than Intel...

AMD Shortage

[obender]

I am not sure if this is of interest but in Europe I have noticed a shortage of Opteron dual core chips. So far the only place where I could find some was a couple of shops in Germany and they went out of stock quite fast.

All shops in USA that I could find and that had them did not ship overseas.

Re:AMD Shortage

[Psiren]

Ebuyer (UK) currently have 18 of the 270's in stock. Generally they're good on price, although I don't buy server cpu's so I can't say if its true for these or not.

Re:Way To Jobs!

[Sique]

No. This is the perfect time for Apple to go with Intel. Intel needs to do something to save its ass in the desktop market (even with sliding market share, it's still the big revenue and profit), so they will try to keep Apple happy as long as possible. And if it doesn't work out for Apple with Intel, they can switch to the binary compatible AMD chips at any point.

Apple Inc. sells Apple computers with Apple Mac OS X. Apple doesn't sell Intel Inside computers.

I have always drooled over

[thammoud]

The power Macs but could never justify the price/performance ratio. Now I can ith the new Intel chips. A Unix based OS with a great GUI running on fast Intel hardware makes me a convert. I develop Java apps for a living so it really makes no difference to me as to what OS I use for development.

Re:I have always drooled over

[justsomebody]

The power Macs but could never justify the price/performance ratio. Now I can ith the new Intel chips. A Unix based OS with a great GUI running on fast Intel hardware makes me a convert. I develop Java apps for a living so it really makes no difference to me as to what OS I use for development.

I was already wanting to say you don't know much about fast. But,... then the sad truth pops out. You develop Java applications. Yes, Java is faster on Intel. It can't be optimized for PPC and runs against PPC rules. I even agree, in any CIL or JIT world you're far better with Intel/AMD.

Hardcoded software (one example is Photoshop, where all Mac fans are proud) can be optimized for PPC, and this is why PPC can often leave any PC in the dust.

Re:I have always drooled over

[fdisk3hs]

Two issues related to Java on Macs (may change with the Intel version):
 
You have to get your Java from Apple. I was stuck all through OSX 10.3.x with an old Java. Sun doesn't offer a build. Apple did bundle an update with OSX 10.4.
 
Gosling mentioned in an interview that he builds the latest Java on his Linux box, and copies the JARs over to his PowerBook. Somehow I haven't found enough spare time to try to do that myself. However I think Gosling slowly tried this over a long period of time that he spent using X tunneling to use his native Linux build of Java on the Linux box, and experimenting with the individual JARs one at a time. Theoretically I would think using a slightly older JVM with a newer environment would not be a problem, except for the Cocoa hooks. Maybe you can use X for the gui stuff? As I said, don't know, haven't tried. Lots of people have failed to make this work, though I don't know how much effort they put into it. Netbeans works fine on OSX BTW, there is a release on their site.

Re:I have always drooled over

[level_headed_midwest]

I have been running a Unix-based OS with a great GUI running on fast Intel hardware since about 2003. Mac OS does have some differences with the *nixes (such as Quartz vs. X11), but a Unix-derived OS with a great GUI running on x86 hardware is certainly not Apple's new or unique idea. Other firms even sold their *nix OS only on their own boxes (Sun comes to mind.) What I can say that Apple did differently was to market it to the general public- that had not been done much before.

Re:Point of interest

[Anonymous Coward]

But TFA also says that AMD has >52% of the desktop CPU market.

No, TFA says that AMD had 52% of the retail desktop PC market sold in B&M stores. Online sales are not included in that figure. Overall Intel had 53% of the desktop market that month.

Re:Point of interest

[Jesapoo]

It's a shame they can't compete with Centrino. I'm not really an intel fan, I don't like they way they do business. Indeed, I use AMD in my desktop - but my laptop came with an intel processor, and I love it to bits. They really have nailed down the laptop market.

I can't even tell WTF is going on

[autopr0n]

I stopped paying close attention to the hardware market just after then P4 came out. The last motherboard I bought knowing what model it was an ABIT-KT7-RAID (the raid feature of which, I've never used). I have a 600mhz Duron in there to this day. I bought a 1200mhz duron, and burned it out over clocking the day I got it :P. I got another machine later, with just a generic mobo and an a '2400' chip. I know it wasn't 2400 but it was easy to think of it that way (it was supposed to compare to the P4's boosted clock speed).

But now chips have no identifying performance number at all, and I have no idea what they represent. I also have no idea what sockets or motherboards go with what. It's really quite annoying.

Re:I can't even tell WTF is going on

[Guysmiley777]

You stopped paying attention to the hardware market and now you're annoyed because you don't know what's going on in the hardware market?

If you don't care, don't get annoyed about it. If you want to know but just fell behind, Tom's Hardware, Ars Technica and Anandtech have good articles to get you up to speed. If you're annoyed because you want to know but don't want to do any reading, well, I can't help you there.

Re:I can't even tell WTF is going on

[WuphonsReach]

But now chips have no identifying performance number at all, and I have no idea what they represent. I also have no idea what sockets or motherboards go with what. It's really quite annoying.

The solution to that is motherboard bundles (like what MWave sells) where the vendor puts together compatible CPUs / RAM / MBs in a list you can pick from. Some will even assemble and test the components before shipping them to you.

Once again...

[minginqunt]

*cough*excludingdell*cough*

I love statistics.

The K5?!

[C10H14N2]

AMD has been around since the 60's, whereas I've been using their chips for, oh, 25 years. At the time of the K5, they were just scraping up the pieces from a massive legal battle with Intel, which they won, btw. If you think that was "day one," pfft.

Re:The K5?!

[Twid]

Um, yeah, how about that second source 8088 chip they made in 1982. What a screamer!!! And who remembers the glory days of the Am386 and Am486?!? :)

Re:The K5?!

[Bing Tsher E]

Well, those of us with deep, rich collections of all the weird variants of x86 processors, do. Not the 'glory days' in terms of performance necessarily. But in the '386 era when hardly anybody was running much '386 code at all anyways (due to Microsoft's crippled legacy-based OS software) a non-Intel 386-workalike was a GOOD value. And it's fun to collect all the weird variants.

Regarding AMD Advertising

[gallwapa]

The first time I was in Rome in 2003, we walked out of the Sistine Chapel at the Vatican down to St Peter's Square...I look to my left before entering the massive cathedral and what do I see? Two GINORMOUS AMD banners. I looked over, said "HOLY (crap), rock on!" took a picture (of the square...of course...) and went inside. But it was pretty sweet to see like 100ft AMD logos :o

Re:Regarding AMD Advertising

[cpt_rhetoric]

I know AMD got a lot of good press during the Tour de France. Had AMD logos all over Lance and crew. The team also had a sort of infomercial about their traing techniques where they mentioned AMD.

Re:Regarding AMD Advertising

[Anonymous Coward]

AMD is a sponsor of the Ferrari Formula 1 team, which is, of course, regarded as a national institution in Italy.

Re:Point of interest

[uniqueCondition]

AMD isn't slashing prices
Processors (74% of their revenue) saw units increase 27% y/y and pricing increase 6% y/y. Net/net their margins increased 500 bps to 46%.

Intel on the other hand saw revenues, units and ASPs (average selling prices) fall bellow their mid-quarter guidance (what they tell investors to expect in the future). They're looking for margins to continue to go lower (note: this could imply an upcoming price war!)

Word of Mouth now finally reaching mainstream....

[KeiichiMorisato]

With the gigantic Intel marketing machine dominating the advertising campaign (I don't ever recall an AMD TV commercial...), I'm glad that the "marketing" by users and the community is finally making a difference.

For a while, AMD was only being used by people who would take the time to research what was the best cpu for their purpose (e.g. Gaming...), however the regular, off the street, go to a major retailer computer shopper would just get Intel, because it was THE brand to get or even the only brand they knew.

With the major retailers now pushing AMD machines and with Dell embracing AMD, I'm sure AMD's market share will only increase.

However, hopefully this will knock some sense into Intel and make them get their act together instead of just increasing clock speed, but increasing efficiency. I'm glad they finally decided to step into the 32bit/64bit compatible era.

As a result, hopefully the consumer will benefit with better chips at lower prices AND actual advancement in technology.

Let's see the data

[burrows]

I'd like to see more data on processor market share as well. In particular, I'm curious about the 64 bit market share. Has anyone seen those numbers? I'd like to settle a bet...

MORE RAM

[autopr0n]

The basic benifit to using a 64bit machine is the ability to use more then 4gb of ram.

Re:MORE RAM

[Anonymous Coward]

That's not accurate; using Physical Address Extension (PAE), a 32-bit processor can address more than 4GB, but applications still have a 32-bit address space.

PAE enables you to have more RAM, hence run more applications concurrently. However, the applications won't benefit directly from the extra RAM, but indirectly, since they are less likely to be swapped out.

Re:MORE RAM

[jandrese]

How many people buy $100 processors for boxes they plan to put more than $400 worth of memory in?

I suppose it's good to know that she got an actual 64 bit chip, not that it'll do her much good with the 32bit version of Windows I installed on the machine anyway.

Death of a slogan

[supabeast!]

Does anyone else think that this had something to do with intel finally killing off the "intel inside" campaign?

Yes.

[Anonymous Coward]

Overclockers vacuumed the market for 165 and 170 Opterons. I've had a 165 on order for two weeks, and it's not expected in for another THREE. Also, AMD are hiking the price on these exact models, so you can draw your own conclusions (yeah, they're blinking *SOLD OUT*)

apple

[Blade80]

Steve Jobs: Damn I new we should have used AMD. But I had a gut feeling Intel would be better.

My two cents

[Anonymous Coward]

Just wondering... is Apple switch to intel a way to get rid of their "closed-hardware-solution" ? Are we gonna see MacOSX running on OEM-PC's hardware anytime soon? Are we gonna be in a real OS war, on the same hardware platform?

Whatever happens, I still wont trust DRM!

Roles reversed

[EmagGeek]

I wonder how long it will be before we're talking about AMD as the giant microprocessor monopolist that makes inferior chips...

Even though...

[tetabiate]

AMD is not offering a free (for personal use) optimizing C++/F95 compiler, nearly three hundred computing processors in our lab are either Athlon or Opteron, compared to a few (outdated) Intel-based ones. Intel gross clients are not end-users nor public research centers but private corporations which make big deals with companies like Dell.

- The French criminality index is raising to alarming levels: p2p download explosion last year.

Re:Who says AMDs are cheaper?

[CaptSnuffy]

You can save a lot on your heating bill just by running an intel.

It's all perception...

[Anonymous Coward]

Note that this article got posted with an Intel logo, not an AMD one.

Purchases aren't made just on the strength of a particular product, but on the strength of the brand name. Intel's brand name, for all of AMD's recent and sensational success, is still much, much more valuable and recognizable to the consumer.

I'd compare AMD to Japanese car manufacturers in the early 70's...although their design
capabilities were fairly equivalent to the American-made cars, they only made smaller, budget
cars, because no one outside of Japan would premium prices for the Japanese brand: "What's a Nissan?" But as they delivered solid products again and again, and American manufacturers stumbled (not mention the oil crisis in that decade, which killed off a lot of energy-hungry car designs...sound anything like the power problems current CPUs face?), the Japanese brands achieved parity. These days they can offer a wider variety of products to compete in multiple
markets (sedans, sports cars, SUVs), just like AMD has grown up to start putting chips in
mobile and servers, not just desktop.

Don't count Intel out just yet: Both companies started at about the same time, and Intel's brand name is more recognizable today, in part due to marketing, but in part due to some really good products. Now AMD has finally come into its own and shown that it too can innovate and deliver reasonable products.

Re:Point of interest

[Sebastopol]

When you match that with their server options AMD are wiping the floor with Intel at almost every level.

Retail Desktop - Intel
Server - Intel

Corporate Desktop - Intel
Mobile - Intel

AMD is making headway in retail and server (intel has squat on their roadmap).

However, AMD is making much less on the segments they are competing in. Server is high ASP, but very low volume. Retail desktop is high volume, and razor thin ASP.

AMD needs to focus on being competitive in price to dominate corporate desktop (Intel's fab capacity means they can easily underprice AMD in this arena). Everyone keeps quoting the CPU price for a boxed part, but that is the HIGHEST POSSIBLE PRICE Intel will charge for a CPU. It can be 50-60-70% cheaper per CPU for high volume corporate sales. AMD is fukked in this area because in 30 years, they have still failed to even come close to Intel's volume. AMD hasn't had enough R&D dollars to compete here, but that can change.

And AMD also needs a competitive part in mobile, where the volume is growing every year and ASPs are sky high. This is where Intel is focusing. AMD is years behind Intel in mobile power-miserly processors.

So it is shaping up to be an interesting battle. Lets see if AMD can hang on to their lead this time.

Waiting for the real news

[Nom du Keyboard]

Intel losing market share to AMD is not the news I'm waiting for. What I want to see is:

Dell losing desktop market share to AMD system vendors.

Then something interesting might happen.

They used to call it the "Motorola Curse"...

[javaxman]

then, it happened to IBM.

Now, Apple is shipping with processors from Intel. So, if the curse holds, AMD is going to be mopping the floor with Intel... at least until Apple decides to ship some machines with AMD processors as well...

The other read could be that, well, this story is BS because it's not counting online sales ( Dell, duh ). Although, it's true that AMD has been gaining a little ground, due to having cheap laptop chips on one end and hot, killer 64-bit gaming-rig CPUs on the other end.

Whatever. It's hardly as if Intel is "in trouble", they just aren't growing as fast as they ( or Wall Street ) would like. Let's hold of on talking about Intel's demise for a while... let's wait until Dell starts shipping AMD systems, or has chips running 30% faster than Intel's, before freaking out, OK ?

AMD just kicks Intel's butt, period

[neamerjell]

AMD has been kicking Intel's butt for a long time, mostly because of their price for performance ratio. Intel's pricing has always been higher per performance unit (however you wish to measure it) than AMD. AMD has always been finding ways to boost performance and efficiency in order to stay ahead of Intel: Back when Athlon's first came out, their numbers were to signify that their processor were equivalent to an Intel of a certain speed (The Athlon 1800+ could keep up with an Intel 1.8GHz, but only ran at 1.53GHz) because of their architecture redesign.

Wordwide scam or crappy code?

[MindPrison]

You know what guys? I used to be an Intel fan. I worked at a PC-store assembling and servicing customers computers for years. We dealt in both AMD, Cyrix and Intel at the time. At that time Amd was often prone to overheating and crashes while the Intel processors where very stable and easy to install.

All that changed...

The Athlon came along and my personal "intel-fanboyism" was over. I saw that you could get the same performance for 3 times less money if you did your homework and got it right with the bios settings - the reason I mention Bios settings is...youd be amazed if you knew how many people in this world that cant get it right. Even applying artic-silver (the cooler paste between the cooler and the cpu head) is an artform not to be taken lightly.

Price...

...doesnt matter nearly as much today as performance. And the AMD processors surely can perform. Let me give you a real-life example. Im an animator by trade and as you probably can guess...I need RENDERPOWER...uhmmm..that sweet sweet taste of renderpower and custom optimization. Today I use Linux because Ive compiled the whole kernel and modules for Amd athlon_xp and Ive compiled my favourite opensource 3D program Blender with the same optimizations. Belive it or not - I got over 30 percent better rendertime overall, in some cases even 70 percent! Now thats power that you can feel and taste. Intel didnt even come close.

Another real life example. Im in a community house that houses 40 computers or so mostly Intel computers. I tried my "old" Amd Xp 2500+ up against an Intel 3.2 ghz cpu. (yes..thats a 3.2 ghz cpu versus an 1.8 ghz Amd!) Guess who won? No... this is not a joke... the AMD won by ca. 5 percent. but keep in mind...this was an 1.8 GHZ computer versus an 3.2 ghz. What gives? Ive also tried several other Intel computers with various bus-speeds, ram-size etc. No deal - AMD wins hands down each time.

Now... Im no coder anymore (Used to be way back in Amiga/C64 days) so I dont know squat really, but it amazes me to see this working with 3d all the time that AMD wins the battle again and again.

Maybe theres better code for the processor? Maybe its my Linux optimizations that does the trick? I really dont know - but I am impressed.

Re:Wordwide scam or crappy code?

[jonesy16]

I'd like to agree with you but you've left a lot of details out of your "real life scenarios." You don't mention any attempts at optimizing for the Intel processor for one. I also work in a compute-intensive field where floating point operations dominate. We've got a bank of 10 AMD Opteron 250 machines as well as a slew of Intels (we were always Intel based since our early experience with AMD had left us with overheating / lockup prone boxes).

Performance-wise, our Intel 660 series chips (about $340) perform almost TWICE as well as the Opteron 250's (~$442, prices from pricewatch). The intel code optimized with Intel's compiler, the AMD code optimized with Portland Group. We've tried every optimization we can think of (we even wrote out own compiler). The AMD chips just cannot keep up with Intel for OUR application.

It amazes me that every time an INTEL/AMD story comes up, or a MAC/WIN/LINUX story comes up for that matter, everyone flies to the keyboard to rant about how great their choice is without any consideration for the niche that they're in.

For gaming there is no comparison, the AMD chip wins hands down (though Intel's dothan/yonah architecture is changing that, and let's see an Athlon64 produce only 26 watts of heat disspation). AMD's smaller cache size pays a penalty though when dealing with large data arrays where 3.6 GHz and a 2MB cache smoke AMD.

AMD used to be the king when it came to price, but that's not the story anymore. Call it whatever you want or claim I'm wrong, but check the prices on your favorite crawler. Bottom of the line prices were discussed in an earlier post, and top of the line is no different. In fact AMD's latest offering is more expensive than a P4 Extreme Edition ( Athlon64 FX60 @ $1163 vs 3.46 GHz P4EE 1066 bus @$1060 ). And the server market is no different, the Opterons are not cheap and never have been. Even in the dual core market Intel's lowest offering is cheaper ( Pentium-D 820 @ $244 vs Athlon 64 x2 3800 @ $310 ).

What does it all mean in the end, who's winning, what should you buy? There is no clear-cut answer. Either processor could be 2 or 3 times faster than their competitor and performing the task that you need it for. AMD has come along way in terms of reliability and compatibility since their K5 days, but so have their prices. Their architecture is terribly efficient at many things, more so than the aging P4 NetBurst. The end of 2006 should see those tables balance out again as the Dothan/Yonah architectures make it into mainstream desktop use. So keep your eyes peeled, and your ears open. But watch our for fanboys.

Re:Wordwide scam or crappy code?

[MindPrison]

Quote "Performance-wise, our Intel 660 series chips (about $340) perform almost TWICE as well as the Opteron 250's (~$442, prices from pricewatch). The intel code optimized with Intel's compiler, the AMD code optimized with Portland Group. We've tried every optimization we can think of (we even wrote out own compiler). The AMD chips just cannot keep up with Intel for OUR application.

It amazes me that every time an INTEL/AMD story comes up, or a MAC/WIN/LINUX story comes up for that matter, everyone flies to the keyboard to rant about how great their choice is without any consideration for the niche that they're in.

For gaming there is no comparison, the AMD chip wins hands down (though Intel's dothan/yonah architecture is changing that, and let's see an Athlon64 produce only 26 watts of heat disspation). AMD's smaller cache size pays a penalty though when dealing with large data arrays where 3.6 GHz and a 2MB cache smoke AMD."

Interesting stuff.

Taking into account that you guys wrote your own compiler for the task (which is something I can only admire) Im somewhat puzzled with your conclusion since you mention gaming in comparison as well.

Its not entirely impossible that the Blender code is way different from the code that you guys have (most likely) so it could somewhat use the features of the AMD chip more.

In all fairness - I would like to point out that the comparison I did was to "standard-windows" machines versus my own blend of natively compiled Linux & software (Blender 3D) so from what youre telling me I should experience even better optimizations on an Intel platform? Its of course worth a shot. But the fact remains the same no matter what story comes up to dismiss this and that, the scenes I make actually renders faster on the AMD machines, now - as I said before - this COULD be due to the fact that its different code than yours. Anyhow, Im not a coder - but Im a bit fuzzy on the whole Gaming versus Floating point issue... I seriously thought games these days where CPU/fpu intensive. I know that a lot of the realtime graphics can be done with the graphic cards own "gpu", but still...calculations remains the same.

Re:Wordwide scam or crappy code?

[jonesy16]

Well, it's definitely hard to compare windows and linux when running applications. They interact very differently on the kernel level so there's not always a direct performance comparison depending on what the code is doing. I can say that the AMD's definitely scream when they are asked to do a series of small tasks. For instance, bash scripts in Linux seem to execute infinitely faster on the AMD than the Intel platform (ok, infinitely is a bit of an exaggeration, but let's say at least twice as fast). Code seems to compile faster. Ray tracing, which is mostly an integer based operation is usually dominated by clock speed, however a 64-bit processor can perform two 32-bit operations per cycle (although the 6xx series pentium 4's do have 64 bit extensions).

I don't know ANYTHING about how games are written, I only know that you'd be hard pressed to find a benchmark on any website where an AMD Athlon lost to a Pentium, it just doesn't happen. I can't even describe exactly why our company's code runs so damn efficiently on the Intel platform, but it just does, and no amount of optimization and different compilers can seem to close the gap, which is really too bad since almost half of our computer farm are Opterons. All I ever really try to get across when I start on these stupid rants is to get fanboys/fangirls of either platform to realize that neither company is the perfect solution and neither one even has that much of an advantage. But you know the slashdot crowd, if it's got intel/microsoft/sco associated with it then it's a spawn of satan and if it's got apple/google/linux on its side then it can't lose.

As a side note, one of my buddies at work does a lot of work with Blender and he prefers his SPARC machines, which run at about 400 MHz and can clean up on a 2 GHz P4.

Re:Wordwide scam or crappy code?

[MindPrison]

Cant blame us geeks for being a little biased on our geekhood now can you? ;)

Personally I go the road of "speed and stability" is king. I am definitely guilty of being a fanboy sometimes especially when I get caught up in good results that I can actually use in real life situations.

Most of the times I optimize my systems because of desperation, lack of money if you will - so whats a guy to do? Use the knowledge that I have of course ...and if I dont... I learn it. Belive me - I would not be much of an AMD fan if I discovered that in "my case" ...renderpower...Intel chips would help me speed up my rendering times tremendously - heck...Id change and ditch my AMD in a hoot. Ive got a power-pc based machine too, curiously that one is only 1.3 ghz and renders like a speed-daemon with another render engine called Yafray... but is slow like an geriatric on turtlewax when it comes to Blenders internal rendering engine. On the Intel box ...its exactly the opposite, and again..very curiously yafray is still fast under Linux, but not as fast as the powerpc processor. So again - for me...speed is king. You know - I have to sit here and twiddle my thumbs while rendering so the faster I can get the job done - the more work I get done, so I really cant afford being a fanboy :)

Re:Wordwide scam or crappy code?

[jonesy16]

Haha, well put. Yeah, us geeks always love our gadgets. But I do agree with you about speed and stability being king, and luckily in this day and age, there are a plethora of options which are all stable, and it's just a matter of finding out which is fastest for the stuff you gotta do. Happy rendering!

Re:Consumer vs Corporate? Not True!

[Nom du Keyboard]

Intel sometimes acts like they're a government agency.

Not true at all. If Intel were a government agency, AMD would simply be regulated out of existence by requiring them to provide all the same benefits, unions, and paperwork that government agencies put on themselves.

Which would be followed by the Pure Food & Processor Act requiring them to list the contents of their processor, and would not allow them to claim to be Intel compatible unless they ran exactly as badly as the Intel processor itself does.

You're reading too much into this

[Anonymous Coward]

Don't give too much credit where it isn't due. If you RTFA: Intel itself declared with its Q4 earning statement this week that it couldn't supply enough chipsets to meet demand. In the middle of the year, they turned to ATI to supply chipsets for Intel-manufactured budget motherboards. But ATI was late to deliver, meaning that Intel couldn't ship as many MBs, and as a result, a tone of OEMs cancelled their late Q4 orders. No motherboards means no computer, hence there were less available Intel machines out there to buy. If people couldn't buy an Intel box, they went to the shelf right next to the empty one and bought an AMD one. That being said, I'm sure that not all of the market share loss was due to lack of Intel boxes. AMD has a competitive product, and that some non-trivial percent of the cause.

Re:So

[zodiaccat]

Doritos are overrated. I'm using a Pringles 64 3200+ (SourCream build). It rocks, and only set me back $2.39!

Everyone loves an underdog

[mrrock]

If Intel was the underdog this group would be singing thier praises. Just like at one time everyone loved Microsoft when they were the underdog. Soon we will have Google joining Microsoft and Intel. (probably rightfully so) I am sitting here surrounded by Intel boxes and none of them have their cases open due to heat nor have the poor performance and lack of linux support like the one sitting next to me.

Dell should get used to being excluded...

[Anonymous Coward]

Here's a chart for you...

http://finance.yahoo.com/q/bc?t=1y&s=AMD&l=off&z=l &q=l&c=intc%2Chpq%2Cdell

Funny how HP sells AMD and Dell doesn't. It's keeping Dell out of a lot of accounts lately. Sales guys in Round Rock had to move from Lexus to Toyota when their car leases turned over - they are NOT happy.

Dell will ship AMD in force by end of Q2 this year. Bank on it.

Dumb Intel Commercials

[Anonymous Coward]

[Gasmonso wrote:]I don't think I have ever seen an AMD commercial, whereas Intel was all over the TV.

I always thought those Intel commercials were extremely ridiculous and self-indulgent. In the mid-90's, they had a bunch of guys in space-suit costumes dancing around. How much of middle America understood that those were clean room suits that are used in semiconductor fabs? Why on earth would anybody pick a computer based on those dancing guys?

I don't even understand the more recent commercials with the Blue Man Group.

However, since I RTFA'd, I would point out a caveat with the original FA. AMD beats Intel's marketshare in retail sales; however, mail-order sales, via venues such as Dell, were not included in the result.

Now that Apple has embraced Intel, Dell needs to embrace AMD.

If Intel is So Bad

[Heembo]

If Intel is so Evil, then why did Apple (whom most Slashdotters worship) go with Intel and not AMD?

Re:If Intel is So Bad

[theendlessnow]

If Intel is so Evil, then why did Apple (whom most Slashdotters worship) go with Intel and not AMD?

Apple == Hollywood == MPAA/RIAA == DRM == Intel!

Makes perfect sense to me.

Re:If Intel is So Bad

[Anonymous Coward]

Obviously, Apple is becoming evil. We'll probably have to wait for Steve's son to become a Jedi.

How about better CPU design?

[MtViewGuy]

I think what has made AMD such a competitor is that ever since they rolled out the original Athlon CPU in 1999 with its far more modern CPU core design than its Intel competitors, AMD has demonstrated that you don't need ridiculous clock speeds to get superior performance.

It's more a case of an efficient CPU core, more efficient access to L1 and L2 on-die cache memory, and now on-die memory controller that has given AMD CPU's major performance leaps. For example, my home computer runs an AMD Thunderbird-core Athlon CPU running at 1.4 GHz; it will handily beat out the final generation of Intel Pentium III's running at the same clock speed, and in fact even compares very well with the early generations of the Intel Pentium 4 CPU's.

desktop vs. diy?

[alexq]

"Sales of Intel-based desktop PCs fell 22.3 percent during the fourth quarter, according to Current Analysis. As a result, sales of AMD-based desktops took the lead during the pivotal fourth-quarter holiday shopping season. AMD chips were found in 52.5 percent of desktop PCs sold in U.S. retail stores during that period."

This appears to apply to premade PCs only. I wonder how the statistics differ when you look at the raw number of processors purchased - I imagine that that market could be pretty different.

Isn't it obvious...

[corngrower]

Apple is even more evil than intel.

Just wait...

[ImaLamer]

Soon Intel will have the entire Apple PC/Laptop marketshare!

Distribution is the key

[snuf23]

I think the fact that AMD chip based computers are now being sold by all the leading PC manufacturers except (for the time being) Dell probably has helped AMD significantly.
Due to the antitrust lawsuits, Intel is not going to be able to beat up on companies that want to source chips from AMD.
The other point is that the brand "Pentium" which was once upon a time what people asked for when they went to purchase a computer, is not the force it used to be. For most people now the price and the features (can it burn dvds? use the Internet? can I plug my camera into it) are what people are looking for when shopping. They don't really care if it's a Pentim or an AMD.
But the vast majority of non-techies do not know who AMD is or that they might have an Athlon in their system. And they don't care.

Re:Point of interest

[corngrower]

Three things here. First, yes Intel sells more server chips than AMD, but AMD is making strong inroads in this area. In fact, AMD has been limited by their lack of fabrication capacity. The margins in this arena are fairly high, and with the new Dresden Fab coning on line this year, Intel is going to take a big hit. If I'm AMD, and have limited production capacity, I'mg going to be producing chips that give me the biggest margin, and that's going to be the high end chips. Because of their superior performance, AMDs chips will command a price premium to Intels chips.

Secondly, corporate desktops. The best that AMD can do is to try to underprice Intel, which will be difficult since Intel does have better process technology. Expect prices of the midline chips to fall as Intel lowers prices to maintain market share. With margins as thin as they are in this arena, AMD needs to work to maintain its performance edge on the high end chips where it can command better margins.

In laptop processors, the Pentium-M's excellent perfomance/power ratio means that AMD is not about to overtake Intel's number one position. AMD's Sempron may have better performance, but it also 25% (AFAIK) more power hungry. This is an important market segment, and while AMD puts up some competition, Intel is still the strongest. The price margins in the market aren't as large as those of the server market, but they're still better than the margins desktop market.

It's Intel's more advanced process technology that gives them the edge in producing the low power laptop chips, not the manufacturing volume. I wouldn't say that AMD is years behind Intel, just 10 months behind, which is far enough behind to be at a definite disadvantage. AMD should be concerned with improving its process technology while also trying to improve production capacity.

Re: Good quality and low prices work :)

[dakirw]

I have always thought that Intel has a high yield, quality process. However, their prices have always been higher than AMD, Cyrix, and any of the other competitors over the years. If their yield was good, then there was no reason for them to be higher priced. I always felt that they were gouging the customers, so I quit designing them into my systems. Since they are still higher priced than AMD for competing parts, I wonder if their yield has always been not so good.

Their yields have been good, but I suspect that Intel is just following basic marketing and sales. Why sell for cheaper if you don't need to? If the customer is willing to buy an inferior product at a similar or higher price, Intel can get away with this for a short period of time, much as IBM did with its PC division in the late 80's. However, sooner or later customers will realize this and buy more competitive products.

Fortunately for Intel, they have a lot of money and can spend to overcome these issues by having greater production capacity and lots of marketing.

Re:So

[pianomahnn]

All I have to add is that I'm laughing at this. On a conference call...thanks for the mute!

I love AMD

[Anonymous Coward]

I'm not a rich guy and I don't live in a rich country. Before I buy a new PC, I look for info and benchmarks for weeks! In cost vs speed, AMD processors have beaten Intel's (I mean desktop PCs, I know nothing about laptops).
  Few days ago I finally got a new PC with 3000+ (Venice core) Athlon 64. It's awsome! After burning Prime95 for half an hour, temp is stuck at 45C with box cooler, and that's with +450 MHz overclock. AFAIK AMD dominates in this price range with their Athlon 64s (3000+ to 4200+). They run very cool (literally), from what I've read you can give your processor +50% overclock and still run rock stable, if you have good enough motherboard and RAM.

  I hope AMD continues to compete with Intel and stays in business for many years to come. It's true that Intel is much larger and works in many areas aside from desktop PC processors, but AMD can focus in this one area.

Re:I have always drooled over(kernel performance?)

[Steavis]

I also develop java apps, many of which run on XServes. However, in our case, the OS made a HUGE difference. So much, in fact, that we switched (wherever possible) the XServes to Linux.

Does anyone know if the x86 builds have addressed the thread creation problems mentioned here? It made a huge difference for our CPU/Thread bound java apps, even in simple benchmarks. I'd be curious to see someone run benchmarks on the new x86 Apple machines running OSX86 vs some version of Linux (if they can get it installed).

For example, in our case (identical Dual G4 1.42GHz xserves):

[Linux]
SciMark 2.0a

Composite Score: 182.71861801546197
FFT (1024): 147.06919818652634
SOR (100x100): 221.18400271177688
Monte Carlo : 17.996476897927952
Sparse matmult (N=1000, nz=5000): 162.3786017210325
LU (100x100): 364.96481056004615
java.vendor: IBM Corporation
java.version: 1.4.2
os.arch: ppc
os.name: Linux
os.version: 2.6.8-1.ydl.7smp

[OS X 10.4.2]
SciMark 2.0a

Composite Score: 102.68623228630156
FFT (1024): 64.50593864434353
SOR (100x100): 211.84642346103976
Monte Carlo : 11.988007039210785
Sparse matmult (N=1000, nz=5000): 103.01163174770112
LU (100x100): 122.07916053921254
java.vendor: Apple Computer, Inc.
java.version: 1.4.2_07
os.arch: ppc
os.name: Mac OS X
os.version: 10.4.2

Simpleton benchmarks, yes, but they reflected the real-world performance problems with surprising accuaracy.

This is not news! Its just NORMAL!

[I.M.O.G.]

This article from MAY OF 2005, shows why this is just spin - anyone who has been paying attention to AMD in the retail outlet sector should know that AMD has done well in this area for quite some time! The exact figure from last May was... You guessed it... 52%!

Re:Point of interest

[Anonymous Coward]

I think people are comparing whatever they want to get the results they want. Is there any real study about the performance vs cost for AMD and Intel processors when it comes to common users? (when I say common users I mean people with less than 4GB of RAM, $1k worth graphics adaptor, etc)
Recently I've come to believe that the prices of AMD chips are not as competitive as they used to be. True they are a little cheaper and cooler but the difference is not all that much anymore. Also I've been hearing people complain a lot about the quality of the motherboards for AMD chips.
I personally don't care if it is made by Intel, AMD or an ape with a very small hammer. I want the best for the $$$ I'll spend on cpu, mobo and ram.

Because of the mobile chips

[nacs]

I think one of the biggest reasons is because AMD's laptop processor lineup sucks hard and Apple really needed to deliver with their new Powerbooks / "Macbook".

I'm sure Intel's DRM technology and production capability also played a factor.

Re:Because of the mobile chips

[tomstdenis]

Laptop processors suck hard? Get yourself a Turion MT series. They run as fast as 2.2Ghz [though 2Ghz is more common to find] and are rated for 25W TDP which is basically what a PentiumM is.

Oh and the Turions are 64-bit processors. So you can run in a proper long mode.

That said, my last laptop was an Athlon-XP-M and even that was not so bad. It had a TDP of 45W but I could easily eek 4 hours out of it when editing source code. It usually ran 1150mAh in the lowest mode out of a 4.3Ah battery.

My dell with a Dothan 2Ghz runs about 950mAh out of a 7Ah battery. So really the Intel laptop gets a longer run not really because it takes less power but because the battery is more efficient [and physically larger].

The turions are even more efficient than the XP cores. I don't have the current rating yet [hard to figure that out in Windows] but I suspect that it's comparable. I get 6.5 hours with the spare battery in [and this lifebook is still very light]...

Anyways, the reason they went with Intel is because they want some kichy design that the zealots can get up all in a roar over. Both Intel and AMD execute the same ISA so from a users stand point there is no different. A MacOS binary for Intel would work on an AMD cpu.

It does make a difference for the OS/packaging folk but that just means they have to "earn their pay".

Apple would be wiser to not deal exclusively. If Intel is the better choice as they claim let the market determin that.

Tom

Re:Who says AMDs are cheaper?

[Afrosheen]

Not really, because the electricity bill offsets any savings you might have gleaned.

Re:Who says AMDs are cheaper?

[JanneM]

You can save a lot on your heating bill just by running an intel.

You would only save in total if your electricity is cheaper than your heating system per unit of heat. Most likely it's the other way around, and you lose money. It's the same tired argument in favour of lightbulbs, saying that their inefficiency doesn't mappter.

Re:Point of interest

[default luser]

I believe you meant to compare AMD's TURION to the Pentium M.

While the Turion MT was a good match for the Pentium M, the Turion ML was far more common, and used more power.

It seems the real thing holding the Turion back is that Via and ATI have not built power-optimized chipsets for it yet. However, I expect this to happen in 2006. The chipset makers know how critical power consumption is to compete with Intel, plus DDR2 will reduce the power required for memory by a lot (much lower voltage).

The only question is: will Turion dual-core be able to compete with Yonah? AMD has a lot of work to do if they want to be taken seriously in the mobile world. More along those lines: how is AMD going to sell themselves on the desktop once Intel catches up on the power consumption and performance front later this year with Conroe?

Gonna be an interesting year for processors.

Leap ahead? Whatever.....

[blankoboy]

...AMD is already leaps beyond

/AMD fanboy.
http://www.leapsbeyond.com/.

Re:Point of interest

[froschmann]

If you can drop the process size, you up your production capacity. Smaller chips, right?

Re:Point of interest

[Bing Tsher E]

I never thought I would have to say pricewatch.com is behind the times. If you wait, within any two week period, you can buy a motherboard bundled with a Sempron 2600 or 2800, from Frys for ~$79.00. I'm not sure anymore why I would pay any attention to pricewatch, at least regarding low-end low cost hardware. Of course, you have to have a local Frys, which we do here now.

Re:Point of interest

[kesuki]

toledo is still on 90 nm tecnology, power consumption will go down a lot, and clockspeed will go up when amd makes a 65 nm based x2.

so yeah, conroe kicks butt over the existing x2's because the product amd will be selling that is comperable to it hasn't been released yet ;)

Me and my bias

[billcopc]

I'm a pretty big AMD fan, mostly because I'm sick of setting up Intel's stupid huge clunky heatsinks that arch the motherboard like a hunchback. It just looks retarded and I'm sure it's not exactly healthy for the finely layered circuitry lurking beneath. It's also quite charming to have a screaming dual-core AMD with an absolutely inaudible fan, and an NVidia-designed chipset that's got almost everything right.

In contrast, every late Intel board I've loaded will just plain bluescreen during Windows Setup if you don't load the floppy for its SATA controller, whether you use it or not. And that stupid round fan with no sides that loves to chew the power cables. Puh-leeze!

And who can forget the nightmare of mating chipsets to CPUs ? Sometimes I wonder if the processing power is really in the motherboard, and the actual CPU is just a key to unlock various features.. it's such a mindfuck. My AMD board from 2004 took the new dual-core chip without fussing, and I can drop just about any old Socket-A chip on an aging board.. sure, it might not always run it full speed but at least it runs. Try sticking a Prescott chip on an ancient 478 board and see how far you get.

Long story short: Intel may be the household name, but they've reached a point where they make too many "mistakes", intentional or not. It's alienating customers and techies alike.

VIIV - Marketing LXIV, shipping XXXII

[vincecate]

As reported in SlashDot almost a year ago, Intel's idea behind VIIV was that VI is roman numerals for 6 and IV is for 4, so it is sort of saying 64-bit, though the correct way to write 64 is LXIV. Intel's web site description of VIIV says,

Processor
Dual-core 64-bit processing power means smooth performance and complete control of your digital media.

Intel launched VIIV at CES with the "Centrino Duo" platform that uses their "Core Duo" processor (also known as Yohan). But as with all Intel notebook chips, Yohan is 32-bit only (look for "1" footnotes). So the VIIV that is really shipping is 32-bits. And the correct way to write 32 in roman numerals is XXXII.

Re:Point of interest

[Sebastopol]

Go one step further, Compare Turion to Yonah, forget pentium M.

You're right: this IS going to be an exciting year.

Finally

[Green Llama]

I've been predicting this for years - good to see it finally happened. More powerful chips for less $$ should sell more.

Re:Who says AMDs are cheaper?

[KermitJunior]

You assume he's running it only to heat. The actual factor would be the difference of use of electricity between the intel and AMD compared to price per heat and difference of heat of the two, etc., blah blah blah.