AMD issues dramatic price cuts for triple-core CPUs

Meh, those "very low-end" chips just don't interest us elitists here in this forum.

If they were selling it at a loss, or below manu cost, they couldn't sell it legally. On the other hand, now we see exactly how cheap AMD is able to make the 65nm X2's for.

Sounds good to build some low end systems.

I wounder what the production cost of the Phenom X4s are because to already have a chip thats barelyt a year old and being able to cut the price this low?

Especially on a monolithinc die and new chip. Just seems weird to me really.

I mean look at the Q6600. it took it about 1 year to drop to half price then about 6 more months to reach the sub $300 range.

As I said it just seems weird that they can cut the price of the Phenom s so low so fast.

Also, keep in mind that Intel doesn't have to do a price war. They historically charged quite a bit for their CPU's because they were always ahead in the market (even when they were down in performance, as during Netburst).

I don't think that B3 yields are all that bad. We are seeing some tweaking in triple and quad clock speeds, unlocked multipliers and thermals, but we are not seeing a massive dump of triple cores on the market with a dearth of quads.

Intel's quads aren't native. Until this point, if one core needs disabling, they'd just release it as an old fashioned budget single core Celeron. There's no reason to package a triple core C2D.

When Nehalem arrives, they could disable one core and release a triple if they have yield issues, or if they want to show up AMD in that market niche. Intel's said they're considering triple cores.

At any rate, over the next few years, I expect to see either triple or quads replace dual cores, with six and eight cores at the mainstream and enthusiast end. Software should catch up soon enough.

Look here or on Tom's hardware news.
http://www.tomshardware.com/news/a [...] ,6364.html

A 3.0 Ghz AM2+ Deneb (Phenom X4 TBD) sounds like a nice upgrade for me in Jan 2009.

Not against the 4870x2 and the coming 4850x2. My point wasnt going in that direction anyways. Allowing nVidia to lower prices as ATI releases new cards, like their (nVidia) entire line, or the 96GSO against 4670, if thats ok, then why not this? Sure high end is important, but actually the higher you go, the lessor the returns, both on cpus and gpus, as their potential buyers drop in those segments. I see someone allowing nVidias current strategy but shaming AMDs at the same time as unfair, illogical and biased. I like fast gpus, and I like fast cpus, but if you argue that a GSO is a decent buy at much lower pricing that it was, so it could compete with one of the lowest cards ATI produces, then that same philosophy should be used with AMD and its cpus vs Intels.

No, as ATI has the high end, and just released on the low end, so to has Intel done the same, and people scoff at AMD, yet say good going nVidia. The 4670s only response in that market was to drive down one area or segment ATI lacked in, and nVidia responded, as does AMD. Does that change my disappointment towards AMD concerning the high end? No. Does it make sense for both AMD and nVidia to compete at the low end, even tho both have products that are being reduced in price to compete? Yes. Tho its not good, it makes sense. Does it deserve slamming? Cause if it does, then its easy to point a few things out about peoples bias vs real business parctices.Like Ive said before, Im in the middle, and see a few things differently than some. I think this is an easy conclusion to draw, if you followed what Id written. And Im discounting your point at all, like I said, Im disappointed with AMD highend, but you havnt addressed my point at all.

Thanx for the downgrade whoever it was, heheh, I must be in the middle, one side or the other doesnt like my comments heheh

As I pointed out, ATI released the 4670, a superior card to what nVidia had at that price point. What did nVidia end up doing? They lowered the card that was there, then lowered a HIGHER card that cost more to compete at that level. Sound familiar? Im in no way ignoring, missing, discounting or forgetting the AMD doesnt have a presence at the high end. My point is, people laugh, slam jab and whatever AMD for doing what they do, which is exactly what nVidia is doing. If one person defends nVidias position on what theyve done with the GSO and the 9500 which just came out (again, sound familiar?) then turn around and run AMD into the ground for doing the same thing AT THAT PRICE POINT, shows a bit of a non logical response, attribute it to whatever you wish. Everything doesnt start from the top down. Lets say when Deneb arrives. It may not compete at the top, but say 3 levels higher, from that price point on down, there is still competition. Get it? So as each company adjusts their pricing accordingly, from that price point on down is how theyll compete. Tell me this then. Since P4 EE cpus didnt compete with the FX series, was it seen as amart or ridiculous as having those EE cpus at 1000$ ? Surely those EE's didnt compete, yet Intel held their pricing. Granted, the gap wasnt as far apart then as it is today, but did it matter? Its not as tho IM missing the point, its more Im looking at the larger picture here, and trying to point out certain things.

If it ocs as good, if it has performance gains, and if its a decent price, theyll sell. Lotsa ifs. But making them competitive will help. Question. Does anyone think Intel will ever sell a tri or a hacked dual? Or would they just throw them all out and take the loss? Im thinking that IF (that word again) AMD ever gets a niche market for these chips, that Intel might do it

You have seen there financial statements, right?

My "triple cripple" is out for delivery today, but as it's 7 pm already, I probably won't get it till tomorrow. I'll benchmark the old setup on Saturday and then set up the Gigabyte GA-MA78GM-S2H board and 8750 at stock, then benchmark that.

Then, I'll see if I can get the fabled 2.8 on air that review sites got last April. Edited to add, that the overclock might not be worth it, see Extreme Tech's review last June with the same board I'll be using:

http://www.extremetech.com/article [...] 186,00.asp

Other sites had better results in April but they may not have kept it at 2.7 or 2.8 for long. At any rate, I got this CPU to compare an Athlon X2 and a Phenom X3 at the same clock. 2.4 is the minimum I generally accept for CPU's nowadays.

I tired of most single player games, I'll benchmark Oblivion and The Witcher but haven't played them in ages. Same for KOTOR (and I haven't gotten Mass Effect because of DRM issues). Spore's on my kid's PC, won't be playing unless I play his copy (more DRM issues there).

What I've been playing 20+ hours a week is LOTRO. My wife and I are lifetime members and can't wait for the expansion -- adventuring together is kind of romantic in an odd sort of way. Plus we have other real life friends who play.

So, the triple core will be useful for my other hobby, which takes a bit of time too; unlicensed anime fansubs in H264 when available. My goal is to get a 24" LCD and a Blu-ray player for my PC, so I can watch more than just my DVD movie collection. I definitely want to buy anime and movies in Blu-ray rather than DVD this holiday season.

Besides, the notion that most games are single threaded is kind of obsolete. Many use 2 cores and some use 3 or 4. When I do take the time to play newer games, I'll take both the number of cores they use and whether they're native DX10 into account (as long as they're CRPGs).

At 17% faster per core than an X2, I expect to get some mileage from the 8750 until it ends up in a dedicated HTPC when I go Deneb. Another X2, even a 3.2 is kind of beside the point right now. .

The motherboard I have does support a couple of 125 watt CPU's but the whole thing will be in my box for only about 6 months before it does HTPC duty back at stock.

Do you understand their financial statements? They're paying for ATI, which was overvalued to begin with. Those paper losses have little do to with the manufacturing cost of their CPU's and the price points they sell them at. If AMD had not bought ATI, they'd be profitable now. If they'd paid the right price, it would already be over as far as charges against earnings.

TC's up to his old tricks again ... just bashing AMD.

Getting monotonous ... havn't you got anything else to do?

I pray your Opty blows up ... it would be Karmic If you ask me.

We wouldn't have to read your regurgitated cr@p ... over and over again.

If AMD has dramatically reduced prices then great ... someone might get a cheaper entry level system or a good deal for a HTPC rig.

Gamers will all yawn though ... nothing here for them.

that's about it for this story ... nothing else to tell ... no point in the fanbois gathering like moths to a flame.

Sorry, damn facts...

BTW: I recommended an AMD dual-core to him, if you were taking notes!

Be reasonable TC. You can quote their price two years ago and one year ago, but find another quote instead of today. My 401K lost $385 in one day and gained a hundred back yesterday. The bears are partying like it's 1929.

I don't own AMD stock, don't even own stock in my own company, I went for one third moderate risk and two thirds conservative (the moderate risk lost). The stock market is like Vegas nowadays and analysts who understand the tech less than my A+ and hobbyist level issue opinions on Phenom's, Larrabee etc. that influences stock prices.

AMD's lost market share and is bleeding cash, but less than before and they have some clear wins that even analysts can see (ATI's achieved a coup that's probably influenced layoffs at Nvidia). If their CPU division can improve Deneb made with SOI over B3 the way B3 improved B2, then we'll see. The second version of Deneb with some version of the high k process should be even better. Then, they'll be set to introduce a truly competitive architecture.

Don't count them out yet and don't consider stock prices to be the be all and end all. Stocks rise and fall, and it's the profit or loss that matters when you decide to cash out.

AMD's min/max in:

2000: 13.56/48.50
2001: 7.69/34.65
2002: 3.10/20.6
2003: 4.78/18.5
2004: 10.76/24.95
2005: 14.08/31.84
2006: 16.9/42.7
2007: 7.26/20.63

The minimum this year was on Aug 1st at 4.05 IIRC. Today it closed at 5.42. That's a 33% increase in just over a month and a half. Don't you wish you had bought then?

I did. Sucks to be you.

Truth is, AMD's stock has been upgraded by many major analysts lately (it's actually rated better than Intel on average), and they are expected to be back in black very soon, after 7 straight quarters of losses. So, AMD's stock prices will likely rise further. Then again, as JDJ says, stocks are mostly speculation, so who knows what may happen.

Yeah, that I have. And I've seen my monthly and quarterly financials too. Debt can do a lot to make things look bad. *student loans are a gripe for me for example*

Otherwise, what Yomamafor1 said about operating cost.

Yeah for the most part it is. Same with oil and commodity prices. Mostly based off of what speculators think the supply of oil will be in 10 years for example, same for any consumable item. But with stocks in companies, it has a lot to do with projected income and speculated product success and sales. One of the major factors for killing AMD stock was the failure to execute on time with the k10, almost 2 years late to be exact. Then the whole TLB thing killing server market share, and bad press from that. Now that they have ATi executing, with successful products at that. It's keeping them for going lower, now as long as Deneb/Shanghai comes out on time it should help them a bit more, regardless of whether nehalem is higher performance or not.

Now lets not get into people intentionally trying to drive stock prices down. Thankfully, if you watch the news the government has at least tried to do something about that now.

I also wish I bought Googles stock when it came out for $15/ share and then sold it when it was in the high $600s. But guess what? You cannot predict how a stock will go without insider information.

TBH those little bumbs are nice but you have to look at the reality. If AMD consitently is cutting prices and is selling below operational costs and continues to have losses their stock will never see the leap back to $40/ share. The main reason it went to $40+ a share was due to them actually taking market share from Intel very fast and selling almost all of their stock and of course not being able to keep up with the demands.

Right now you could buy AMD and in 2-3 months it could go up or it could go down. Or you could buy Intel and get the dividends. Thats all the choice of the buyer though.

^Actually I doubt they will. The 22nm chips they have are not even working silicon from what I have heard. Intel currently has 32nm working chips and is starting to mature the process for the Q4 2009 release.

I am sure that during the next coming months or maybe early to mid next year Intel will announce it. Intel announced the working SRAM 32nm chips back in about November of 2007 and they arent even releasing chips on it until 2009.

I also look at it that AMD tends to start or becomes part of a consortium of companies that research new processes and technologies, mainly because AMD could not afford to do it on their own. Not that thats a bad thing but then I look at Intel who has been doing it on their own for many years and has always had the most advanced process, not only size wise either. Intels process is so efficient that even when its new their yeilds are much better than other companies.

But either way whoever gets the 22nm CPU out first it still benefits us end users because it will use less power and produce less heat.

Then again though Intel will be ahead in the HK/MG tech that IBM has yet to get out and that will be a big factor in the heat/power of the chip.

Intels stock is still a pretty strong buy. You get dividends. In fact though it should be much higher than it is now. They have posted, for two quartes in a row, record breaking profits but since they fell short of analysts expectations, their stok dropped down instead of going up.

But thats the stock market for ya. Everything is based on the analysts idea of what they should post and if they do better the stock goes up, if they do worse it goes down even if they break their own internal estimat which is always more accurate than a analysts.

From what I have seen with the "anti trust" cases against Intel AMD doesn't see a dime of any of the fines, or has yet to. So even if Intel is found guilty, although the recent one in New York is obvious guys going for office and the FAB that could bring jobs there, AMD may never see a dime of any fines or fees. It will go to the government who will of course waste it on needless spending.

I am kind of off of predicting Larrabees performance. All I can see is that Intel is trying to stir things up. Maybe for the better because as it is now you have nVidia in the high end for the insane price and ATI in the low-mid-high range at the decent price. But we have been on the same standard of GPUs for a while with no change for the past 10 years or so.

I can't say Shanghai will compete with Nehalem. Only because I have yet to see Nehalems performance in a server. But don't forget server Nehalems will have quad channel DDR3 and also 4x Quickpath as compared to desktop chips which will have 1x Quickpath.

To compare, the first memory bandwidth Nehalem test I saw was on a 1x Quickpath 4.8GT tri channel DDR3 @ 1066 and it reached 18.8GB/s. So adding quad channel DDR3 and 4x Quick path? That makes Nehalems server chips seem like complete power beasts that might make a change in the leader of servers. Of course that depends on how well Nehalem scales too.

At any rate, if AMD does not get in the black soon they wont be doing well enough to see frutition of a lot of their ideas. I hope they start to at least break even.

That sounds quite right, generally intel still get better market then AMD, this price drop is just their strategy, it is said that AMD is going to bankrupt, I don't think so though they are not that good at the moment.

Nope.

Intel uses MCM....Multi Core Module. An Intel quad uses 2 dual core dies sharing the same package to make a quad. AMD uses a native quad, meaning all cores are on the same die. If an intel die is bad, they only need replace it with another. AMDs native quad cores dont have that luxury...either all the cores work or they do not.

All things are relative. Phenoms arent really crap overclockers. The K10 B2 steppings had several problems and were for the most part clock limited. Not all were limited, but most were. Initial reports on the the new steping, B3, show it to be the improvment over the B2 stepping that AMD intended it to be, with reports of up to 3.4GHz stable. At this point, as the B3 stepping is less than 2 quarters old, and given some of the sources, the veracity of the claims of B3s improvment is subject to question, but on the other hand, there have also been positive reports from reputable sources, so overall B3 looks good.

Relative to C2D, Phenom cant acheive the same percentage increase, but relative to itself, its really not crap.

As far as AMD caring, I dont see that they would. The overclocking comunity is a tiny fragment of the market. If every overclocker alive were to never buy another CPU, I doubt either AMD or Intel would even notice a blip on their sales reports

'Overclock-ability' may sell to individuals who frequent forums such as this, but it doesnt push systems out of Walmart, Dell or HP, and it certainly doesnt put workstations out in enterprise...where the real money is.