Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
90nm power consumption numbers from Anand
Wow. The 2.2Ghz/512k/2-DDR 90nm A64 consumes significantly less power than 1.8Ghz/512k/2-DDR 130nm A64. I'm impressed. Can't wait to see what AMD does with 90nm in the mobile space, the last stronghold of Intel.
Things are sure looking up for AMD. I think I'll be holding my shares quite a while longer.
Especially interesting is gaming performance for the top games (HL2, D3, Far Cry, UT), power consumption numbers for 90nm etc.
Yes amazing gaming benchmarks. It looks like A64 is the top choice for all the prime gaming engines out there. The games benchmarked in this article will represent nearly all other 3d action games for the next couple of years. The Doom, Unreal, and Source engines are highly sought after by other developers and will be used in 90% of the games coming out, and probably 100% of the top games.
Also the 90nm power numbers are even more interesting given this quote:
I'm a little disappointed in our power consumption testing because we ran into a big, hairy snag. This problem affected the previous Athlon 64 90nm tests we performed, although we were unaware of it at the time. It seems the CPU voltage option on our Asus A8V Deluxe motherboard isn't actually functional; we could set the proper voltage for each CPU in the BIOS, but no matter what, both the AsusProbe monitoring software and CPU-Z reported 1.6V for the CPU. It's possible this problem was caused by the revision 1008 beta 1 BIOS we were using, but that BIOS was necessary for compatibility with the 90nm Athlon 64. This problem was exacerbated by the fact that the A8V Deluxe is the only motherboard we had on hand that would POST with a 90nm Athlon 64 processor. As a result, all of the readings you'll see below were taken with the CPU voltage at 1.6V (although we did set the proper value on the BIOS for each processor). Generally, 130nm Athlon 64s are supposed to run at 1.5V, and the 90nm flavors expect 1.4V. Take the results as you will, or ignore them if they offend your precise sensibilities.
So the 90nm system power of 108/145 is too high, and it still looks quite good compared to the equivalent 121/172 for 130nm.
today upholds the PC performance crown with the introduction of the AMD Athlon(TM) 64 FX-55 processor, the world's ultimate PC processor for enthusiasts and hard-core gamers. The AMD Athlon 64 processor 4000+, also announced today,
Well we all knew it was coming, but I just have to say, dang, AMD looks to be in a sweet position with its processors for the foreseeable future.
I think the leadership AMD has been showing in its product offerings will translate to increased profits over the next 6-12 months. The tsunami has just not hit yet, but it is out there looming deep, ready to spring, and when it does, we should get a nice revaluation upwards. Q4 may be the first indication to outsiders and analysts of this coming wave.
I remember saying "I could do it, I just don't WANT to," too.
Way back in gradeschool.
lol.
In Intel's case here, it is probably true that they could do it, but it would mean a 9-12 month delay in product release to do another revision of the chip they hadn't planned upfront. In a year 4Ghz isn't going to cut it on the top end, so they probably decided not to bother. Of course they probably realized the current chip revision's shortcomings 3 or 4 months ago, so they could have already been well into another revision by now to fix the problems. They really dropped the ball here, and I think the impact on AMD is going to be very significant. AMD should be able to grow their high-end marketshare very quickly which is where all the $$$ is at.
There were no technical or thermal limitations that prevented Intel from releasing a 4-GHz product, Kirby says. But there were practical limitations, so that to release such a product Intel would have to devote time and energy to tweaking circuit designs and testing those chips, he says.
It sure sounds like they hit a wall on the current revision of Prescott of around 3.8Ghz, and the only way over it was to make some circuit changes to the chip, which would require another chip revision. This means design work, tapeout, validation, etc., which is typically a 9 month turnaround to a product in the most aggressive situation.
The statement Intel makes would be the same as if AMD said, "There are no technical or thermal limitations that prevent us from releasing a 3.0Ghz chip". It doesn't mean they have a current revision of a design that will run at that speed.
I'm really surprised by Intel's move on this. I would think they would pursue all avenues to ensure they hold the top end performance crown, but it appears they are conceding it temporarily (a year or so) to AMD.
Me: This is why I am looking for at least $.15/share today on flat revenue.
You: The reason I think the earnings growth will be more modest than that
Looks like you were right, but I will say I would have been dead on if marketing expenses had not jumped $20 million. I wasn't expecting that much of an increase. That was $.04 right there.
It is pretty clear to me if AMD can grow revenue just a little more, these nitpick details to the earnings report will be less important to overall profit. At that point AMD will be making $.30 or $.40/quarter.
I think Q4 will be quite good, assuming flash can hold the line.
Overall I would say the report was quite good. I like that gross margin was up significantly (Revenue - Cost of Sales). The increases in R&D and MG&A, however, drove down operating income to be lower than Q2. I hope that they don't start spending like crazy capping net income going forward.
Every dollar of CPU sales that offsets a dollar of flash sales represents incremental profit, because flash profits have a minority interest component that does not appear at the top line.
Yeap. CPU group gross margin is running close to 60%, and Flash is more like 30% because of the minority interest. So there could be a huge swing in profit depending on the magnitude of the revenue swap between divisions. This is why I am looking for at least $.15/share today on flat revenue.
Every $100M flash is down reduces profit by about $30M, but every $100M increase in CPU group revenue increases profit by about $60M, netting $30M. I expect the revenue swap this quarter will be roughly $100M.
Try the press release:
http://biz.yahoo.com/bw/041004/46257_1.html
SUNNYVALE, Calif.--(BUSINESS WIRE)--Oct. 4, 2004--AMD (NYSE:AMD - News) today announced that net income for the third quarter ended September 26, 2004 is anticipated to increase from the $32 million reported in the second quarter of 2004 despite slightly lower sales.
Slightly lower sales in this sentence implies lower than Q2. It doesn't say lower than projected here. A paragraph later, Rivet does say lower than projected, but lower than projected can also mean lower than Q2.
I'll admit there is a small hole in this, but I think you are reading way too much into it. My guess is $1250M in total revenue, and $.15/share earnings, which to me would not be a disappointment.
Yes, it is allowable to take the extra day into the last quarter
This would imply quarters sometimes end in the middle of the week (like on a Wednesday), which I don't think is true. That is why I suggested every once in a while a quarter will have an extra week. I'm not saying I'm right, I just haven't seen evidence that points to quarters ending in the middle of the week.
There is no need for a 14 week quarter unless you think the one spare day a regular year or 2 spare days every 4th year is significant. I'd still call it a 13 week quarter even if it was 92 days instead of 91.
Well look at this year. Q4 ends on December 26th, and there are still 5 days left in the year. It adds up, 1 day a year as you point out. They obviously realign this at some point or the quarters wouldn't fall nicely the way they do now. My question is, when is the leap quarter that has an extra week? The quarters always end on the same weekday, so it has to be a whole week that gets added.
Nice try. There's a lot of "funny" math there. It just isn't equivalent to running something at 6GHz. Sorry.
True. FX-53 performance per Ghz scales better than Prescott performance per Ghz does. That is, running a FX-53 at 50% faster clock speed will result in a higher performance increase than running a Prescott 50% faster clock. So a 3.6Ghz FX-53 may perform more like a 6.5Ghz or 7Ghz Prescott. I can't even imagine the current demands of a 7Ghz Prescott. I doubt there is a motherboard out there that can handle it regardless of supercooling. Sorry.
AMD (and Intel) run on 13 week quarters, not 3 month quarters!
So do they must have a quarter that is 14 weeks at some point or they wouldn't stay in sync with a calendar year. I guess that is just a bonus quarter when it happens as they get an extra week of sales.
AMD pretty much had hit a brick wall now. Channels have short supply, many titles backordered, pricewatch is empty.
What parts are in short supply? From a consumer aspect, I had no trouble locating an Athlon 64 3200+ for what I considered a good price. I guess OEMs have a different view, but aren't they in general ordering direct from one of AMD's main distributors? How exactly do we know what inventory levels they are holding? It seems we only have some limited data from smaller distributors. I might have missed some valuable information in one of the posts, so please correct me if I'm wrong.
It really looks to me like AMD is in the sweet spot. Every different Athlon 64 SKU is available, and at a good ASP.
The only question in my mind is what is going to be the unit ratio of k7 to k8 in Q3. k7 ASPs are dropping like a rock, so there may not be enough k8 sales to make up the ground. It will at some point be a huge rise in revenue and ASP. I'm thinking Q4 is going to be a blowout. Q3 might simply just meet expectations.
Yep. OEMs are tired of accumulating Intel inventory and have stopped ordering...
Yes. You could clearly see an inventory build up by Intel in the face of increasing competition, and now they are paying the price. It may be that Q3 is going poorly as well, but I think Intel is going to feel it more than AMD this time.
You didn't highlight 'as it expected'.
AMD has had a very slow July and August as it expected
July and August are usually very slow. I don't see what the big deal from the report is.
AMD Demonstrates World's First x86 Dual-Core Processor
http://biz.yahoo.com/bw/040831/305770_1.html
Nice. Not much in the way of details, and products are not expected until Mid-05. I guess it is nearly September, so that isn't that far off.
edit- I'm too slow - subzero beat me to it...surprisingly with no comment.
Sun knows the DAY sunw ports Solaris to X86 is the day they are out of the hardware business.
Not necessarily. They might be out of the chip business, but there are a lot of system architecture issues that Sun is very good at where they can add a lot of value. They can be more than just a clone with this model. I expect them to start making large Opteron MP machines that no other OEM will have, and they could charge big $$$ because there would be no competition. Being out of the chip business might not be a bad thing for them either. There is so much fixed cost associated with chip making that it just doesn't make sense to carry on with the low volume sparc.
And Dothan is kicking Athlon's butt every way to Sunday. Not bad for a "slight advantage."
Of course you are comparing a 90nm part to a 130nm part. This advantage will soon diminish. I would bet AMD's 90nm A64 is going to be close to Dothan in power consumption. We'll know in a couple of weeks (perhaps even next week?).
I went out on a limb today. I think we're at a bottom for both the market and AMD. I picked up a bunch of Jan 04 $15 calls. I'm also short August $13 puts and it is pretty clear I'm going to get assigned the shares. At that point I'll be fully loaded once again. I had previously bailed after the Q2 earnings which I thought were disappointing enough to drive the stock down short term. I think we're just about through that. Others may disagree, but I think the year will end with a bang, especially for AMD with their very strong product lineup.
Thanks so much for posting this Buggi!
Once again the 10Q has tons of information they refuse to give out in the conference call. I don't understand why they do this. Maybe they figure nobody reads the 10Q.
a one-time retirement benefit expense of $5 million
Jerry? :)
It only seems terrible to those who had unrealistically high
expectations.
The problem is you may have just described the average AMD investor, in which case there may be a lot of disappointed investors selling their stock. This is the reason I bailed. I think I can get a lower reentry point than $13.50. If I'm wrong, so be it. I don't care to risk a 3 point drop and lose my entire profit in a mass selloff. If the stock shows some strength over the next week, I will get back in.
For now I'm on the sidelines to see how it plays out.
After closer review of the financial statement, I no longer think the quarter was a disaster. Operating income was up over Q1, and margins increased. Expenses seemed to be kept in check. I may have been too hasty in the sale of all my shares. I'll be looking for a reentry over the next few days. I'll probably use an aggressive put writing strategy which would result in me having a lower cost basis than before, although I just blew my long term gains tax advantage.
Just dumped all my shares of AMD. Some of these shares I had held for over 2 years. Maybe I'm overreacting, but I think we are going to see significant selling after this.
Made an overall gain of 35%. I should have dumped a couple weeks ago when it was 60%. :(
I really expected better than this for Q2. AMD did not capitalize on the opportunity presented.
I haven't lost complete hope for the future of AMD, I just think I'll be able to reenter nearly the same position with a put writing strategy over the next few months.
Good luck to those who have the cojones to hold. I'll sit this one out.
I've seen several posts about Dell's adoption of Opteron being insignificant because of the low volumes involved. It seems to me, though, that Dell has a more important concern: they don't want their large, tied-up customers to add HP or Sun as approved corporate suppliers.
You are on to something here.
I know of a couple of companies who buy Dell exclusively because of some kind of service arrangement they have with them, and they will not consider using other vendors unless they are forced to. One of the IT people at one of these companies has said they are reconsidering their policy because they are needing some Opteron Linux boxes for 64-bit development. Once the door is open to another vendor, it could mean a whole lot more machines are purchased through that alternate vendor.
Dell has already offered some larger customers Opteron machines manufactured by another company but under their service contract just to keep these customers happy. This is what was witnessed not too long ago with that Opteron page on Dell's website.
I think it may have finally come to the point that Dell is seeing enough customer demand for Opteron systems that they are going to introduce their own Opteron lineup.
solved their issues and will reach 4Ghz easily this year
Bring it on, I say. 90nm Prescott at 4Ghz is not even going to stack up to an FX-55, which is due out in the fall, and on 130nm no less. Now if Intel has 4Ghz and 2M L3 together, then there may be some trouble, but it still isn't a terrible threat.
AMD may have a total market cap exceeding a half of a trillion dollars by then.
I would be sitting on the beach drinking $50 drinks surrounded by gorgeous women staring at my 100 ft. yacht long before this ever happened!
We can all dream. I would be happy with $30 at this point which would be a triple for me. I expect we'll see $30 before the Q4 earnings announcement in january, and then I'll reassess AMD as an investment at that time. 2004/2005 look great. My big question is, what is going to happen to AMD in 2006 once Fab 36 is online. With 2 fabs running and one being 300mm, AMD could potentially supply about 40-50% of the entire CPU market, but not without an ugly fight. I'm not sure I have the stomach to hold AMD through such a period. This is also about the timeframe Intel will start coming on strong with cores beyond the lacking P4 derivatives, and that could be a whole new playing field.
How many things have to go right for AMD to benefit from
an advantageous position? And how long can all these
factors realistically stay right?
Contrast that with how many things have to go wrong for
Intel to suffer from a disadvantageous position. And how
long do all these factors realistically stay wrong?
I think AMD has already established a place for itself. If things stay the same as they are now, they should be quite profitable for the foreseeable future. I don't expect Intel to keep making mistakes. I wouldn't really say everything is going right for AMD either. Where is 90nm? It should have been out 6 months ago.
I think we'll see a more balanced game with no clear winner. Intel will continue to crank out profits based on superior manufacturing prowess, but I don't think they will have the market stranglehold they had several years back where they could control prices, and I don't think they'll ever get it back. Winning for AMD is to be profitable year after year, not to destroy Intel. I should have been more clear. My conservative target for AMD is to see them make $1/share/year. They are set up perfectly for this and possibly more, so I think it justifies owning a lot of AMD shares at current prices regardless of what Intel is doing longer term. Intel doesn't have a decent response in the short term.
Interesting article. I liked this part:
"With real-world applications we have tested, the uglier the code, the bigger the advantage is for Opteron over Xeon."
I have had similar findings. It is a shame many of the public benchmarks out there don't really tell you which chip is best for your needs. The good thing is OEMs will give workstation and server samples to large companies on a trial basis to determine this. Opteron is winning a lot of prestige in these situations.
Those who cannot remember the
past are condemned to repeat it
- George Santayana
Did he say anything about those stuck in the past?
I agree remembering the past is good. It shows how well positioned AMD currently is. Sure they could still fumble the ball and probably will at some point, but it doesn't mean they won't win the game. Contrast AMD today with AMD 10 years ago. Contrast Intel today with Intel 10 years ago.
first step to doinga 64-bit build.
What's really funny is that I have a huge build running on the 2.0 Ghz A64 and it's cool and quiet andjust typing this in a web
browser on te 2.0 P4M has it fairly warm with te fan
I think your A64 notebook is having a problem. It is skipping some h's and spaces in your message. j/k :)
But seriously, I have found A64 machines to be the fastest compiling machines period. I'm glad to hear this doesn't just apply to Linux, but I find it appalling that there is not a native AMD64 .net at this point. MSFT bragged a couple years back that it took them only a couple of days to get the kernel ported to A64, yet it has been 3 years and there is no .net, native compiler support, etc. What a joke. I'm glad my work is on a Linux box. I don't think I could put up with what you are having to go through.
what AMD needs now is a nice jump in profits from $45 million to $100 million...anything less than $60 million and this stock will stay at $15 a share for awhile and I might lose interest in this stock as an investment play
The seeds are in place for large increases in GM going forward in both CPU and memory. I don't think we'll see $100M in earnings in Q2, but it is likely for Q3, and I think Q4 will be over $.50/share.
I think patience is key for a long term AMD investor right now. Trying to trade this stock right now is going to be tricky. It could run to $25 quickly, or stay at $15 for another couple of quarters. I'm holding out until at least January before selling any of my shares.
So is this the correct mapping from models to frequency/cache?
3700+ (754) 2.4Ghz 1M
3500+ (939) 2.2Ghz 512k
3800+ (939) 2.4Ghz 512k
FX-53 (939) 2.4Ghz 1M
Older parts
3400+ (754) 2.2Ghz 1M
FX-53 (940) 2.4Ghz 1M
I haven't been following all the various models to know for sure, so please correct me if I am wrong somewhere. I've been waiting on the 939 platform before building a new system, and I want to know what my options are.
I wonder what PR rating FX-53(939) would have if it was to get one. 3900+? 4000+? I also wonder if the ratings above have been adjusted somewhat for Prescott Mhz instead of P4 Mhz. The lineup above looks really strong. FX-53 on 939 is going to be way out in front.
OT AMD common
LOL. Please do not discuss the stock on this board. It clutters up all the other relevant material. :)
More OT. I think an upwards breakout in AMD is imminent. I still don't get the analysts. They are either complete idiots or geniuses (if they are artificially holding their ratings down while their companies stockpile shares). I guess they'll all pile on at some point. Probably when it is time to sell.
But there are times when interruptions save you from bad trades...
You mean I can't ignore those? I thought the norm on stock boards was to report your gains and ignore your losses? :)
There was a period of time of several months last year when a
bunch of us would await the monthly short interest reports
I love the little trading niches like this you can find in stock trading. I just don't have much time to play around right now though. Sounds like you are in the same boat.
AMD is clearly a buy right now, technically and fundamentally.
The short interest has drifted higher over the past month. A huge spike is worrisome, but a steady rise is bullish.
The short ratio is amazingly high right now. This is only going to help the push to get over $20, when it comes. I would expect sometime in Q3 this will happen.
I know folks that have the same problem where work gets in the way of trading.
I can't even remember how many times people have walked into my office, talked to me for 15 minutes, and cost me several thousand dollars. It has happened so much, I started keeping a tally of how much each person has cost me. I jokingly call it my blacklist. Someday when it becomes a big enough number for a certain person, I'll let them know how much money they have cost me with all their interruptions and how I expect reimbursement. For now I keep it secret. :)
This is one reason I have been tempted to exit corporate life and enter some business where I keep my own hours. I like the office BS so much though that I'm not sure I'll find that as fun.
Intel sure took off today! Up 2.34, or 9.08%
I think your quote is off.
28.10 +0.36 (1.30%)
Onward and upwards until Q2 earnings, mid-July. I'm expecting to see $22.00 before then. Put your chips down if you want to enjoy the ride!
Well I'll throw in my outlook.
I think it will be tough to crack $20 before Q2 earnings, but I do expect upper 20s/low 30s before the end of the year. The summer can be brutul to chip stocks. Q2 is not traditionally a good quarter. I expect AMD to outperform the industry due to gains in the server space. ASPs are rising fast for AMD. 100K additional Opterons at $500 profit is $50 million extra in operating profit, which is about $.12/share. I'm really feeling that Opteron is starting to take hold and making a name for itself. By the end of the year this momentum will be obvious. Q4 is going to be a blowout.
to inline charts so easily.
I welcome your charts. Keep em coming!
I only follow a few indicators myself and they have served me very well for spotting trend changes. I pointed out a problem in the chart last week, got thrashed for it here, and then AMD tumbled a point and a half. I use simple things like MA (20, 30, and 200 day), Stochastic (%k 21 %d 14), and Bollinger Bands on a 20 day MA. I'm always looking at others, but I'm so used to looking at these that others don't seem to add much value. After looking at charts for so many years, I'm also finding I don't need the indicators other than MA curves.