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.
Economaniac -
Do you really think that extra .7% makes the difference between "edging out" and "kicking butt"? Is it just possible that you aren't being entirely objective in comparisons between AMD and Intel?
Well... there is an outside chance....
Economaniac -
Intel has been getting base and peak scores which are essentially identical, which has always suggested to me that they simply added optimizations within the commercial compiler that detected the test programs and implemented the peak tuning.
I'm not willing to leap to outrageous assumptions with no data whatsoever to back me up.
It is mostly a matter of bragging rights with limited value in evaluating real world performance, which can be directly measured on the actual applications.
The significance is that AMD used SPEC scores (and base at that) to tout Hammer's performance many months ago. I'm pointing out that AMD failed to deliver on their promise, even by their own measure.
Economaniac -
It is generally believed that base scores are the better choice when comparing. Base scores require that the same set of compiler options be used for all tests. Peak scores allow a different set of compiler options be used for each individual test. Peak scores are much more finely tuned and may not be as representative.
Tom's Opteron review -
http://www.tomshardware.com/cpu/20030422/index.html
Joe -
BTW, what are the base and peak scores of the mysterious 3 GHz P4 with the patch applied?
Don't know but I don't expect much change, if any. We'll see.
Additionally I would expect to see a new speed boost from P4 soon. They've been at 3.06GHz for a long time and I don't think the process has run out of headroom yet.
Let me say that there's no denying that Hammer looks pretty damn good in 2 & 4 way configurations. Now, can they produce it?
EP
Joe -
The fact that all the Intel processors combined are able to steal a win here and there vs. 1 concrete Opteron processor doesn't mean that Opteron is not the overall winner. With your approach, you could even add Alpha to the mix and point to some benchmark where Alpha is still leading and say that Intel is ahead.
Yes you're right but this small set are AMD handpicked benchmarks. Don't you think Intel could handpick a set of benchmarks that told a completely different story?
There are some issues here that tell a different story when you look a little deeper. I'm sure we'll have lots of comments on these scores and those that emerge from independent websites. Have you noticed that AMD is quoting peak scores here? Doing a comparison between base scores is not so flattering to Opteron.
Economaniac -
you're gonna gloat cause Intel got out a software patch in time to claim that the 800FSB P4 was a real product?
Intel beat Opteron in SPECint with the 3.06GHz P4. They didn't need the higher performing 3.0GHz P4 with 800MHz FSB.
All in all some very impressive scores but a very limited set. I expected a little more. Where are the SAP scores? Now let's see how these are interpreted by others and also let's see if they can actually ship this product.
Let's digest for a while.
More details -
Despite claims that Hammer would beat anything Intel has to offer, Hammer gets it's but kicked by Itanium in TPCC and edges out Xeon (for now)
http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_8796_8800~69696,00.html
More details -
Despite claims that Hammer would beat anythin Intel has to offer, Hammer get's it's but kicked by Itanium in TPCC and edges out Xeon (for now)
http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_8796_8800~69696,00.html
Official Hammer Benchmarks -
Thanks to Joe -
As I predicted, Hammer fails to match Intel SPEC performance despite their promises.
Opteron SPECint_base 1095
Intel 3.0GHz P4 1152
Intel 3.06GHz P4 1099
Opteron SPECfp_base 1122
Intel 3.0GHz 1202
Intel 3.06GHz 1059
Many other benchmarks where Opteron looks very good. We'll need some time to digest this.
http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_8796_8800,00.html
Official Hammer Benchmarks -
Thanks to Joe -
As I predicted, Hammer fails to match Intel SPEC performance despite their promises.
Opteron SPECint_base 1095
Intel 3.0GHz P4 1152
Intel 3.06GHz P4 1099
Opteron SPECfp_base 1122
Intel 3.0GHz 1202
Intel 3.06GHz 1059
Many other benchmarks where Opteron looks very good. We'll need some time to digest this
Chipguy -
If true the current process + current stepping yield vs frequency curve peak occurs around 1.5 GHz or so.
I wouldn't read that much into the pricing. At this point pricing is as much Marketing posture as anything else. We'll find out over the next few weeks.
Chipguy -
Anyone want to bet that Opteron will reach 2 GHz before P4 reaches 3.2 GHz?
I won't touch that bet.
Not a Short
Do you expect those scores to be acurate?
Sure.
The rereleased cpu has a software patch. I'll wait for new benchmarks to be done on a patched system if you don't mind.
I don't mind.
Intel resumes 3-GHz MPU shipments after delays
http://www.siliconstrategies.com/story/OEG20030421S0062
Now that Intel has resumed shipping the 3GHz P4 with 800MHz FSB, the new SPEC targets that AMD promised Hammer will beat have now risen.
P4 3.00GHz 800MHz FSB Base scores.
SPECint 1152
SPECfp 1201
http://www.intel.com/performance/resources/desktop/tables.htm
Intel resumes shipments of Pentium 4 with 800MHz bus
http://www.ebnonline.com/showArticle.jhtml%3Bjsessionid=1H3ARJ2UQHIRKQSNDBCCKHSCJUMEKJVN?articleID=8...
belgiangenius -
So, you're actually trying to prove Hammer's performance by quoting AMD officials?
No. My post was sarcasm. I'm pointing out what AMD promised.
Sgolds -
I am under the impression that the engineering samples were the 256KB cache version, so they are not comparable to Opterons (in that respect) as Ephud thinks.
I will assume this is an honest mistake and not the obvious lie it seems at first glance. If you will look at the below link you will see that it is a 1Meg L2.
http://www.xbitlabs.com/articles/cpu/display/athlon64_2.html
As I know you are a man of integrety, you obviously didn't intend to misrepresent the facts, despite appearences and no matter how tempting. Too bad you are unwilling to extend the same consideration to others...
CJ -
"But would floating gates' speed be helped by floating loans?"
Good point. Yes, but limited currency mobility can cap the performance gain...
And your profits could go in a Flash...
SGolds -
I take the cache assertion as an honest oversight as Athlon64 is suppose to have less cache than Opteron.
I believe you are in error but I trust it was an honest mistake. There have been numerous references to Athlon64 at 256K, 512K and 1Meg L2 cache.
Mr Golds.
An A64 has the same core, cache and similar memory controller as an Opteron plus it's on the same process. If they aren't debugged yet they AMD is going to be very embarrassed Tuesday...
Spokeshave -
Gee, Elmer, please remind us what it is you have been saying all along. You haven't mentioned it in, oh, say, three or four minutes or so...
That's because I know that some of you will forget and I just have to say "I told you so...."
The chances of a diehard AMDroid admitting they stood in line for the chance to buy the marketing hype hook line and sinker is near zero.
All this assumes of course that the rumors will turn out to be true...
Paul -
I don't get what all the negativity is about. @ 5% improvement(core) over equal clocked XP was expected, right?
One of the glaring issues is that SOI hasn't bought AMD anything. It cost them millions, over a year of delays and lowered frequency. Not much of a deal. AMD could have done a bulk silicon Hammer much faster, much sooner, much cheaper and it probably would have performed and yielded much better.
EP
wbmw -
I plan to sell my shares as soon as the initial Opteron news passes, because there is little chance that AMD will be able to ramp the product for revenue before more bad news floats this way.
I suggest you consider selling in the money (ITM) CCs. For example you can get ~$3.10 for a May $5 call. If you get called away then fine, you got your money plus a little extra. If AMD drops you are protected to below $5.
You could even get ~$0.90 for a May $7.50. Do the math...
The Junes will open Monday morning and will offer still more opportunities...
CJ -
Repeat after me, engineering silicon and 6 months to product launch.
Engineering sample? Opteron is being released in 2 days. Athlon64 has the same core and same L1/L2 or if the L2 is smaller then the performance numbers will take a further dive. Don't expect a major change as AMD brings yields up. Meanwhile, I can assure you, Intel will not be standing still.
belgiangenius -
AMD ATHLON64 seems to get totally destroyed by PIV in most benchmarks:
BS meets reality...
http://www.infoworld.com/article/02/01/07/020107hnchipwar_1.html
"All in all, Clawhammer, based on what AMD has said about it, could be a real barn burner. And if that's the case, then just as the original Athlon was faster than the Pentium III, AMD expects Clawhammer to dramatically outperform the Pentium 4," said Brookwood.
"Clawhammer will definitely be revolutionary in terms of performance," said AMD's Crank.
wbmw -
Everybody knows that this is Elmer's First Law of AMD Behavior. <VBG>
Thanks for the credit. While he captured the spirit, I'd have phrased it a little less politely...
EP
Greg -
A good summary of what some of us (Myself included) have been saying all along.
CJ -
AMD never promised an Opteron at 3400+, that was the Athlon64.
The one that's still almost 6 months off assuming AMD can actually produce it? The one with recent benchmarks showing rather disappointing performance, often compared to a 2.53GHz P4? The one AMD promised to introduce EOY 2001? That one?
Let's be clear. AMD published estimated SPEC benchmarks of a 2.0GHz Opteron. The clear impression was that this was the device that AMD intended to produce. Now, a delay later AMD can't deliver what they lead everyone to believe they would. You can argue the fine points but you know that AMD promised something far different from what it looks like they will deliver, even with delay after delay after delay.
All this assumes that the rumors will turn out to be true. We'll find out something Tuesday and later we'll see if they can actually deliver what they claim or if it's still vapor.
wbmw -
This was the 1.8GHz part, formerly the 3100+. Still, you make a good point. It's not quite as fast as Intel's best SPEC scores, but since it is a server CPU, you should compare with Intel's Xeon processor, and use SPEC_CPU2k_rate, rather than SPEC_CPU2k. The Opteron SPEC_CPU2k_rate scores are truly impressive.
SPECint and SPECfp are the only benchmarks estimates AMD released when bragging about how Hammer would be the highest performing processor in the world. I think it's fair to use their own measure when pointing out that they are not delivering what they promised, as I predicted from practically day one.
I think faster frequencies from AMD could lead in 2-way SPEC scores, until Intel releases the supposed 1MB Prestonia core. The 1MB of cache will help Xeon to scale much better. That and 3.2GHz later in the year should be able to match anything that AMD has to offer.
I have not seen any public document where Intel clarifies what this product will be or what we can expect from it.
So you might be right about AMD not quite fulfilling their promises of "world's highest performance", but we'll have to see if other benchmarks this next week tell a similar story.
Yes, I just might be right and by AMD's own measure. Assuming of course that these rumors turn out to be true...
EP
More Benchmark rumors for Opteron -
http://www.theinquirer.net/?article=9039
The Opteron 244, equivalent in frequency terms to a 1.8GHz chip, is a Newisys 2100 dual CPU system, and the base values as posted are for SpecInt2000 1048 (1 CPU), SepFP2000 998 (1 CPU), SpecFPrate 18.4 (2 CPUs), and SpecINTrate 22.4 (2 CPUs).
So the mighty Opteron 3400+ that AMD promised is about equivalent to a 2.9 GHz P4 (if there was such a thing) running on Intel's older chipset in SPECint and about equal to a 2.53 GHz P4 in FP. A far cry from the "world's highest performance" that they promised years ago.
Assuming of course that the benchmarks posted above hold to be true, and assuming AMD can actually produce this part at the rated speed.
EP
wbmw -
To me, Athlon 64 is unimpressive.
Add to that unmanufacturable and it's really a bomb.
IDC: Dell's back on top of worldwide PC shipments
http://www.computerworld.com/hardwaretopics/hardware/story/0,10801,80464,00.html
Neye_eve -
Heh, I remember buying the Qs at $103 - ugh (nt)
As they say, don't feel like the Lone Ranger...
BTW, what does (nt) mean?
Economaniac -
you must have quite a bit of equity tied up in something for your broker to let you write naked puts (let alone naked calls).
I learned the hard way the necessity of cash reserves as opposed to heavy margin. The tech slide since 2000 + 9/11 almost did me in. I now keep substantial cash on hand. When I write puts I limit my exposure to about 125% of my cash reserves and I generally write them for 2-3 months out. I'd be willing to accept assignment on the Qs @ <$25. I think QQQ is the best of all vehicles because both the shares and the options are extremely liquid with incredible volumes and open interest. Additionally the options are traded in $1 increments and the shares are equivalent to the Nasdaq 100 so you have good diversification. If only SPY would offer the same...
Having watched the Qs go from about ~$125 to <$20 I now feel the downside risk is manageable and I'm preparing to get back in. Another lesson I learned the hard way, never write puts on something you don't want to own because you just might. I do not generally write naked calls but I've been known to.
Your decision to have most of your long positions in AMD & ATYT seems awfully risky. Not that they can't perform but you have no diversity. Why are you putting yourself at such risk?
Thanks for commenting. You'd think that on an investment forum we'd have more discussions about investing?
EP
Economaniac
Writing a Put is almost exactly equivalent to buying the stock and selling the covered call at the same price.
The operative word here is "almost". Selling a Put does not require tying up cash although in this interest market there is little lost. Buying shares on margin requires interest charges, selling a put doesn't. Selling a put involves only one commission where buying shares and selling calls involves 2 commissions, at least. Either can be rolled as needed. I guess it depends a lot on your expectations. I expect a generally flat market with a slight upside bias. I like straddles now because they guarantee one position will win and the other can be rolled.
What are you doing anyway?
wbmw -
you do know that you can't make money in the stock market unless you sell (assuming no dividends, which is the case here).
I'm making money and I'm not selling (shares).
Edgar -
think it's hilarious how when AMD was back in the 3's,4's and 5's for that matter that people where talking about the demise of this company. Going bankrupt and never staying a thorn in Intel's side like it does.
http://www.insanely-great.com/news.php?id=1658
IBM: AMD Will Be Dead in 5 Years
by Remy Davison, Insanely Great Mac
February 7th 2003
Advanced Micro Devices (AMD) will be out of the chip market within five years, predicts Bill Zeitler, Senior Vice President & Group Executive, Server Group. Zeitler made the comments to Fortune magazine, and the article is quoted in an Inquirer story. [attrib. MacBidouille].
more...
YB -
sgolds, probably no winner will ever exist. Intel will never drive AMD to bankruptcy, AMD will never take 50% of the market. But AMD may fluctuate between 15% and 30% of the PC and x86 market, and that will also change their profits between minus and plus infinity.
So play the range.
My April $5 Puts close out this week. Nice little premium stays in my pocket.
The CCs I rolled to July gave me a nice little premium too.
I'm now short AMD July $5.00 Calls and Puts. July $7.5 Calls and May $7.5 Puts.
On the INTC side I've been writing CCs each month at $20 and I'm pleased with the outcome. May start writing Puts at $17.5. I'm also writing lots of QQQ puts at various prices in the lower $20s. Also QQQ CCs @ $27.