Analysis on IT trends and competitive strategies, with emphasis on micro processors, computer systems and networks. Based on latest news, backed up with real data, this site intends to provide a true and realtime picture of the fast changing IT landscape. This journal strives to be accurate on facts and sharp on criticisms. You may email your opinion to sharikou@yahoo.com or post comments here, be cool and intelligent.
Freelance journalist on IT matters. Some of my writings have been published on online IT journals. Any original content on this journal is Copyrighted, but it's free for non-commercial use. Any Trademarks used on this site belong to their respective owners. Some of the pictures are links. If there is any issue with the content of this site, please email sharikou@yahoo.com .
The K10 is here to frag Clovertown. The story was a 2.3GHZ K10 is 50% faster than a 2.6GHZ Clovertown. In other words, a 4GHZ Clovertown is needed to compete against a 2.3GHZ K10, and a 3.5GHZ clovertown is needed to compete against a 2GHZ K10.
posted by Sharikou, Ph. D. @ 7/03/2007 06:53:00 AM
26 Comments:
The funny thing about that Inq story is that they say those were done a while ago, so perf MAY HAVE increased.
Um, these are PAPER estimates. Nothing real. AMD hasn't shipped one Barcelona yet, and the Barcelona ramp seems to be doing QUITE badly for AMD: http://www.theinquirer.net/default.aspx?article=40738
Now mind you these are PAPER numbers for the 2.3Ghz Barcelona which isn't scheduled to ship until sometime in LATE OCTOBER.
But when has that ever stopped Ph(ake)d from publishing pure bullshit on his so called blog.
When Intel doesn't go BK in 2008 we can all pray that Ph(ake)d goes and offs himself in some non-spectactular factor.
Not like anyone would notice this flame mouth going away...
Of course, now all his screaming fanbags will be showing up to defend fucknuts...
I don't think Evil_Merlin likes the good doctor very much.
Your math still is lacking.
Evil Merlin: Hey man, don't let these guys get to you, it's only CPUs. (Plus Sharikou just looks dumber every day too).
Christian m. Howell the 3rd: Performance may very well have decreased, and in fact that is more likely. Why? Well several months ago they were only running simulations of ideal chips. The fact that I can't buy a K10 now shows that there were logic bugs in the CPU, and often times the fixes for those will hurt performance since the ideal simulations were made to be as aggressive as possible, while the fixes usually slow things down.
A good parallel could be in multithreaded code: It might perform faster with fewer locks on shared objects, but then timing bugs and race conditions make the code crash or deadlock in bizarre ways. Adding more mutexes will fix the bug, but also slows down the 'theoretical' performance since threads have to serialize. Timing bugs in CPUs are not exactly the same, but they can get similar race conditions as well that require more latching/locking that slows the overall chip down.
If AMD is going to get anywhere, it needs to fire all its marketing stuff. Intel is still selling dual core 32 bit junks like hot cake.
AMD's desktop offerings from HP and Dell are the best values that informed buyers grab without any hesitation. AMD has miserably failed with rest of the buyers. No matter how much better AMD's offerings are or will be, with the current sales and marketing stuff it is doomed. It is not going to go BK, but it will barely manage to stay alive.
Isn't the 20% faster at integer & 50% faster at floating point some claims that appeared in April but have never been proved ?
The figures given comparing Opteron & Barcelona state that it was for a dual socket and 4 socket systems.
Wouldn't that mean that in a 2p system they were comparing performance of 4 opteron cores vs 8 Barcelona cores (& 8 vs 16 in a 4 socket system)?
yipikayey you mf
Why should I like Ph(ake)d...
1.) he lies publicly about his education.
2.) He lies about Intel any chance he can get.
3.) According to Ph(ake)d, Paper launches for AMD are perfectly fine, but not so for Intel.
4.)According to Ph(ake)d, Erratum are perfectly fine for AMD but not so for Intel.
5.) If a web site does a Pro-Intel review, they are paid Intel pumpers. If the same exact fucking site does a pro-AMD review, all is forgotten and its the word of Gawd.
6.) The fanboi's around here are just blind. They worship this web page as if they were Fundie Christians and the Bible. AMD can do no wrong and everything Intel does is wrong.
But since I like getting fanboi's all riled up, its a good place to hang out and watch people like Pezal, Penix, Howell, onlyamd and expert make totally asses of themselves...
evil_merlin, you are the most pathetic of all the idiots on this page. Look at you, proudly displaying your little "Ph(ake)d" line where ever you can. Did you think up that witty little piece of shit all by yourself?
I used to think your fanboyism was a show you put on to troll this blog, but I'm now convinced you truly that confused. Your ignorance is nothing less than vile and I am ashamed to have to share this world with you.
Just as I said here come the liar's supporters.
Is your life THAT bad Penix you need to support proven liar?
If my ignorance is so vile, one has to wonder just how bad your little world really is, as EVERYTHING posted by me is 100% fact (well other than Intel's profit's which were over 5 BILLION dollars last year, not 8.5 Billion)...
So let me get this right, you would rather plug your fat little fingers in your ears cover your eyes and deny the TRUTH to keep supporting your precious Ph(ake)d?
What a fucking douche.
Great life! Keep ignoring the facts and maybe some day you can move out of mom's basement and head to the loony bin where you belong.
Evil_Merlin said...
6.) The fanboi's around here are just blind. They worship this web page as if they were Fundie Christians and the Bible. AMD can do no wrong and everything Intel does is wrong.
I swear I could detect rage and frustration in this guys comments!
Look in the mirror Merlin (oh so evil). Who worships this web site more? You or "us" fan boys?
Take a breather before you burst a vein in your forehead. Go read a few blogs on the bible for a change and come back when you've calmed down :)
Has the reality of a 'possibly' amazing Barcelona product just suddenly hit you?
A paper launch is NOT:
a) a company that has delayed or delays a product launch. That is admittance!
b) a company that has unexpected production glitches and prevents them from producing enough products. That is execution errors!
c) a company that can't supply the stronger than expected demand. That is management problems!
A paper launch is when a company says it is shipping and pretends to accept orders but has knowingly nothing to deliver.
evil_merlin is as blind as a fanboy can get, and he has a big hard on for Intel. His opinions are completely worthless and should not be taken seriously.
It's important to make sure that every post you make is completely without bias and based on facts and reason. I make it a point to do so before each and every post I make, and I suggest everyone else do the same.
Often I run into some old geezers who say ...AMD ...hmmm can it run Windows ??? I think evil_merlin is one of them.... :-)
Isn't the 20% faster at integer & 50% faster at floating point some claims that appeared in April but have never been proved ?
Yes, but it was saying that it was at the same clock. The new claim is that 2.3 GHz AMD trumps 2.6 GHz Intel by 20% to 50%.
In other news, yesterday I implemented a 128-bit x86 CPU on an FPGA, and it will outperform Barcelona by 3000%. Even funnier is the fact that my CPU's performance automatically increases every day 5000%.
jeach! you ignorant little fanboi, you need to look up what PAPER LAUNCH means.
1.) Can you buy ANY Barcelona processors yet? Nope.
2.) When Barceleona becomes generally available, will you be able to get a 2.3 Ghz unit? Nope, not until 3-4 months following it's inital release.
3.) Can't meet exceptional demands for the prodcts? OMFG get real. AMD can't even meet the demands of Sun and Cray, both who have suffered stock market impact due to the inablity to deliver products.
Once again, AMD fanboi's running to defend AMD's fucking barn sized screw ups.
But had this been Intel screwing up, holy fuck boys, there would be a good ol' lynching.
Again Penix you fucking dripping douche, show me ONE post I've made that isn't dead on accurate. In the meantime I can show plenty of yours which are based on supporting nothing more than the verbal diarrhea that pours out of the diluded mind of someone who CLAIMED to have a doctorate, but like a moron, was found to be utterly lying about said doctorate, and who makes totally off the wall, bunkish claims such as Intel is going bankrupt in 2008.
Who the FUCK is stupid enough to support that kind of crap...
Oh yeah AMD fanbois!
Good going!
All these performance claims from AMD are proved totally false:-
http://forumz.tomshardware.com/hardware/Intel-lose-market-share-Barcy-ftopict243067.html
AMD is making wild claims with CPU speeds at 2.3 and 2.6GHz. The thing is, Barcelona is still not out. When it is out, in August, if they don't paper launch it, the best you'll get is a paltry 2Ghz.
I think the AMD numbers actually point to it being pretty even to Intel on widespread release.
If the range is 20% to 50% improvement over Clovertown, depending on the benchmark being integer or floating point then lets just split down the middle and say 35% in general.
A 2.4GHz K10 will be quivalent to a 3.2GHz Clovertown, a 2.6GHz 10 will be a 3.5GHz Clovertown. However these AMD speeds will be out when Penryn is out which is about 10% faster so you have 3GHz to 3.3GHz Penryn which is actually where this Intel chip is likely to be (11x333 or 8x400).
If they are roughly the same then it will be interesting to compare the power use and cost.
AMD is behaving unethically again:
"I’ve seen benchmarks get cherry picked and twisted before but this is just outrageous. AMD is deliberately leaving out Intel’s best scores, leaving out Intel’s best products that shipped months ago, and putting in theoretical Barcelona scores for products that don’t even have a ship date. After Henri Richard (AMD executive) came in front of our ZDNet cameras to slam Intel for “un-ethical behavior” and promising not to do the same, we have caught them on four separate occasions behaving unethically. After this latest incident, it’s clear that AMD has no intention of behaving honestly or ethically."
http://blogs.zdnet.com/Ou/?p=567
Alright, it's 4th of July people, get out and enjoy some fireworks (and I don't mean the flames on this board)!
Here's a little paper shredding of some overhyped theoretical numbers.
Evil_Merlin said...
you need to look up what PAPER LAUNCH means.
Right, and I guess you invented the definition, so yours must be right!
Can you buy ANY Barcelona processors yet? Nope.
To be release idiot! Of course you can't buy it it's not yet released!
When Barceleona becomes generally available, will you be able to get a 2.3 Ghz unit? Nope, not until 3-4 months following it's inital release.
Your so amazing that you can predict the future... wow!
Can't meet exceptional demands for the prodcts? OMFG get real. AMD can't even meet the demands of Sun and Cray, both who have suffered stock market impact due to the inablity to deliver products.
And you proclaim that all your comments are sane and logical ???
Sharikou, not to impose my standards, but anyone wishing to demonstrate emotion by typing F!@#$ once will get the point across. I'm not sure why you don't sensor this idiot's comment for using the actual word directed at writers, especially so many times??
[...] mind of someone who CLAIMED to have a doctorate, but like a moron, was found to be utterly lying about said doctorate
Using the term 'DR' in as an avatar/alias/username/handle (pick one) is NOT pretending to be a doctor or have a doctorates degree!
Do you go around and blast all those who pretend to be elves in fantasy worlds like WoW or Everquest? No I didn't think so! So just come to this blog with an open mind and have some fun for once!
When Sharikou proclaims to be a doctor and tries to get you to become a paying client... then you can start to yell foul. But I doubt he'll do that because at least HE can distinguish the difference between 'real-life' and being 'on-line'.
Who the FUCK is stupid enough to support that kind of crap...
You are! I mean your here every day supporting it (by reading) and worst, you are one of the biggest contributers to all of this so called 'crap'!
jeach!... yet another fanboi who doesn't understand facts and hangs off of every lie that utters forth from Ph(ake)d's pie hole.
Dude, you REALLY need to learn what a paper launch is.
You also need to learn what estimated (IE PAPER) benchmarks are.
You also need to learn to think for youself.
Patty where is he?
Here he is..
http://www.intel.com/quadcoreserver/index.htm?iid=technology_architecturesilicon+body_quadmainstream
"When Barceleona becomes generally available, will you be able to get a 2.3 Ghz unit? Nope, not until 3-4 months following it's inital release."
"Your so amazing that you can predict the future... wow!"
AFAIK AMD themselves announced that barcelona would only be around 2.0 ghz at launch...
Evil and Jeach, are you quite finished? AMD admitted that these numbers are as or more cooked than Enron's books. They've even said they'd take them down. Bleh.
Everybody, have a drink.. this is only the internet.
AMD can't even get their benchmarks right. Got silicon in hand and still have to resort to "simulated" benchmarks. They can't even over clock their engineer samples to 2.6 how sorry is that.
Can you say speed path issues
Can you say thermal issues
Can you say big yield issues
Can you say a 3 billion dollar fab emppt
Can you say BK?
AMD below 10 bucks by Christmas. Once they compare 2.6 Barcelona against Penrym it'll be a joke and AMD will be making firesale material.
AMD will lose another 10 points of MS in serverys by 2008
Post a Comment
<< Home