Sign in with
Sign up | Sign in
Your question

Intel Frauds Woodcrest Performance Over Opteron 64

Last response: in CPUs
Share
May 24, 2006 7:21:18 PM

I changed this a little, hopefully to shut up a few more Intel Fanboys from making pointless posts.

http://www.intel.com/performance/server/xeon/database.h...

That is straight from Intel.com, and they said this about the 2 systems:

Woodcrest - 64-Bit Windows & 64-Bit SQL Server

Opteron - 64-Bit Windows & 64-Bit SQL Server

Intel also linked to this website (The system they referenced of the Opteron) - http://www.tpc.org/tpcc/results/tpcc_result_detail.asp?...

That website says it's an Opteron w/ 32-Bit Windows & 32-Bit SQL Server, not 64-Bit. It appears Intel has changed the data and says it is comparing 2 64-Bit Platforms when in reality, they are comparing a 64-Bit Platform to a 32-Bit Platform and saying it's 2 64-Bit Platforms.

This coincides with Intel's previous lies, where it claims the Xeon Lindenhurst performs above AMD Opteron's when they used 10K HDD's vs. 7.2K HDD's in the Opteron's.

Opteron 64 is typically 20-40% faster in 64-Bit, and those show the Xeon is 33% faster in 64-Bit than the Opteron in 32-Bit, so if you adjust for Tru64 performance, it's really 7% Slower than the Opteron 64's. Remember, this is 20GB/s Bandwidth vs. 9GB/s Bandwidth in the Opteron and it is still winning, just wait until 4P Tests and 64-Bit Socket 1207.

Woodcrest is also known *from Daily Tech and Tech Report* that in Multi-Threaded apps, it is on par with an Opteron 64 while the Opteron 64 has half the bandwidth.

More Intel FUD to throw in the mix.
May 24, 2006 7:27:20 PM

As an Intel Fanboy, I blame Senator Lieberman.


huhuhu
May 24, 2006 7:30:19 PM

Intel compared a 32-bit Opteron to a 64-Bit Woodcrest, even though they say it's a 64-bit Opteron. If you have doubts (Intel Fanboys), check out the .PDF of Full Disclosure, it displays that it is running x86 versions of OS and SQL Server 2005 as it was referenced from, while Intel says referenced from there was 64-bit, even though it says multiple times it is x86.

I am waiting for defendence from Intel Fanboys as to why their great, honorable company lied and manipulated data to better sell a product inferior to the Hammer64.
Related resources
May 24, 2006 7:39:32 PM

If you compare the costs, $2.99 for the Opteron vs. $2.93 for the Woodcrest, it shows Woodcrest in 64-Bit is 2% better in Price/Performance compared to AMD's Opteron 64 in 32-Bit.

No Intel Fanboys on this one? Come on guys, show your support for the fraudalent company!
May 24, 2006 7:39:36 PM

I already told you; it's all Al Gore's fault. Because he's Senator Lieberman's clone. The bastard.


There's a fine line between good marketing and bare-faced lies. That's all I'm going to say for now, since it could just as easily be something along the lines of a typo. Also, I heard The Furniture Guy(TM) is having a sale on queen-sized pillows, that guy is the bomb. Paperclips (gotta be pink) are cooler though.
May 24, 2006 7:41:53 PM

Quote:
I already told you; it's all Al Gore's fault. Because he's Senator Lieberman's clone. The bastard.


There's a fine line between good marketing and bare-faced lies. That's all I'm going to say for now, since it could just as easily be something along the lines of a typo. Also, I heard The Furniture Guy(TM) is having a sale on queen-sized pillows, that guy is the bomb. Paperclips (gotta be pink) are cooler though.


From Intel:

Dual-Core AMD Opteron* Processor Model 285 based platform details: HP Proliant DL385 G1* server platform with two Dual-Core AMD Opteron* processor 2.60GHz, 32GB memory, Microsoft Windows Server* 2003 Enterprise x64 Edition, Microsoft SQL Server* 2005 Enterprise x64 Edition. Referenced as published at 113,628 tpmC; $2.99/tpmC; Availability Date as listed in the submitted report is May 5, 2006. Results at http://www.tpc.org/tpcc/results/tpcc_result_detail.asp?...

On that website, not to mention the documents, it says BLATENTLY, that it is 32-Bit. So, either 1) Intel Lied, or 2) They are a bad company because they do not clsoely look at data and just put what they think it is.
May 24, 2006 7:43:11 PM

Quote:
Intel compared a 32-bit Opteron to a 64-Bit Woodcrest, even though they say it's a 64-bit Opteron. If you have doubts (Intel Fanboys), check out the .PDF of Full Disclosure, it displays that it is running x86 versions of OS and SQL Server 2005 as it was referenced from, while Intel says referenced from there was 64-bit, even though it says multiple times it is x86.

I am waiting for defendence from Intel Fanboys as to why their great, honorable company lied and manipulated data to better sell a product inferior to the Hammer64.


Opteron runs 32bit code just as well as 64bit code as does Woodcrest there isnt a 32bit only Opteron... Now if you mean the OS they had installed and drivers where 32bit on the Opteron side and the OS and Drivers where 64Bit on the Woodcrest side I would agree something is VERY fishy.
May 24, 2006 7:45:24 PM

Quote:
As an Intel Fanboy, I blame Senator Lieberman.


huhuhu
THAT BASTARD!

Is he such an ass**** that he even has to interfere with the last free area of geekdom?(He f'ed with us in videogames,music,movies,ete etc now he's going after computers!)
May 24, 2006 7:46:10 PM

Quote:
Intel compared a 32-bit Opteron to a 64-Bit Woodcrest, even though they say it's a 64-bit Opteron. If you have doubts (Intel Fanboys), check out the .PDF of Full Disclosure, it displays that it is running x86 versions of OS and SQL Server 2005 as it was referenced from, while Intel says referenced from there was 64-bit, even though it says multiple times it is x86.

I am waiting for defendence from Intel Fanboys as to why their great, honorable company lied and manipulated data to better sell a product inferior to the Hammer64.


Opteron runs 32bit code just as well as 64bit code as does Woodcrest there isnt a 32bit only Opteron... Now if you mean the OS they had installed and drivers where 32bit on the Opteron side and the OS and Drivers where 64Bit on the Woodcrest side I would agree something is VERY fishy.

By 32-bit, I meant it is an Opteron 64 w/ 32-Bit Windows Server and 32-Bit SQL Server vs. a Woodcrest w/ 64-Bit Windows and 64-Bit SQL Server, but Intel says it's Opteron 64 w/ 64-Bit and it isn't.

Intel has lied, knowingly.
May 24, 2006 7:51:27 PM

Quote:
As an Intel Fanboy, I blame Senator Lieberman.


huhuhu
THAT BASTARD!

Is he such an ass**** that he even has to interfere with the last free area of geekdom?(He f'ed with us in videogames,music,movies,ete etc now he's going after computers!)

Hahahahaha !!! I heard he wants to team up with Al Gore to "invent" Internet 2 !!!
May 24, 2006 7:54:38 PM

It is your post that is a fraud. I'm sorry you don't like the results, but thems the facts. There is a typo in one the descriptions but the scores for all of the systems are real, and the Woodcrest score and description in particular are quite accurate.

Regardless of server configuration, 113628 is the best TCP result AMD and HP have been able to produce... and Woodcrest decimates it with a 169360.

This whole thread is silly and should be locked.
May 24, 2006 7:54:50 PM

You've been banned once for trolling ... do we have to do this again Mike?
May 24, 2006 7:55:56 PM

So at worst either Intel made a typo or HP improperly reported their OS. Both scores are genuine, and the HP DL385 still has the highest TPC-C score for an Opteron based 2S system and Woodcrest easily beats it.
May 24, 2006 7:58:14 PM

Quote:
It is your post that is a fraud. I'm sorry you don't like the results, but thems the facts. There is a typo in one the descriptions but the scores for all of the systems are real, and the Woodcrest score and description in particular are quite accurate.

Regardless of server configuration, 113628 is the best TCP result AMD and HP have been able to produce... and Woodcrest decimates it with a 169360.

This whole thread is silly and should be locked.
'

I'm sorry you are stupid, but let me fix you.

DL385 G1 - From Intel.com

Intel linked to this website: http://www.tpc.org/tpcc/results/tpcc_result_detail.asp?...

Which is the SAME SYSTEM.

So, what it is YOU who is the idiot, and I am not trolling Jake, I am saying facts.
May 24, 2006 8:00:12 PM

You're saying it's fair to compare 32-Bit to 64-Bit and say "64-Bit is better" but at the same time, lie and say it's 64-Bit vs. 64-Bit when it ISN'T? You're also saying it's impossible for HP to put on a 64-Bit OS? Wow Itty, you're a genious!
May 24, 2006 8:06:36 PM

Quote:
You're saying it's fair to compare 32-Bit to 64-Bit and say "64-Bit is better" but at the same time, lie and say it's 64-Bit vs. 64-Bit when it ISN'T? You're also saying it's impossible for HP to put on a 64-Bit OS? Wow Itty, you're a genious!

what would intel gain from that(falsifying a comparative test)???honestly? amd is going to be behind for a while magic.i think its just a typing error.

You are truly an Intel Fanboy who is ignorant.

Intel has plenty gain, such as market share!

AMD is already ahead of NGMA and will never be behind, sorry.

This is not a typo, it proves 1 of 2 things:

1) Intel lied (more likely)

2) Intel did not read the website about the data and assumed it was 64-Bit and just took the numbers

Either way, Intel is not honest (wow, something new).
May 24, 2006 8:14:02 PM

Good to see you haven't killed yourself over the benchmarks MMM.
May 24, 2006 8:14:04 PM

Quote:
As an Intel Fanboy, I blame Senator Lieberman.


huhuhu
THAT BASTARD!

Is he such an ass**** that he even has to interfere with the last free area of geekdom?(He f'ed with us in videogames,music,movies,ete etc now he's going after computers!)

Hahahahaha !!! I heard he wants to team up with Al Gore to "invent" Internet 2 !!! I heard that they struck a deal with the telco's in order to make us pay more money! All i can say is



THOSE COMMIE BASTARDS!
May 24, 2006 8:16:30 PM

I give accordance to Mistakes, not Competance.

Intel is incompetant and this proves it. Nobody is this stupid and cannot read X86 and puts down X64 and thinks it's a mistake, if they had read the entier documents that gave the numbers, or even looked 10" down on their screens and saw X86, they'd know the truth.

The truth is: Intel lied....Again.
May 24, 2006 8:18:54 PM

Quote:
As an Intel Fanboy, I blame Senator Lieberman.


huhuhu
THAT BASTARD!

Is he such an ass**** that he even has to interfere with the last free area of geekdom?(He f'ed with us in videogames,music,movies,ete etc now he's going after computers!)

Hahahahaha !!! I heard he wants to team up with Al Gore to "invent" Internet 2 !!!


God Gates invented Internet 2 already!
May 24, 2006 8:21:27 PM

Quote:

I own AMD and plan to upgrade to AMD sooo,i think i am bieng pretty realistic,and giving consideration for human err,if intel used a 32bit amd in the test vs a 64bit intel,wouldnt that anger investors and accomplish the reverse of generating sales exitement?????????

The tests were done by the manufacturer, in this case HP.
May 24, 2006 8:22:35 PM

Quote:
It is your post that is a fraud. I'm sorry you don't like the results, but thems the facts. There is a typo in one the descriptions but the scores for all of the systems are real, and the Woodcrest score and description in particular are quite accurate.

Regardless of server configuration, 113628 is the best TCP result AMD and HP have been able to produce... and Woodcrest decimates it with a 169360.

This whole thread is silly and should be locked.
'

I'm sorry you are stupid, but let me fix you.

DL385 G1 - From Intel.com

Intel linked to this website: http://www.tpc.org/tpcc/results/tpcc_result_detail.asp?...

Which is the SAME SYSTEM.

So, what it is YOU who is the idiot, and I am not trolling Jake, I am saying facts.
I know they are the same system.

As I said above, there is a typo in one of the descriptions. The scores are still correct. Woodcrest wins big time, I know it must really hurt your fanboy heart.
May 24, 2006 8:23:16 PM

Quote:
As an Intel Fanboy, I blame Senator Lieberman.


huhuhu
THAT BASTARD!

Is he such an ass**** that he even has to interfere with the last free area of geekdom?(He f'ed with us in videogames,music,movies,ete etc now he's going after computers!)

Hahahahaha !!! I heard he wants to team up with Al Gore to "invent" Internet 2 !!!


God Gates invented Internet 2 already!

I thought you'd have said Lord Linus :lol: 
May 24, 2006 8:24:13 PM

Quote:
IBTL !
What does that mean?
May 24, 2006 8:26:08 PM

Quote:
As an Intel Fanboy, I blame Senator Lieberman.


huhuhu
THAT BASTARD!

Is he such an ass**** that he even has to interfere with the last free area of geekdom?(He f'ed with us in videogames,music,movies,ete etc now he's going after computers!)

Hahahahaha !!! I heard he wants to team up with Al Gore to "invent" Internet 2 !!!


God Gates invented Internet 2 already!

I thought you'd have said Lord Linus :lol: 


Oh no, God Gates is the one true god!
May 24, 2006 8:28:11 PM

Hey MMM I thought since intel 64 bit was so inferior, does it really matter if the AMD system was 32bit? I guess it does since it got its ass kicked.
May 24, 2006 8:29:26 PM

Quote:
It is your post that is a fraud. I'm sorry you don't like the results, but thems the facts. There is a typo in one the descriptions but the scores for all of the systems are real, and the Woodcrest score and description in particular are quite accurate.

Regardless of server configuration, 113628 is the best TCP result AMD and HP have been able to produce... and Woodcrest decimates it with a 169360.

This whole thread is silly and should be locked.
'

I'm sorry you are stupid, but let me fix you.

DL385 G1 - From Intel.com


Intel linked to this website: http://www.tpc.org/tpcc/results/tpcc_result_detail.asp?...

Which is the SAME SYSTEM.

So, what it is YOU who is the idiot, and I am not trolling Jake, I am saying facts.
I know they are the same system.

As I said above, there is a typo in one of the descriptions. The scores are still correct. Woodcrest wins big time, I know it must really hurt your fanboy heart.

You're being too stubborn.

There is no typo, this is Comparing a 64-Bit to 32-Bit, so regardless of the numbers, and if you adjust for 64-bit in AMD, the AMD is 7% faster.

The only fanboy heart broke is yours, because you're so obsessed with Conroe's fake performance that it's blinding you. Open your eyes son, and see what the world is really about.
May 24, 2006 8:32:59 PM

DIE TRAITOR!!!!
May 24, 2006 8:39:36 PM

Quote:
Hey MMM I thought since intel 64 bit was so inferior, does it really matter if the AMD system was 32bit? I guess it does since it got its ass kicked.


The Opteron is 7% faster if they compare 64-bit to 64-bit from NGMA. Who's ass got kicked? Oh ya, your Intel Fanboy ass did.

Not sure why you're all defending Intel so much, aren't you the same genious's who said "I'm a fan of performance"? So why the hostility when I prove Intel wrong? Isn't this a good thing according to you people? Or is an AMD Innovation a bad thing?
May 24, 2006 8:41:35 PM

Quote:
IBTL !
What does that mean?


if he told you youd have to disappear into the superconciousness for knowing too much. But then how is he still here?
May 24, 2006 8:44:16 PM

Quote:
and if you adjust for 64-bit in AMD, the AMD is 7% faster... [some other crap]

LOL, what adjust it by doubling its score? You are seriously grasping at straws....
May 24, 2006 8:49:19 PM

Quote:
and if you adjust for 64-bit in AMD, the AMD is 7% faster... [some other crap]

LOL, what adjust it by doubling its score? You are seriously grasping at straws....

Woap! You got me Itty, your genious never fails to impress!
May 24, 2006 8:49:38 PM

Quote:

There is no typo, this is Comparing a 64-Bit to 32-Bit, so regardless of the numbers, and if you adjust for 64-bit in AMD, the AMD is 7% faster.

How'd you get that figure? A 7% advantage would give the DL385 a score of around 180K tpcM, which is close to the scores obtained by the 4S/8P DL585. In reality for TPC-C 64-bit doesn't give a significant boost as we can see from HP's submission for the DL585.

http://tpc.org/tpcc/results/tpcc_result_detail.asp?id=1...
4 Opteron 2.2DC/64GB RAM/ SQL Server 2000/Windows 2003 x86

187,296 tpcM

http://tpc.org/tpcc/results/tpcc_result_detail.asp?id=1...
This DL585 uses faster 2.4DC Opterons, doubles the memory to 128GB and uses SQL Server 2005 x64/Windows 2003 x64. It scores:

206,181 tpcM.

An increase of roughly 10%.
May 24, 2006 8:53:55 PM

Quote:
I changed this a little, hopefully to shut up a few more Intel Fanboys from making pointless posts.

Wazzzup MMM? When are you going to stop bullsh*ting around?
Hey, what do you think about my rig?
I am replacing it, whenever Conroe comes. Do you think it can fit in my old can?
May 24, 2006 8:54:35 PM

If you compare AMD's 64-Bit performance, it can gain avg. 40% in 64-bit over 32-bit, and that Woodcrest in 64-bit is 33% faster than the Opteron 32-bit, and it's a 64-bit w/ 64GB RAM vs. 32-bit w/ 32GB RAM, running a benchmark that increases score w/ more RAM.

If you do simple math (you're all capable of that, right?) you see AMD w/ a 7% advantage.
May 24, 2006 8:55:16 PM

Quote:
Hey MMM I thought since intel 64 bit was so inferior, does it really matter if the AMD system was 32bit? I guess it does since it got its ass kicked.


The Opteron is 7% faster if they compare 64-bit to 64-bit from NGMA. Who's ass got kicked? Oh ya, your Intel Fanboy ass did.

Not sure why you're all defending Intel so much, aren't you the same genious's who said "I'm a fan of performance"? So why the hostility when I prove Intel wrong? Isn't this a good thing according to you people? Or is an AMD Innovation a bad thing?


MMM I am a fan of performance, I just couldn't resist the urge to throw a little gasoline on your fire. I must admit, man, you are quite humorous getting so worked up, about to have some massive coronary over this stuff.

Just chill out man. If it works out that you are right, I will be the first in line to give you credit for your call. But seriously, why would intel knowingly falsify this kind of data ? They know somebody is gonna catch it. Even if it was a mistake, I would label it more than likely an honest mistake (Even though it nearly gave you a heart attack)
May 24, 2006 8:58:30 PM

Funny thing is all you Intel fanboys ganging up on me (remind us of the old days?) and it's when I'M RIGHT (still can't find a time when I wasn't though). I bet if Intel said publicly they were wrong, you still would think "it's an honest mistake" when saying "honest" and "Intel" in the same sentence is as funny as Conroe being a good processor, LOL!

Accord, you're comparing SQL 2000 to SQL 2005 and 32-bit to 64-bit, something that should not be done.
May 24, 2006 9:03:11 PM

Quote:

Accord, you're comparing SQL 2000 to SQL 2005 and 32-bit to 64-bit, something that should not be done.

Why not? The second DL585 gets faster CPU, more memory, faster DB and 64-bit and can barely muster a 10% increase and you still think the DL385 will be able to get a 40% increase in tpcM just from 64-bit.
May 24, 2006 9:06:19 PM

How about the fact both those systems have PC2700 RAM and using 4GB Modules, which means slower bandwidth and higher latencies.
May 24, 2006 9:07:47 PM

MMM if you haven't already done so, I would recommend that you report this misinformation to intel using the link below. This would at least give them an opportunity to fix their error.

http://www.intel.com/intel/report.htm


Secondly, I don't think it is fair of you to judge an entire company by what was most likely an honest mistake of a single individual (I could very easily see myself making this kind of error).

You act like intel has some great big conspiracy to ruin your life.
I bet Hector Ruiz didn't get half as worked up over this as you did.
May 24, 2006 9:08:28 PM

Dude how can u call someone a fanboy when YOU have a conroe avatar.
LOL
May 24, 2006 9:11:40 PM

Quote:
Funny thing is all you Intel fanboys ganging up on me (remind us of the old days?) and it's when I'M RIGHT (still can't find a time when I wasn't though). I bet if Intel said publicly they were wrong, you still would think "it's an honest mistake" when saying "honest" and "Intel" in the same sentence is as funny as Conroe being a good processor, LOL!

Accord, you're comparing SQL 2000 to SQL 2005 and 32-bit to 64-bit, something that should not be done.


Calm down man, I'm sure someone else has already got the benches for the opteron that you are looking for. It's a matter of comparing them to what intel got. This reminds of when intel setup a conroe against an OC FX60 at the IDF...every amd fan was claiming bloody murder and that the FX60 was given a disadvantage. :wink:
May 24, 2006 9:12:51 PM

Quote:
How about the fact both those systems have PC2700 RAM and using 4GB Modules, which means slower bandwidth and higher latencies.

A limitation of Opterons. 16GB is about the max you can expect for PC3200, you want more memory/DIMMs you have to use slower speeds. For TPC-C, and most enterpise databases, memory capacity beats memory speeds.

Besides, it doesn't matter since you claim that Opteron should run 40% faster going into 64-bit mode. Why isn't the second DL585 40%, or even 25% faster?
May 24, 2006 9:17:22 PM

You're trying to compare SQL 2000 32-Bit to SQL 2005 64-Bit....this tells me that SQL 2005 is 10% faster than SQL 2000, nothing for the CPU's.
May 24, 2006 9:20:24 PM

Quote:
You're trying to compare SQL 2000 32-Bit to SQL 2005 64-Bit....this tells me that SQL 2005 is 10% faster than SQL 2000, nothing for the CPU's.

It tells me that for TPC-C, Opteron going from 32-bit to 64-bit has little impact and that it has no hope of catching Woodcrest at 2S.
May 24, 2006 9:24:29 PM

Quote:
You're trying to compare SQL 2000 32-Bit to SQL 2005 64-Bit....this tells me that SQL 2005 is 10% faster than SQL 2000, nothing for the CPU's.

It tells me that for TPC-C, Opteron going from 32-bit to 64-bit has little impact and that it has no hope of catching Woodcrest at 2S.

And it tells you that because you're a moron.
May 24, 2006 9:26:12 PM

Quote:
It is your post that is a fraud. I'm sorry you don't like the results, but thems the facts. There is a typo in one the descriptions but the scores for all of the systems are real, and the Woodcrest score and description in particular are quite accurate.

Regardless of server configuration, 113628 is the best TCP result AMD and HP have been able to produce... and Woodcrest decimates it with a 169360.

This whole thread is silly and should be locked.


How do you know that it is just an innocent typo? Furthermore, how do you know the results are valid? Who signs your pay check-Intel??? :roll:
May 24, 2006 9:29:42 PM

Quote:
You're trying to compare SQL 2000 32-Bit to SQL 2005 64-Bit....this tells me that SQL 2005 is 10% faster than SQL 2000, nothing for the CPU's.

It tells me that for TPC-C, Opteron going from 32-bit to 64-bit has little impact and that it has no hope of catching Woodcrest at 2S.

I just found out more info to justify me saying you're a moron.

System 1:

32-Bit Opteron system shown at Intels website is servicing 4 clients and 8 Processors

System 2:

64-Bit Opteron system shown in your posts is servicing 8 clients and 16 processors

I should also note the 2.2GHz Quad-Socket system is using Version 5.3, the other Quad-Socket 2.4GHz is using Version 5.5 while the one Intel displays is using Version 5.6.

That's why there isn't a large increase.
May 24, 2006 9:32:33 PM

Quote:

And it tells you that because you're a moron.

The moron would be the one who added 40% to the score of the DL385 because Opterons run 40% faster in every application while in 64-bit mode and proudly claimed that it would be 7% faster than the ML370 with no evidence whatsoever.
!