Tom's Hardware > Forum > CPU & Components > CPUs > EIGHT CORE SERVERS: OPTERON 880 vs. XEON MP 3.0 GHZ

EIGHT CORE SERVERS: OPTERON 880 vs. XEON MP 3.0 GHZ

Forum CPU & Components : CPUs - EIGHT CORE SERVERS: OPTERON 880 vs. XEON MP 3.0 GHZ

Tom's Hardware: Over 1.4 million members in 6 different countries available to answer all your high-tech questions. Sign up now! Its free!
Page:    Previous 1 2 Next Bottom Search this thread
Word :    Username :           
 

Now this is what a call a grand rape from Opteron.
I hope Intel uses vaseline the next time. :wink:

http://www.anandtech.com/IT/showdoc.aspx?i=2745

Quote :

The Opteron leads the way in every test, which isn't surprising. As we've said in the past the current Xeon is hindered by the front-side bus. Trying to shove a giant beach ball into a basket ball hoop - with thirty other people waiting to try the same thing - has a serious impact on scalability. All of the tests are fairly consistent, and the Opteron averages a 36-51% lead over the Paxville system.


LOL. 8)

Sponsored Links
Register or log in to remove.

Hey Action: Did your mom recovered from her hymenoplasty procedure??
Could she get it fix?
I'm kinda concerned since many on the forum got keyboards from her. I imagine the though labor those doctors had to go through. 8)

Reply to 9-inch

Quote :

Hey Action: Did your mom recovered from her hymenoplasty procedure??
Could she get it fix?
I'm kinda concerned since many on the forum got keyboards from her. I imagine the though labor those doctors had to go through. 8)



Hey Action, did your mom recover from her hymanoplasty procedure?
Could she get it fixed? I'm kinda' concerned since many on the forum got
keyboards from her. I can imagine the labor those doctors must of had to
go through.



Less links, more practice at english.

Reply to KingLoftusXII

A Newtonian reflector telescope uses a parabolic mirror instead of a lense to gather and focus light. This light is focused to a point near the front of the telescope tube and is then redirected to the side of the tube with an elliptical diagonal mirror at a 45 degree angle. The focuser is mounted at the front of this tube and holds the eyepiece which further magnifies this focused light from the parabolic primary.

Newtonian reflectors, aperature per aperature, are far more affordable than a comparable refractor telescope, and are not plagued by the effects of chromatic aberration (false color)

Reply to gr8mikey

Quote :

9-inch, my pizzle lovin' buddy, great post!!! Wow, dazzling, simply dazzling.

Say, I was wondering since you have established yourself as an expert in the field of semiconductors if you could solve me a puzzle. It should not be too difficult, one that should whip right off the top of your head (pardon the pun).

Ok, here it goes --

I was thumbing through some sites on the internet trying to educate myself to your level. A damn near impossible task, by the way. Anyway, I came across this article that says they expose the wafer to nitrous oxide (I think the formula is N2O, not entirely sure). They expose it after making something called a gate. If I am not mistaken this is laughing gas, why on earth would they do this? Are they trying to make the wafer laugh?

Hmmmm, puzzling, simpling puzzling...but I have faith, I have faith that our own 9-inch can give us the answer. Butterflies tickle in my stomach as I, no WE, all await your reply.

Jack



Nitrous Oxide is used in the semiconductor mfg process? I did not know that! Come on 9-Inch! Bestow upon us this great pearl of wisdom! A response to this question will really help me sleep tonight.

Reply to gr8mikey

Quote :

9-inch, my pizzle lovin' buddy, great post!!! Wow, dazzling, simply dazzling.

Say, I was wondering since you have established yourself as an expert in the field of semiconductors if you could solve me a puzzle. It should not be too difficult, one that should whip right off the top of your head (pardon the pun).

Ok, here it goes --

I was thumbing through some sites on the internet trying to educate myself to your level. A damn near impossible task, by the way. Anyway, I came across this article that says they expose the wafer to nitrous oxide (I think the formula is N2O, not entirely sure). They expose it after making something called a gate. If I am not mistaken this is laughing gas, why on earth would they do this? Are they trying to make the wafer laugh?

Hmmmm, puzzling, simpling puzzling...but I have faith, I have faith that our own 9-inch can give us the answer. Butterflies tickle in my stomach as I, no WE, all await your reply.

Jack

Lol, you scrutinize him so 8) .

Reply to Caboose-1

9-inch dont listen to action man he's just playing with you :lol:

Reply to dvdpiddy

Quote :

sphincter

Paul lol "SPHINCTER" :lol:

Reply to dvdpiddy

Quote :

Yeah, some companies will use N2O or NH3 or some other nitrogen containing molecule and "nitride" the gate, sometimes called an oxynitride. It has the affect of making a better SiO2 to Si interface.



N20 is used during tunnel odixation not the gate ox. For the gate step, you actually use good old hydrogen and oxygen gases, trick here is that it has to be reacted in a torch chamber, with the correct ratio of course, by the way when the gases are reacted in the torch, they make a pretty blue flame that flows into a 950C reactor tube. NH3 is ammonia which we use to make a nitride deposition, not an ox, but some other places might use it for ox. Actually we use much more interesting chemicals, such as HCL (a nasty acid we use during one of our dep steps) and DCS (gas ignites when exposed to oxygen). Oxynitride??? never heard of that process, is it a deposition or oxydation???

Reply to qurious69ss

Alas, I see my job is already done. :lol:

Newtonian Telescope...is that in comparison to a Galilean Telescope?

Reply to joefriday

Quote :

Yeah, some companies will use N2O or NH3 or some other nitrogen containing molecule and "nitride" the gate, sometimes called an oxynitride. It has the affect of making a better SiO2 to Si interface.



N20 is used during tunnel odixation not the gate ox. For the gate step, you actually use good old hydrogen and oxygen gases, trick here is that it has to be reacted in a torch chamber, with the correct ratio of course, by the way when the gases are reacted in the torch, they make a pretty blue flame that flows into a 950C reactor tube. NH3 is ammonia which we use to make a nitride deposition, not an ox, but some other places might use it for ox. Actually we use much more interesting chemicals, such as HCL (a nasty acid we use during one of our dep steps) and DCS (gas ignites when exposed to oxygen). Oxynitride??? never heard of that process, is it a deposition or oxydation???


Are you refering to an Intel process ? Intel does not use oxynitrides for their gate material, they use plain old SiO2 and do not nitride it. :) ...

However, forming gate oxynitride is used in the industry. Here is a quick applied phys. lett. reference. Again, it is a subscription site but you can get the jist from the abstract.

http://scitation.aip.org/getabs/se [...] s&gifs=yes

The reference I site above is studying Nitric Oxide for a gate material, it does reference N2O for the most common oxynitride producer. I could pull out more references if you wish. This particular reference was published by, dun dun dahhhhh - Motorola. :)

Hey, look, a 9-inch thread that is actually producing some informative discussion--go figure.

EDIT: Not that you are incorrect, in fact, you are quite on the money however there are more than just a few uses for N2O :) What the heck, here is another one: http://ieeexplore.ieee.org/xpl/abs [...] ber=237142

Ahhhh crap, here's another one, this from IBM: http://www.research.ibm.com/journal/rd/433/ellis.html

Jack

I will look at those articles as soon as I get a chance......thanks jack, by the way I'm not a process expert, my job is to make sure the equipment is operating correctly at the right temps, gas flows, and pressure, and of course the damn automation system. Anyways back to work.

Reply to qurious69ss

are there even moderators on this forum?

Reply to Legenic

9 inch may not have givien us a great post. He has given us a simple reminder.
In the real world, the best chips available today, for desktop and server aplications, are made by AMD.
It seems unlikely that Intel will be able to catch 8 way opteron for the forseable future. That is the reality. There is no smoke nor mirrors involved.

Reply to endyen

Cough *woodcrest* cough

Quote :

Now this is what a call a grand rape from Opteron.
I hope Intel uses vaseline the next time. :wink:

http://www.anandtech.com/IT/showdoc.aspx?i=2745

The Opteron leads the way in every test, which isn't surprising. As we've said in the past the current Xeon is hindered by the front-side bus. Trying to shove a giant beach ball into a basket ball hoop - with thirty other people waiting to try the same thing - has a serious impact on scalability. All of the tests are fairly consistent, and the Opteron averages a 36-51% lead over the Paxville system.


LOL. 8)

Reply to apache_lives

Woodcrest isnt out yet and it'll have a hard time in 8 way configs with the aging fsb. :roll:

Reply to Action_Man

Quote :

Cough *woodcrest* cough



*COUGH* *COUGH*
THG's Conclusion

"At present, Opteron has a commanding lead in 8-way configurations. Looking at the prices of the processors we used, AMD has the advantage there as well. The Xeon MP 3.0 GHz processors currently cost upwards of $3000 each, whereas Opteron 880 are about $1000 less. Of course, for enterprise servers price is less of a concern; performance, scalability, and reliability take precedence. Reliability might be a tie, but that's about as close as Intel's Xeon MP currently comes to Opteron."


Ok I'll give you that it isnt woodcrest but this time next year AMD will have quadcore 4 socket systems with higher HTT 3.0 specs and FB-DIMM's. Not to mention all of the architectural improvements with K8L. AM2 will supposedly give us quad core as well. Think Intel will make you buy a new motherboard for quad core :roll:
You bet your ass they will.

Reply to K8MAN

Didnt so much mean it would ace it but competition would be better - Pentium M's do well with the old FSB and woodcrest with DIB will do ok till ~2008 with CSI or whatever.

What i really mean is ANYTHING IS BETTER THEN P4.

Quote :

Woodcrest isnt out yet and it'll have a hard time in 8 way configs with the aging fsb. :roll:

Reply to apache_lives

Quote :

"At present, Opteron has a commanding lead in 8-way configurations. Looking at the prices of the processors we used, AMD has the advantage there as well. The Xeon MP 3.0 GHz processors currently cost upwards of $3000 each, whereas Opteron 880 are about $1000 less. Of course, for enterprise servers price is less of a concern; performance, scalability, and reliability take precedence. Reliability might be a tie, but that's about as close as Intel's Xeon MP currently comes to Opteron."



What does that have to do with woodcrest?

Quote :

Ok I'll give you that it isnt woodcrest but this time next year AMD will have quadcore 4 socket systems with higher HTT 3.0 specs and FB-DIMM's. Not to mention all of the architectural improvements with K8L. AM2 will supposedly give us quad core as well. Think Intel will make you buy a new motherboard for quad core Rolling Eyes
You bet your ass they will.



Intel isn't going to move forward either? :roll:

Reply to Action_Man

Quote :

"At present, Opteron has a commanding lead in 8-way configurations. Looking at the prices of the processors we used, AMD has the advantage there as well. The Xeon MP 3.0 GHz processors currently cost upwards of $3000 each, whereas Opteron 880 are about $1000 less. Of course, for enterprise servers price is less of a concern; performance, scalability, and reliability take precedence. Reliability might be a tie, but that's about as close as Intel's Xeon MP currently comes to Opteron."



What does that have to do with woodcrest?

Quote :

Ok I'll give you that it isnt woodcrest but this time next year AMD will have quadcore 4 socket systems with higher HTT 3.0 specs and FB-DIMM's. Not to mention all of the architectural improvements with K8L. AM2 will supposedly give us quad core as well. Think Intel will make you buy a new motherboard for quad core Rolling Eyes
You bet your ass they will.



Intel isn't going to move forward either? :roll:

You took it the wrong way. This time next year AMD will have a platform that is already 40% ahead + all of those improvements.
Conroe is the savior of the dual core world yes because its mighty damn fast and better than anything AMD will have for dual core for @ least a year but because it fits into boards already available on the market designed for netburst already tells me that it was never designed for scaling past 1 core. Intel wanted to go IA64 for MC CPU's and all this sharing FSB DC and soon to be MC solution on SCKT 775 is a hack to keep up with something that was designed for 8 core operation from the getgo.

And the same can be said about anything bearing the name "Xeon"

Reply to K8MAN

Quote :

What i really mean is ANYTHING IS BETTER THEN P4.



Goes without saying.

Quote :

This time next year AMD will have a platform that is already 40% ahead



...of?

Quote :

but because it fits into boards already available on the market designed for netburst already tells me that it was never designed for scaling past 1 core.



Netburst? Well yeah.

Reply to Action_Man

Quote :

Xeon MP 3.0 GHz processors currently cost upwards of $3000 each, whereas Opteron 880 are about $1000 less


You forget, Intel is going to do a major price cut, on April 23rd. (looks at date) Yup the 23rd is the day to wait for. Intel pricing, mmmm, good honest value.

Reply to endyen

What AMD needs to do is make the z-ram they will be adding as l3 to K8L as efficient in design as Core's L2 as that IMO is the whole secret behind Conroe and the lot. Its a great architecture for sure but just slapping that cache onto Dothan improved superpi times and numerous other benches by gigantic margins.

Reply to K8MAN

Quote :

Core's L2 as that IMO is the whole secret behind Conroe and the lot.



:roll: :roll: :roll:

Quote :

Its a great architecture for sure but just slapping that cache onto Dothan improved superpi times and numerous other benches by gigantic margins.



Banias was already good and Dothan also got a good 400mhz speed bump over Banias whilst using less power.

Reply to Action_Man

Quote :

Netburst? Well yeah.


Exactly they are slapsticking this new chip onto the netburst PLATFORM as Intel never wanted to go multicore x86. The P4 was supposed to be a single beast @ 10ghz someday. AMD has had the platform in place for years now. It wouldnt suprise me the least to see AMD realease some lower-end quad-core chips for socket 940 if it will fit into the thermal specs. And with AMD's .065nm chips supposedly only chewing 1.1V compared to Intel's 1.15 Conroe and alleged o/c'es on 3.8ghz on air with dual core........... :wink:

Reply to K8MAN

Quote :

Exactly they are slapsticking this new chip onto the netburst PLATFORM as Intel never wanted to go multicore x86.



So? Conroe still wins.

Reply to Action_Man

But Yonah is what? Dothan with Intel's NGMA Unified cache thrown on. It improved yonah in super pi by something like 6 seconds clock/clock. Yonah is not a 4 issue core as far as i know so what else would make crunching a certain well known figure to 1 million decimals so much faster all of a sudden? I know it wasnt magic.

Reply to K8MAN

Quote :

Exactly they are slapsticking this new chip onto the netburst PLATFORM as Intel never wanted to go multicore x86.



So? Conroe still wins.
You obviously read very thoroughly. I already told you conroe will eat anything else alive for the next 6-12 months in dual core as the platform limitations dont start showing up until the bandwidth starts choking with 4core. Conroe is hungry for bandwidth and will be even more so when code is optomised for it.

Reply to K8MAN

Quote :

But Yonah is what? Dothan with Intel's NGMA Unified cache thrown on. It improved yonah in super pi by something like 6 seconds clock/clock. Yonah is not a 4 issue core as far as i know so what else would make crunching a certain well known figure to 1 million decimals so much faster all of a sudden? I know it wasnt magic.



SSE got a lot of improvements, there's probably other things but details are thin.

Reply to Action_Man

Quote :

You obviously read very thoroughly. I already told you conroe will eat anything else alive for the next 6-12 months in dual core as the platform limitations dont start showing up until the bandwidth starts choking with 4core. Conroe is hungry for bandwidth and will be even more so when code is optomised for it.



I did but it was mostly boring. The platform will improve and they will get CSI in like 2008. 45nm should help as well to do proper quad core and increase cache sizes.

Reply to Action_Man

Intel is gonna have to start using a superscaler pipelined northbridge soon too keep a 16 core conroe system fed beyond the point of near starvation with that ol' northbridge.

Reply to K8MAN

Quote :

You obviously read very thoroughly. I already told you conroe will eat anything else alive for the next 6-12 months in dual core as the platform limitations dont start showing up until the bandwidth starts choking with 4core. Conroe is hungry for bandwidth and will be even more so when code is optomised for it.



I did but it was mostly boring. The platform will improve and they will get CSI in like 2008. 45nm should help as well to do proper quad core and increase cache sizes.
So in otherwords they will be using their manufacturing muscle(made possible by years of anticompetative buisness practises as proven already in the japan suit) to make up for their platforms shortcomings?

Reply to K8MAN

Seriously dont get me wrong i'm all for conroe as its gonna be killer for DC chips will will tide most people over for @ least another few years. AMD is gonna be very competative in the superhighend quad-core and even more so in the server area even with what they've got right now.

Reply to K8MAN

Quote :

So in otherwords they will be using their manufacturing muscle



You'd prefer they did nothing instead?

Quote :

(made possible by years of anticompetative buisness practises as proven already in the japan suit)



Yes, that is the sole reason. :roll:

Quote :

to make up for their platforms shortcomings?



Till CSI, yeah. Again you'd prefer they do nothing at all? What would you prefer they do?

Reply to Action_Man

Wusy have you seen the performance of MM over @ XS's dual socket dual core opty's o/ced to over 3.2ghz? We'll have that from 1 chip sooner that anyone thinks. And any home user that relly wants to show off something big could simply get a dual socket quad core K8L setup that'll surely spank a conroe box.

Reply to K8MAN

Quote :

Now this is what a call a grand rape from Opteron.
I hope Intel uses vaseline the next time.



Intel will finally have something to sue AMD over now... Statutory Rape... I hear it's a good 5-7 years in prison for that shit :lol:

Reply to knightrous
Previous
1 2
Tom's Hardware > Forum > CPU & Components > CPUs > EIGHT CORE SERVERS: OPTERON 880 vs. XEON MP 3.0 GHZ
Go to:

There are 654 identified and unidentified users. To see the list of identified users, Click here.

Please mind

You are about to answer a thread that has been inactive for more than 6 months.
If you still wish to proceed, please ensure that your posting is original and does not duplicate or overlap any prior responses to this thread.

Add a reply Cancel
Sponsored links
  • Ask the community now
  • Publish
Ad
They won a badge
Join us in greeting them