Nehalem-EP......BLOOMFIELD

Quote:

---

Originally Posted by informal

Yep strange benchmark,but it ends quickly, at least :).
Anyone care to comment what system scored 6s in wprime?I see you people are talking about 2Ghz 2P Geinstown but all i see is 2x3.07Ghz system in that screenie (page 30) with power saving on...

---

I wondered about that too :up:

Yes, it was a 3GHz system, I was looking at NB clocks :D

I have compared most of the benchmarks posted in this thread. I hope I am not offending anyone by borrowing their pics.

Listing the benchmark results by best first - worst last:

EDIT: REMOVED BECAUSE OF HURT FEELINGS.

BTW: I felt my benches run smoother than what they would have done on an Intel system. :yepp::p:

next

EDIT: Look above..

next again

EDIT: look at post 804

final

Edit: See post 804

some explanations:

I wanted to compare 8 Nehalem cores (HT imagined or not, they are listed as cores by apps) with 8 lower clocked Barcelona cores on dual socket (L1N64WS/B) with 4gigs of reg ram under Vista x64. I included the higher clocked qx benches to compare as I find it interesting :)

BTW: The third truecrypt pic and the last pic is of a QX6700@3GHz iirc

Were all HWiNFO32 benchmarks run as Multi-threaded?
I suggest running Single-threaded test to get pure core performance (however, in this case the Turbo will most probably engage).

Quote:

---

Originally Posted by LIKMARK

some explanations:

I wanted to compare 8 Nehalem cores (HT imagined or not, they are listed as cores by apps) with 8 lower clocked Barcelona cores on dual socket (L1N64WS/B) with 4gigs of reg ram under Vista x64. I included the higher clocked qx benches to compare as I find it interesting :)

BTW: The third truecrypt pic and the last pic is of a QX6700@3GHz iirc

---

I can only speak for myself, but yes mine were run multithreaded. Running single threaded is.. ..what for? We all know Intel is better at single threaded apps, and single threaded apps is not the discussion here :)

Except for FPU.. They made it quite weak in NetBurst, somehow improved in Merom. They're still betting on SSEx.

Quote:

---

Originally Posted by LIKMARK

I can only speak for myself, but yes mine were run multithreaded. Running single threaded is.. ..what for? We all know Intel is better at single threaded apps, and single threaded apps is not the discussion here :)

---

Jcornell: would you be able to do some of the benches at lower clocks? to compare?

Now if onl somebody would have the patience to make a small graph depicting all this beautiful Nehalem scores vs. other configurations.
:D

Quote:

---

Originally Posted by XS Janus

Now if onl somebody would have the patience to make a small graph depicting all this beautiful Nehalem scores vs. other configurations.
:D

---

sounds like you're the chosen one :D (That'll learn ya to stick your head up :p: )

hmm ok if noticed something, seems HW32 cant dedect power saving stat or dont switches to full speed on my QX9650. I get 35000 point point more when cpu runs at 3ghz fixed.

ok if done the charts, just a quick hackjob but still. :p:

i only have added the dualsocket Opteron and the QX9650 (exept for truecrypt), cause this systems had full sets of test.

How did you do the BOINC benchmark, did you add all the 'CPUs' (logical cores) together?

I'm still wondering if that wouldn't give SMT too much credit, though.

Quote:

---

Originally Posted by bowman

How did you do the BOINC benchmark, did you add all the 'CPUs' (logical cores) together?

I'm still wondering if that wouldn't give SMT too much credit, though.

---

The boinc score is per core.

Quote:

---

Originally Posted by Hornet331

Boinc only tests one cpu (core).

---

Yeah, but, as you can see when SMT is on the score is hardly any lower, so if one added those 8 logical cores together SMT would give something like an 80% gain which is just crazy.

Is there any way to test it in a more realistic way?

Quote:

---

Originally Posted by bowman

Yeah, but, as you can see when SMT is on the score is hardly any lower, so if one added those 8 logical cores together SMT would give something like an 80% gain which is just crazy.

Is there any way to test it in a more realistic way?

---

sry i edited the post, i meant boinc tests all cores and then divides the scores, at least i think thats how it happens. But im not sure about that.

Okay, one can only hope for the WCGers it's really that good a boost.. :D

Nice chart :)

Now there's no lower clock Nehalem, all recycled :rofl: ...

Just old benches ... Gainestown 2.4GHz A0 (HT Disabled + crapped BIOS :ROTF:)
http://i245.photobucket.com/albums/g...halem/G_24.jpg

...

Quote:

---

Originally Posted by LIKMARK

http://www.xtremesystems.org/forums/...p;d=1217107562

---

Are all the Nehalem benches with single channel DDR3? JC, what is the problem with the RAM channels is it the CPU or the board?

@Hornet331: Thank you very much hor your trouble. That was very fast.:p:
This really puts it into perspective.
Suddenly the power is becoming obvious.
:cool:

Here the Cinebench chart with 2,4ghz gainestown with 8 threads.

real cores still own HT cores. ;)

edit: fixed some typos (damn it was late when i did this graphs :ROTF:) see here

Thanks for the effort hornet, graphs make things so much clearer than a whole bunch of numbers! ;)

How do you spell ownage? :ROTF:

Quote:

---

Originally Posted by virtualrain

Are all the Nehalem benches with single channel DDR3? JC, what is the problem with the RAM channels is it the CPU or the board?

---

All Gainestown are with triple channel ... Latest Bloomfield benches are triple too ... ;)

Quote:

---

Originally Posted by bowman

How did you do the BOINC benchmark, did you add all the 'CPUs' (logical cores) together?

I'm still wondering if that wouldn't give SMT too much credit, though.

---

Quote:

---

Originally Posted by bowman

Yeah, but, as you can see when SMT is on the score is hardly any lower, so if one added those 8 logical cores together SMT would give something like an 80% gain which is just crazy.

Is there any way to test it in a more realistic way?

---

I asked about this some pages before in this thread but got no viable answer...We need a Nehalem ppd statisctics(actual work it does) in order to measure the SMT influence in WCG.

I see JC has posted that all latest tests were done with 3 channel so that's it when it comes to memory efficiency.

Quote:

---

Originally Posted by STaRGaZeR

How do you spell ownage? :ROTF:

---

Please explain?

I for one is not exeptionally impressed with initial Nehalem performance, even less if this is with triple channel memory performance.

That a lower clocked TLB patch ridden old socket F system with less than poor memory performance does this good compared to Nehalem should worry you more than gloating "ownage". If Shanghai corrects TDP and clocks I do not see Nehalem as superior as most people in this thread.

Intel employee dr.who's appearance in this thread support my assumptions IMO.

PS: This is not for starting flame war, but based on quite obvious results. I spent quite a long time doing benches and comparing yesterday, so please don't misunderstand this as threadcrapping :)

BTW, why does the Gainestown result show 16 threads? Bug?

Quote:

---

Originally Posted by LIKMARK

BTW, why does the Gainestown result show 16 threads? Bug?

---

where?

do you mean the screen that JC provided of the 2.4GHz one?

yes, but I see now that it is the other run with the bloomfield @ 3,07GHz ;)

Quote:

---

Originally Posted by LIKMARK

Please explain?

I for one is not exeptionally impressed with initial Nehalem performance, even less if this is with triple channel memory performance.

That a lower clocked TLB patch ridden old socket F system with less than poor memory performance does this good compared to Nehalem should worry you more than gloating "ownage". If Shanghai corrects TDP and clocks I do not see Nehalem as superior as most people in this thread.

Intel employee dr.who's appearance in this thread support my assumptions IMO.

PS: This is not for starting flame war, but based on quite obvious results. I spent quite a long time doing benches and comparing yesterday, so please don't misunderstand this as threadcrapping :)

BTW, why does the Gainestown result show 16 threads? Bug?

---

You do realize what you are comparing here?

8 Barcelona cores @ ~2.4GHz vs. 4 Nehalem cores @ 2.93GHz.

8 Barcelona cores @ 2.4GHz under 64-bit (which improves performance by 15-20%) are losing to 4 Nehalem cores @ 2.93GHz under 32-bit.

Granted your numbers are a bit lower than they should be for a Barcelona system. If you read Anandtech's MP Barcelona review, Dual Opteron 8356 (2.3GHz) score 14,487, under 64-bit. Were Nehalem also in a 64-bit environment, that would mean that 4 Nehalem cores @ 2.93GHz would be a good 15% faster than 8 Barcelona cores @ 2.3GHz. Work that out and that means that 4 Nehalem cores are not that far behind clock/clock parity with 8 Barcelona cores.

If you look at the Gainestown DP results, they are just as impressive. With SMT enabled (full 16 threads, for some reason only 8 threads were run), Dual Gainestown @ 2.4GHz will effectively tie Quad AMD Opteron 8356 @ 2.3GHz in Cinebench 64-bit.

Now imagine the performance of Quad Beckton (32 cores, 64 threads, 24MB L3/CPU) and you know that AMD is in trouble.

Quote:

---

Originally Posted by LIKMARK

yes, but I see now that it is the other run with the bloomfield @ 3,07GHz ;)

---

yeah he was only shwoing the database for the result. :up:

This is a comparison of 8 nehalem threads @2,93GHz vs 8 barcelona threads @2,375Ghz. HT or not, a thread is a thread. That is a more correct description.

Quote:

---

Originally Posted by LIKMARK

This is a comparison of 8 nehalem threads @2,93GHz vs 8 barcelona threads @2,375Ghz. HT or not, a thread is a thread. That is a more correct description.

---

? Not sure if you were replying to me or not.

A thread doesn't mean anything, because the 8 threads in Nehalem don't cost you anything. Quad core Nehalem is fully comparable to Quad core Barcelona. DP Gainestown (8 cores) is fully comparable to DP Barcelona (8 cores).

(The only area where a thread would "cost" you is if you were paying for a OS per CPU, then HT would probably be disabled. But that doesn't apply to anything here.)

i think its not that easy, cause when you look at the gainestown score and then at the bloomfield score you see where the difference lies. Gainestown has 8 threads on 8 physical cores, while bloomfield has 8 threads on 4 physical cores and 4 virtual cores.

And if you compare gainestown 8 threads (aka HT disabled) with your opteron 8 threads scores, the result of gainestown are quite impressive.

Quote:

---

Originally Posted by LIKMARK

This is a comparison of 8 nehalem threads @2,93GHz vs 8 barcelona threads @2,375Ghz. HT or not, a thread is a thread. That is a more correct description.

---

Kind of weird why you compare the Bloomfield(High End Desktop, nearly WS) with your AMD Server ...

Yeap, you're right : thread is a thread, but need to differentiate a Physical Core and Logical Core, there're not the same :shrug:

Quote:

---

Originally Posted by Hornet331

bloomfield has 8 threads on 4 physical cores and 4 virtual cores.

---

8 logical/virtual, 4 physical :p:

Quote:

---

Originally Posted by Hornet331

Here the Cinebench chart with 2,4ghz gainestown with 8 threads.

real cores still own HT cores. ;)

edit: fixed some typos (damn it was late when i did this graphs :ROTF:) see here

---

Thanks for your Chart :up:

Need updated CINEBENCH R10 Gainestown 3.07GHz*2 HT Disabled result again ? :rofl:
Nah, I think thats no need to :D

lol you want put every other system to shame with a score of 22k+ in cinebench?

well, i certainly dont gona stop you. :D

Quote:

---

Originally Posted by LIKMARK

This is a comparison of 8 nehalem threads @2,93GHz vs 8 barcelona threads @2,375Ghz. HT or not, a thread is a thread. That is a more correct description.

---

Exactly right. And I remember the same defense for P4 w/HT vs X2.

Quote:

---

Originally Posted by flippin_waffles

Exactly right. And I remember the same defense for P4 w/HT vs X2.

---

So...because Nehalem supports SMT. You quickly say AMD is better because 2 Quadcores can compete with 1 quadcore?

Try 2 quads vs 2 quads. Or try see what a Bloomsfield cost vs those 2 2000 series Barcelonas.

1 Nehalem Quad@2.93Ghz beats 2 Barcelona Quads@2.4Ghz in Cinebench.
Not even to talk about Barcelona uses 64bit and Nehalem 32bit. (Nehalem would perform some 10% better with 64bit)
1 CPU vs 2 CPUs. 500$ vs 1500$

Or use the Gainestown with SMT disabled. About 50% faster than the dual Barcelonas at the same 2.4Ghz.

It is weird we discuss the numbers Nehalem puts out based on one app that does favor intel hardware more ,and that's Cinebech(what's the use of Cinebench anyway?).
I would like to see some 3D Studio,Lightwave ,Maya etc. numbers.Also some head on comparisons with enc./decod. using optimal compiler tweaks for each uarch(Core2,K10,Nehalem).
You know,the real stuff people actually use.WCG ppd also counts in too.

Quote:

---

Originally Posted by informal

It is weird we discuss the numbers Nehalem puts out based on one app that does favor intel hardware more ,and that's Cinebech(what's the use of Cinebench anyway?).
I would like to see some 3D Studio,Lightwave ,Maya etc. numbers.Also some head on comparisons with enc./decod. using optimal compiler paths for each uarch(Core2,K10,Nehalem).

---

Cinebench is a 3D rendering benchmark. So it is a good benchmark of how well these CPUs will perform with rendering. And I wouldn't say it favors Intel hardware at all. K8 actually isn't too far behind Conroe in Cinebench. Unfortunately K10 does very little to improve performance in Cinebench beyond adding more cores.

To me one of the most exciting things will be that a quad-core Nehalem overclocked to 4.0GHz will roughly equal the performance of 16 Barcelona cores @ 2.3GHz in Cinebench.... 4x Opteron 8356 is currently a $6,000 setup. A 4GHz Nehalem might be possible out of a $284 CPU.... and certainly the $999 XE part. Obviously you can't overclock a CPU when you are doing critical work, but the power that is going to be available to enthusiasts with Nehalem is quite amazing.

Quote:

---

Originally Posted by informal

You know,the real stuff people actually use.WCG ppd also counts in too.

---

I would like to see Folding@Home SMP client :D

3D Rendering :off::lol: , Decrease and to saving time for 3D rendering I think = Time is Money :ROTF:

http://news.yahoo.com/s/nm/20080708/..._dreamworks_dc

http://i245.photobucket.com/albums/g...aneous/Opt.jpg

http://i245.photobucket.com/albums/g...aneous/P_I.jpg

...

Quote:

---

Originally Posted by informal

It is weird we discuss the numbers Nehalem puts out based on one app that does favor intel hardware more ,and that's Cinebech(what's the use of Cinebench anyway?).
I would like to see some 3D Studio,Lightwave ,Maya etc. numbers.Also some head on comparisons with enc./decod. using optimal compiler tweaks for each uarch(Core2,K10,Nehalem).
You know,the real stuff people actually use.WCG ppd also counts in too.

---

its more weird that since 2006 amd enthusiast suddenly claim that every benchmark out there favours intel and therefor isn't valide (exept of siencemark and spec_fp rate).

What do think happens in boinc? E.g. look at seti, the client is optimized by the community to support the specific instructions of certain cpus.

But sadly only intel gets allways the newst optimisations -> amd is at sse3 while the optimized intel client allready supports sse4.1

And no that hase nothing to do with fanboyism of there side, rather then the bigger userbase the can reach when they are optimizing for intel cpus.


Also why should a dev use generic flags for the compilers? To limit his program artificially? If theres is the option i would allways try to provide optimizations for both cpus or even certain architectures.

Quote:

---

Originally Posted by Extelleron

Cinebench is a 3D rendering benchmark. So it is a good benchmark of how well these CPUs will perform with rendering. And I wouldn't say it favors Intel hardware at all. K8 actually isn't too far behind Conroe in Cinebench. Unfortunately K10 does very little to improve performance in Cinebench beyond adding more cores.

---

Actually Cinebench is using scalar SSE instead of vector SSE instructions thus you see no change in K8 vs K10 scores.The use of vector SSE is the main point of even optimizing for that instr. extension set.Thus Cinebench is not a "good" benchmark.
Also i know that people somehow think that SPI represents a valid test to show a strong fp perfromance (intel clearly owns this test),but there are programs that calculate Pi better on AMD hardware but i don't see anyone using these(http://www.xtremesystems.org/forums/...&highlight=gmp).

Quote:

---

Originally Posted by informal

It is weird we discuss the numbers Nehalem puts out based on one app that does favor intel hardware more ,and that's Cinebech(what's the use of Cinebench anyway?).
I would like to see some 3D Studio,Lightwave ,Maya etc. numbers.Also some head on comparisons with enc./decod. using optimal compiler tweaks for each uarch(Core2,K10,Nehalem).
You know,the real stuff people actually use.WCG ppd also counts in too.

---

Then head over to anandtech, they tested 3dsm SPECapc and Intel improved ~42% there compared to core 2. I hope you trust their numbers.

I wouldn't trust Anand with anything hardware or software related :p: .But i do believe Nehalem improves over C2 in 3DS.

Quote:

---

Originally Posted by informal

Also i know that people somehow think that SPI represents a valid test to show a strong fp perfromance (intel clearly owns this test),but there are programs that calculate Pi better on AMD hardware but i don't see anyone using these(http://www.xtremesystems.org/forums/...&highlight=gmp).

---

B'cause SPI is the most popular test, Yeah for now Intel is leading on SPI, but what about the past, before C2D ( AMD clearly owned this test, right ?)

You mean MaxPI Multicore PI calc ? Never heard of it, so this is not strange if you don't see anyone using these ...

So @LIKMARK :) , this is saying that everyone is waiting for your Barcelona to show some MaxPI expression :D

Quote:

---

Originally Posted by JCornell

B\'cause SPI is the most popular test, Yeah for now Intel is leading on SPI, but what about the past, before C2D ( AMD clearly owned this test, right ?)

You mean MaxPI Multicore PI calc ? Never heard of it, so this is not strange if you don\'t see anyone using these ...

---

spi has always been intel\'s benchmark, even with netburst

Quote:

---

Originally Posted by informal

Also i know that people somehow think that SPI represents a valid test to show a strong fp perfromance (intel clearly owns this test),but there are programs that calculate Pi better on AMD hardware but i don't see anyone using these(http://www.xtremesystems.org/forums/...&highlight=gmp).

---

you mean gmp, which is compiled on gcc with k8 flag? way to go to gimp core2. :yepp:

it would be the same as spi, you just complained about, only difference would be that it now would prefer amd instead of intel...

Quote:

---

Originally Posted by Hornet331

you mean gmp, which is compiled on gcc with k8 flag? way to go to gimp core2. :yepp:

it would be the same as spi, you just complained about, only difference would be that it now would prefer amd instead of intel...

---

Gimp core2 :p:
You can compile with appropriate flag if you want:
http://gcc.gnu.org/onlinedocs/gcc-4....002d64-Options

Quote:

---

Originally Posted by informal

Gimp core2 :p:

---

I see ...

But now I'm just curios that JumpingJack saying about Phenom is always 10% faster than Core2 on MaxPI :p:

Quote:

---

Originally Posted by LIKMARK

Please explain?

---

Well, The results speak for themselves. Also I don't know why are you comparing 8 cores vs 4 cores, even if Nehalem has 2 threads per core. The fair comparison is between Gainstown vs your dual Barcelonas. I'm not impressed with Nehalem single threaded perfomance because Core2 is already a monster there and this is only a little improvement, but in multi-threaded apps Nehalem destroys Phenom and Core2. I'd like to see how Shangai performs, but even if it ends 10% faster clock per clock than Barcelona it'd be still slower than a Yorkfield in most cases. In a server environment maybe they'll shine, but not in desktop.

Quote:

---

Originally Posted by Extelleron

? Not sure if you were replying to me or not.

A thread doesn't mean anything, because the 8 threads in Nehalem don't cost you anything. Quad core Nehalem is fully comparable to Quad core Barcelona. DP Gainestown (8 cores) is fully comparable to DP Barcelona (8 cores).

---

According to boinc cpu test, ht seems to cost some performance.

Quote:

---

Originally Posted by JCornell

Kind of weird why you compare the Bloomfield(High End Desktop, nearly WS) with your AMD Server ...

Yeap, you're right : thread is a thread, but need to differentiate a Physical Core and Logical Core, there're not the same :shrug:

---

Asus L1N64WS/B is a WS board ;)

Quote:

---

Originally Posted by Shintai

So...because Nehalem supports SMT. You quickly say AMD is better because 2 Quadcores can compete with 1 quadcore?

Try 2 quads vs 2 quads. Or try see what a Bloomsfield cost vs those 2 2000 series Barcelonas.

1 Nehalem Quad@2.93Ghz beats 2 Barcelona Quads@2.4Ghz in Cinebench.
Not even to talk about Barcelona uses 64bit and Nehalem 32bit. (Nehalem would perform some 10% better with 64bit)
1 CPU vs 2 CPUs. 500$ vs 1500$

Or use the Gainestown with SMT disabled. About 50% faster than the dual Barcelonas at the same 2.4Ghz.

---

Actually those cpus cost me 100USD each, not 1500USD ;)

Of course Gainstown is faster, as it should be, it's just that I don't think it is fast enough to recieve all the "omg wtf bbq" remarks and no critical approach to the numbers provided. That's the reason I compared in the first place.

Quote:

---

Originally Posted by Extelleron

Cinebench is a 3D rendering benchmark. So it is a good benchmark of how well these CPUs will perform with rendering. And I wouldn't say it favors Intel hardware at all. K8 actually isn't too far behind Conroe in Cinebench. Unfortunately K10 does very little to improve performance in Cinebench beyond adding more cores.

To me one of the most exciting things will be that a quad-core Nehalem overclocked to 4.0GHz will roughly equal the performance of 16 Barcelona cores @ 2.3GHz in Cinebench.... 4x Opteron 8356 is currently a $6,000 setup. A 4GHz Nehalem might be possible out of a $284 CPU.... and certainly the $999 XE part. Obviously you can't overclock a CPU when you are doing critical work, but the power that is going to be available to enthusiasts with Nehalem is quite amazing.

---

Guys.. Lets compare prices when they're out, and then the whole systems...

Quote:

---

Originally Posted by informal

I wouldn't trust Anand with anything hardware or software related :p: .But i do believe Nehalem improves over C2 in 3DS.

---

LOL +1 :up:

Quote:

---

Originally Posted by JCornell

B'cause SPI is the most popular test, Yeah for now Intel is leading on SPI, but what about the past, before C2D ( AMD clearly owned this test, right ?)

You mean MaxPI Multicore PI calc ? Never heard of it, so this is not strange if you don't see anyone using these ...

So @LIKMARK :) , this is saying that everyone is waiting for your Barcelona to show some MaxPI expression :D

---

I'll check it out, just don't have time atm.

If there's a store beyond a friend's friend's discarded server equipment that sells Opteron Barcelonas for 500kr I would like a link.. ;) But I'm guessing you got a 'special deal' here.

Quote:

---

According to boinc cpu test, ht seems to cost some performance.

---

He doesn't mean performance, he means price. You can't compare two CPUs to one CPU because the price difference between a UP platform vs. a DP platform, Nehalem or not, is massive.

Perhaps the diagrams should've counted both cores and threads so we'd avoid this bickering? :shrug:

Quote:

---

Originally Posted by bowman

If there's a store beyond a friend's friend's discarded server equipment that sells Opteron Barcelonas for 500kr I would like a link.. ;) But I'm guessing you got a 'special deal' here.



He doesn't mean performance, he means price. You can't compare two CPUs to one CPU because the price difference between a UP platform vs. a DP platform, Nehalem or not, is massive.

Perhaps the diagrams should've counted both cores and threads so we'd avoid this bickering? :shrug:

---

Pretty sure he got them on eBay. Pre-B3 Barcelonas are being dumped on eBay at pretty attractive prices. You can easily get an Opteron 2347 or 2344 HE for $100.

Heck looking now you can also buy 4x Opteron 8346 for $150 each. That would make a pretty interesting setup.

Quote:

---

Originally Posted by Hornet331

you mean gmp, which is compiled on gcc with k8 flag? way to go to gimp core2. :yepp:

it would be the same as spi, you just complained about, only difference would be that it now would prefer amd instead of intel...

---

Nehalem will Rock because Core 2 Q already is.

Of course it wouldn't matter to AMD fans. It is only valid if the software or app favors AMD.:up:

The only good reviewers are those who cripple the competition to make AMD look better. You know, these same folks who acted like they were Vampires and Optimized SSE software was a Cross or Garlic. I'm talking about off the shelf on the market software. Even if the software isn't on the market or readily available, nothing would change AMD fans view. "Just Win Baby". Doesn't matter if it is 4 vs 2 or the cost of 4 vs two, just win baby!

I do wonder where the mods who monitor thread crapping are?:rolleyes: 6 weeks!

To make thing more clear i have add the number of CPUs used on each system in the charts (as bowman suggested).

Quote:

---

Originally Posted by LIKMARK

According to boinc cpu test, ht seems to cost some performance.

Asus L1N64WS/B is a WS board ;)

Actually those cpus cost me 100USD each, not 1500USD ;)

Of course Gainstown is faster, as it should be, it's just that I don't think it is fast enough to recieve all the "omg wtf bbq" remarks and no critical approach to the numbers provided. That's the reason I compared in the first place.

Guys.. Lets compare prices when they're out, and then the whole systems...

LOL +1 :up:


I'll check it out, just don't have time atm.

---

Can we assume that nehalem is an improvement over the yorkfields/harpertowns no matter what percentage it is?
If we assume that, then you pick a series of 3-4 benchmarks and you run with your AMD setup and I'll run with my "slower than Nehalem" Harpertown setup and we'll look at the results.
I am on air cooling and at 3404mhz so not "Xtreme" by anyones definition.
Fair enough?:D

The tears is like ambrosia. Is this another "Core 2 is fake, Intel is using X2s" etc whine again like in 2006? Core 2 spanks K10 hard, Nehalem utterly destroys it. And thats not gonna change a single bit, not with 45nm K10s either. It seems some people are starting to sue every possible excuse. Now also compared 2 K10 quads vs 1 Nehalem quad to win some space with justifying its 8 threads vs 8 threads. Its pathetic and fanboysim beyond reason.

But again, 2009 will be a whole other game. And there will be the number 11 in it, but no K in front of it.

Quote:

---

Originally Posted by Shintai

The tears is like ambrosia. Is this another "Core 2 is fake, Intel is using X2s" etc whine again like in 2006? Core 2 spanks K10 hard, Nehalem utterly destroys it. And thats not gonna change a single bit, not with 45nm K10s either. It seems some people are starting to sue every possible excuse. Now also compared 2 K10 quads vs 1 Nehalem quad to win some space with justifying its 8 threads vs 8 threads. Its pathetic and fanboysim beyond reason.

But again, 2009 will be a whole other game. And there will be the number 11 in it, but no K in front of it.

---

Lol, we'd better hope that doesn't happen though. AMD going bankrupt is the last thing the CPU industry needs. I don't think Intel would let it happen though. There is already quite a bit of investigation into Intel's activities (although I think it is all BS), imagine how much the gov would be breathing down Intel's back if AMD died....

Quote:

---

Originally Posted by Extelleron

Lol, we'd better hope that doesn't happen though. AMD going bankrupt is the last thing the CPU industry needs. I don't think Intel would let it happen though. There is already quite a bit of investigation into Intel's activities (although I think it is all BS), imagine how much the gov would be breathing down Intel's back if AMD died....

---

Honestly, goverments would do nothing. But they would regulate prices if needed.

Quote:

---

Originally Posted by Shintai

But again, 2009 will be a whole other game. And there will be the number 11 in it, but no K in front of it.

---

Ah, a prediction. These are so fun to come back to later.

Quote:

---

Originally Posted by STaRGaZeR

Well, The results speak for themselves. Also I don't know why are you comparing 8 cores vs 4 cores, even if Nehalem has 2 threads per core. The fair comparison is between Gainstown vs your dual Barcelonas. I'm not impressed with Nehalem single threaded perfomance because Core2 is already a monster there and this is only a little improvement, but in multi-threaded apps Nehalem destroys Phenom and Core2. I'd like to see how Shangai performs, but even if it ends 10% faster clock per clock than Barcelona it'd be still slower than a Yorkfield in most cases. In a server environment maybe they'll shine, but not in desktop.

---

I wondered how HT on nehalem performed, therefore I compared with my logical 8 threads system.

Quote:

---

Originally Posted by bowman

He doesn't mean performance, he means price. You can't compare two CPUs to one CPU because the price difference between a UP platform vs. a DP platform, Nehalem or not, is massive.

---

I'm quite sure nehalem will cost more than my setup when it's out..

Quote:

---

Originally Posted by Extelleron

Pretty sure he got them on eBay. Pre-B3 Barcelonas are being dumped on eBay at pretty attractive prices. You can easily get an Opteron 2347 or 2344 HE for $100.

Heck looking now you can also buy 4x Opteron 8346 for $150 each. That would make a pretty interesting setup.

---

Got it in one:up: After all I am not mad either, I would under no circumstances dump up to 800USD on two 2347's today..

Quote:

---

Originally Posted by Donnie27

Nehalem will Rock because Core 2 Q already is.

Of course it wouldn't matter to AMD fans. It is only valid if the software or app favors AMD.:up:

The only good reviewers are those who cripple the competition to make AMD look better. You know, these same folks who acted like they were Vampires and Optimized SSE software was a Cross or Garlic. I'm talking about off the shelf on the market software. Even if the software isn't on the market or readily available, nothing would change AMD fans view. "Just Win Baby". Doesn't matter if it is 4 vs 2 or the cost of 4 vs two, just win baby!

I do wonder where the mods who monitor thread crapping are?:rolleyes: 6 weeks!

---

Look up post number 804 - 807 and compare constructivness to your post, and we can discuss threadcrapping. How much time did it take you to write your post?
Three sec's I assume: One: copy BS post from some AMD thread you have threadcrapped, two: paste into this one, three: "submit reply".

Quote:

---

Originally Posted by Movieman

Can we assume that nehalem is an improvement over the yorkfields/harpertowns no matter what percentage it is?
If we assume that, then you pick a series of 3-4 benchmarks and you run with your AMD setup and I'll run with my "slower than Nehalem" Harpertown setup and we'll look at the results.
I am on air cooling and at 3404mhz so not "Xtreme" by anyones definition.
Fair enough?:D

---

Excuse me I do not understand your intent with this post. As I answered STaRGaZeR, I wondered how HT on nehalem performed, therefore I compared with my logical 8 threads system. I do not want a "showdown" with you. It was my intent to show something to compare nehalem with. This obviously is not something anyone here want to be done, so I'll remove them at once.

Quote:

---

Originally Posted by Shintai

The tears is like ambrosia. Is this another "Core 2 is fake, Intel is using X2s" etc whine again like in 2006? Core 2 spanks K10 hard, Nehalem utterly destroys it. And thats not gonna change a single bit, not with 45nm K10s either. It seems some people are starting to sue every possible excuse. Now also compared 2 K10 quads vs 1 Nehalem quad to win some space with justifying its 8 threads vs 8 threads. Its pathetic and fanboysim beyond reason.

But again, 2009 will be a whole other game. And there will be the number 11 in it, but no K in front of it.

---

Please scroll up to my answer to Donnie27s post. The same applies for you.

Quote:

---

Originally Posted by LIKMARK

I'm quite sure nehalem will cost more than my setup when it's out..

---

I think the word "Out of education reach" can be used here.

You can get 2.66Ghz Nehalems for 286$ at launch in about 2 months.

Yes and you'll have your brand new board and 6 gigs of DDR3 ram to utilize triple channel for free?

i think 286$ is a bit low... maybe in the 300-400$ region, which is still damn cheap (at least in euro land), but if it is really 286$ i have nothing against it. :p:

who said, you need the tripple channel, you also can run it as dual channle i think 21gb/s is enough. :D
also i think 190€ for 6gb ddr3 1333 isnt that expensive. ;)

thought for myself i aim for 12gb :hrhr: (cause 2x2gb ddr31800 is allready at 180€ and in 2-3 months im sure well see moduels for 150€ so 450€ total is not that bad, especial when i compare it to my current 1x2gb ddr3, which i have bought for 360€. :yepp: )

Quote:

---

Originally Posted by Donnie27

Of course it wouldn't matter to AMD fans. It is only valid if the software or app favors AMD.:up:

The only good reviewers are those who cripple the competition to make AMD look better. You know, these same folks who acted like they were Vampires and Optimized SSE software was a Cross or Garlic. I'm talking about off the shelf on the market software. Even if the software isn't on the market or readily available, nothing would change AMD fans view. "Just Win Baby". Doesn't matter if it is 4 vs 2 or the cost of 4 vs two, just win baby!

---

Totally agreed ... :)

Quote:

---

Originally Posted by LIKMARK

Excuse me I do not understand your intent with this post. As I answered STaRGaZeR, I wondered how HT on nehalem performed, therefore I compared with my logical 8 threads system. I do not want a "showdown" with you. It was my intent to show something to compare nehalem with. This obviously is not something anyone here want to be done, so I'll remove them at once.

---

my point was a simple one. Since from what we've learned so far that Nehalem is superior to the current harpertown/yorkfield setups, I was offering to compare my 8 logical cores to yours. That's all.

Quote:

---

Originally Posted by Shintai

Core 2 spanks K10 hard, Nehalem utterly destroys it.

---

pure fanboism? Core2 spanks K10 hard in some apps, but K10 spanks Core2 in other apps, nehalem is a future product, and altough we have seen many benchmarks of it, i haven seen any really good direct comparisons
but as always i didn\'t expect anything more from you

Quote:

---

Originally Posted by Shintai

And thats not gonna change a single bit, not with 45nm K10s either.

---

So you are aware of all changes in K10.5 (Shanghai) ?
You know how that extra L3 cache is gonna effect performance, and which minor tweaks this CPU has, please lend me your crystall ball

Quote:

---

Originally Posted by Hornet331

i think 286 is a bit low... maybe in the 300-400$ region, which is still damn cheap (at least in euro land). :p:

who said, you need the tripple channel, you also can run it as dual channle i think 21gb/s is enough. :D

thought for myself i aim for 12gb :hrhr:

---

Now I confused of all his posted ...
First talking about Performance, then now is talking about price ??? ...

I think someone is confusing about the differences from Bloomfield & Gainestown ?

Quote:

---

Originally Posted by Bellisimo

pure fanboism? Core2 spanks K10 hard in some apps, but K10 spanks Core2 in other apps, nehalem is a future product, and altough we have seen many benchmarks of it, i haven seen any really good direct comparisons
but as always i didn\'t expect anything more from you


So you are aware of all changes in K10.5 (Shanghai) ?
You know how that extra L3 cache is gonna effect performance, and which minor tweaks this CPU has, please lend me your crystall ball

---

Which is why I offerred to compare my harpertowns which are a current product..
hell, I'll even drop the clocks from 3404 back to the stock 3000..:D

Quote:

---

Originally Posted by Bellisimo

So you are aware of all changes in K10.5 (Shanghai) ?
You know how that extra L3 cache is gonna effect performance, and which minor tweaks this CPU has, please lend me your crystall ball

---

we allready saw spi on k10 which increased 10% afaik, lets asume this is best case so any other app will be less then 10% improvement which is enough to fight with kent/york but cant touch nehalem.

At least thats my impression from what i have seen so far on the web.

Quote:

---

Originally Posted by Movieman

Which is why I offerred to compare my harpertowns which are a current product..
hell, I'll even drop the clocks from 3404 back to the stock 3000..:D

---

That would be interesting anyway to compare to Bloomfield and Gainestown, in these benches.

So I say go for it! (if you can spare the cycles ;))

Quote:

---

Originally Posted by Movieman

my point was a simple one. Since from what we've learned so far that Nehalem is superior to the current harpertown/yorkfield setups, I was offering to compare my 8 logical cores to yours. That's all.

---

Weird, just because Bloomfield show 8 thread in OS, then assumed it's pure 8 cores ... Isn't this what he saying from the beginning ??
Then Gainestown will be assume to 4 CPU's ? :p:

Quote:

---

Originally Posted by Movieman

Which is why I offerred to compare my harpertowns which are a current product..
hell, I'll even drop the clocks from 3404 back to the stock 3000..:D

---

stop WCG for like 5 mins n do the benchiesssss

you know you want to :p:

Quote:

---

Originally Posted by bowman

That would be interesting anyway to compare to Bloomfield and Gainestown, in these benches.

So I say go for it! (if you can spare the cycles ;))

---

MovieMan and I are totally know where its go when downgrade Harpertown to 3GHz, Still no chances to compete Gainestown though except Harpertown has been OC to some clock.
He mentioned it before if you are noticing it ...

Quote:

---

Originally Posted by bowman

That would be interesting anyway to compare to Bloomfield and Gainestown, in these benches.

So I say go for it! (if you can spare the cycles ;))

---

That I can do, just waiting for LIKMARK to determine the benches to be run.

Quote:

---

Originally Posted by JCornell

Weird, just because Bloomfield show 8 thread in OS, then assumed it's pure 8 cores ... Isn't this what he saying from the beginning ??
Then Gainestown will be assume to 4 CPU's ? :p:

---

JC: To be honest I'm a tad confused myself so I'm going to go and finish building my new system..
A dual Yonah.. Now there's one for you guys to laugh at.
2-2000/2mb/667 dual core 31w laptop chips on a $430.00 Intel EATX server board..
later..

Quote:

---

Originally Posted by Movieman

my point was a simple one. Since from what we've learned so far that Nehalem is superior to the current harpertown/yorkfield setups, I was offering to compare my 8 logical cores to yours. That's all.

---

Ah ok. For a moment there it seemed remarkably like: "I'm just letting you know mine is better than yours" but being a grown up person you wouldn't engage a conversation like that??:p: Thats kind of.. .

I would rather like that you compared them to nehalem bloomfields 4 logical and 4 virtual cores as this thread still is about nehalem, (isn't it?) but that's just me.. :)

Edit: @MM: just run the benches in the charts to compare.

Quote:

---

Originally Posted by Bellisimo

pure fanboism? Core2 spanks K10 hard in some apps, but K10 spanks Core2 in other apps, nehalem is a future product, and altough we have seen many benchmarks of it, i haven seen any really good direct comparisons
but as always i didn\'t expect anything more from you


So you are aware of all changes in K10.5 (Shanghai) ?
You know how that extra L3 cache is gonna effect performance, and which minor tweaks this CPU has, please lend me your crystall ball

---

With luck it can beat Kentsfield. Maybe touch york sometimes. And nice new account.

Also the word K10.5 is terrible. Perhaps K10.1. 45nm and more cache. Go make a wish. Even if 45nm K10 jumps the same performance 65nm K10 did to K8. Then its still not a performance winner.

Nehalem is a product you have in 2months. 45nm K10 you might have on december 31st.

Quote:

---

Originally Posted by LIKMARK

I wondered how HT on nehalem performed, therefore I compared with my logical 8 threads system.

---

Then you should not compare it with your Barcelonas. Use Nehalem with HT on and off (if possible). You're still comparing 4 physical cores vs 8 physical cores. I count HT as an enhancement to the core, not as another core. AMD could do the same and 1 core will remain 1 core for me. Reality is, in most benchmarks 8 K10 cores lose to 4 Nehalem cores, each at full potential, that means HT on in Nehalem. That's sad.

Quote:

---

Originally Posted by LIKMARK

I would rather like that you compared them to nehalem bloomfields 4 logical and 4 virtual cores as this thread still is about nehalem, (isn't it?) but that's just me.. :)

---

Nah ... again ... How in the hell Bloomfield 1CPUs is the same to Harpertown 2CPU's ??
See, this is what I meant, Bloomfield showed 8threads in OS, so this must be 2CPU's :shrug:

Quote:

---

Originally Posted by STaRGaZeR

Then you should not compare it with your Barcelonas. Use Nehalem with HT on and off (if possible). You're still comparing 4 physical cores vs 8 physical cores. I count HT as an enhancement to the core, not as another core. AMD could do the same and 1 core will remain 1 core for me. Reality is, in most benchmarks 8 K10 cores lose to 4 Nehalem cores, each at full potential. That means HT on in Nehalem. That's sad.

---

Windows XP and Windows Vista counts "Ht-cores" as "another core", wouldn't it be safe to believe that the os uses the virtual core like if it was a physical core, isn't that what HT technology is all about? If not, please explain to me.

EDIT: I compared 8 threads to the 8 threads I have on my system, I don't see what is so wrong about that?? It's called multithreading, not multicpuusage or whatever. I wanted to see how Nealem Ht technology held up against my 8 threads (or cores, it dosen't matter to me, and I'm pretty sure it dosen't matter to the os when its multithreading if the threads is physical or virtual. It applies them the same way)

Quote:

---

Originally Posted by LIKMARK

Windows XP and Windows Vista counts "Ht-cores" as "another core", wouldn't it be safe to believe that the os uses the virtual core like if it was a physical core, isn't that what HT technology is all about? If not, please explain to me.

---

It may use it as if it was another core but it is not and I say that from experience. I have a 2005 SM X7DA8-G2 MB with 2-3600/2mb/800 single core Irwindales(netburst) cpu's in it.
In DC work running 4 WU at a time with HT turned on the machine would do app 20-30% more work than with HT turned off and doing 2 WU at a time.
An increase yes, but not even close to having an additional 2 real cores so to compare your 8 real cores against anything with 4 cores and HT is not a fair comparison.
This is exactly why I offered to compare my 8 core harpertown with your 8 core AMD machine in any group of benchmarks you choose.
Yes, I am aware that the AMD with it's on chip IMC will do some benches better but overall I think I'd squash it like a bug..
OOH, it's nice not to have to be diplomatic once in a while!:rofl:

XP and Vista are 'SMT aware', they know what a 'fake' logical core is and actually take that into account. Not to say all software running would do that, though. They might use it as if it was an entire physical CPU.

http://www.2cpu.com/Hardware/ht_anal...rthreading.doc
http://www.intel.com/support/process.../CS-017343.htm

@MM:

Quote:

---

Yes, I am aware that the AMD with it's on chip IMC will do some benches better but overall I think I'd squash it like a bug..
OOH, it's nice not to have to be diplomatic once in a while!

---

lol, your'e a honest son of a b*** at least :D

Please do the benches in the chart, the more comparing systems the better imo.

As for the HT tech discussion, I do not advocate that it runs as efficient as another core, but as Intel is pretty proud of this "old crap - new package" thing called "new and better than last time" HT tech I wanted to test it up against 8 physical cores. I do not have a nehalem system, so I had to do with what I got. Good night :)

Quote:

---

Originally Posted by LIKMARK

@MM:


lol, your'e a honest son of a b*** at least :D

Please do the benches in the chart, the more comparing systems the better imo.

As for the HT tech discussion, I do not advocate that it runs as efficient as another core, but as Intel is pretty proud of this "old crap - new package" thing called "new and better than last time" HT tech I wanted to test it up against 8 physical cores. I do not have a nehalem system, so I had to do with what I got. Good night :)

---

Good night and remember one thing about me, if AMD tomorrow made a better mousetrap it would be in my house. I am only a fanboy of what is best.
Now is Intels time, next year it may be AMD's as it was 3 years ago and no one, I repeat no one, wishes them good fortune more than I do.

LIKMARK, read this: http://en.wikipedia.org/wiki/Hyper-threading

You can't count or compare 4 cores with HT vs 8 cores. Even if in the XP or Vista task manager are listed as such.

Quote:

---

Originally Posted by Movieman

Good night and remember one thing about me, if AMD tomorrow made a better mousetrap it would be in my house. I am only a fanboy of what is best.
Now is Intels time, next year it may be AMD's as it was 3 years ago and no one, I repeat no one, wishes them good fortune more than I do.

---

:dammit: , Good Morning for me :rofl:
Continue to sleep, :toast: today is typhoon holiday :clap:

Are you intend to benchs it ? If so, just prepare to be commented :p :
1. Nah, that's just not the same clock
2. Oh, You have larger cache
3. OMG, is that 45nm ?
4. etc ... and etc ... :p

I'll run the benches tomorrow at stock speed of 3000mhz
These are 80w E5450's..OEM's not Es's..
Board is a SM X7DWA-N
PCP+C 750w Silencer
4x1 gig DDR2-800 kingston HyperX FBDimms
Air cooled with my own cooling mix
single 36 gig raptor

Quote:

---

Originally Posted by Bellisimo

pure fanboism? Core2 spanks K10 hard in some apps, but K10 spanks Core2 in other apps, nehalem is a future product, and altough we have seen many benchmarks of it, i haven seen any really good direct comparisons
but as always i didn\'t expect anything more from you


So you are aware of all changes in K10.5 (Shanghai) ?
You know how that extra L3 cache is gonna effect performance, and which minor tweaks this CPU has, please lend me your crystall ball

---

I kinda wanted to report your post (so as not to feed the trolls and further derail this thread), but maybe it's not necessary, I'll give you a simple explanation why you are wrong: overall core 2 is superior, the benchmarks do not lie.
though, the uncore part (system interconnect) of k10 is superior to core 2, the most crucial part, the core, is not. k10 wins only in virtualisation and bandwidth heavy tasks. now what does nehalem improve on?

Quote:

---

Originally Posted by Hornet331

we allready saw spi on k10 which increased 10% afaik, lets asume this is best case so any other app will be less then 10% improvement which is enough to fight with kent/york but cant touch nehalem.

At least thats my impression from what i have seen so far on the web.

---

Actually it increased 14%,but hey why say 14 when 10% sounds better in intel fanboy ears ?

Quote:

---

Originally Posted by informal

Actually it increased 14%,but hey why say 14 when 10% sounds better in intel fanboy ears ?

---

It always amuses me to see you call someone a fanboy . . .

Quote:

---

Originally Posted by informal

Actually it increased 14%,but hey why say 14 when 10% sounds better in intel fanboy ears ?

---

Yeah, and that extra 4% is really gonna close the gap on Nehalem right? ;)

You complained about people 'crapping' the Deneb threads, do you spot the irony?! I gotta admit though, its funny seeing a fanboy calling someone else a fanboy... :ROTF:

Quote:

---

Originally Posted by LIKMARK

I wondered how HT on nehalem performed, therefore I compared with my logical 8 threads system.

I'm quite sure nehalem will cost more than my setup when it's out..

Got it in one:up: After all I am not mad either, I would under no circumstances dump up to 800USD on two 2347's today..

Look up post number 804 - 807 and compare constructivness to your post, and we can discuss threadcrapping. How much time did it take you to write your post?
Three sec's I assume: One: copy BS post from some AMD thread you have threadcrapped, two: paste into this one, three: "submit reply".

Excuse me I do not understand your intent with this post. As I answered STaRGaZeR, I wondered how HT on nehalem performed, therefore I compared with my logical 8 threads system. I do not want a "showdown" with you. It was my intent to show something to compare nehalem with. This obviously is not something anyone here want to be done, so I'll remove them at once.

Please scroll up to my answer to Donnie27s post. The same applies for you.

---

You started thread crapping and 807 is a thread crap if there ever was one! If what went on in the Deneb thread is thread crapping then yours and others are just that=P

Your idea of trying to compare two Quad Cores to One breaks any rule of comparisons except for a Desperate AMD Fanboy move. This made even more lame by trying to down play costs. You do something silly like comparing used buggy items to products not shipping or even new is a bad joke:down:

You and your buddies don't understand benchmarking, you don't get price to performance and it is just like I said, just win baby! Nehalem's price lists were given. No one is complaining about yours and other's post "Nit-Picking" because unlike those AMD threads, no one cares that you don't get it:up: Nehalem doesn't need propaganda like K10 or whatever.

Quote:

---

Originally Posted by JCornell

:dammit: , Good Morning for me :rofl:
Continue to sleep, :toast: today is typhoon holiday :clap:

Are you intend to benchs it ? If so, just prepare to be commented :p :
1. Nah, that's just not the same clock
2. Oh, You have larger cache
3. OMG, is that 45nm ?
4. etc ... and etc ... :p

---

According to at least 5 AMD fans posting in this thread, if you use both, you're still an Intel Fan. If you weren't an Intel Fanboy or Girl you'd only buy AMD=P

Most folks in this forum will buy Nehalem if it rocks but if AMD pulled something out of its anus, we'd be on it as well=P Just as I skipped Presshot and bought a 3500+.