Socket 939 Sempron found........

[fierydemise]

Originally posted by: Pabster
On another note, I find it rather pathetic Tom still can't figure out how to cool off the P4. The average temperature is better now (around 68C) but that is still too high. It seems the fan is only averaging 3500rpm whilst the AMD rig's fan is >4500rpm. What gives with that? Also (perhaps coincidentally with the temperature drop) the Intel rig's performance curve is nearly even now instead of jagging up and down wildly. But I still assert that 68C as an average temperature is too high. He needs to get it down around 60C.

I have to agree with you that 68C seems too high, but THG is using the boxed cooler, if anyone knows what temperature their CPUs work at its Intel. This leads us to one of two conclusions;
1. Intel supplied an insufficient cooler
2. The Intel CPU will work adequately at 70C
Either way I think that they should restart the test now that they have everything stable, just so we can see some real results.

 

[AnandThenMan]

Having a cheap cooler is one thing, sure the CPU will get hot. But a cheap cooler should not be responsible for frying a motherboard! I have never seen a processor fry a motherboard, never. Well that's not true, I did see a Xeon bake a motherboard but it was not the fault of the processor, something happened to the voltage and it skyrocketed and cooked the CPU which in turn burned the mobo.

Personally I see the PEE 840 as a desperation product that is not ready for the mainstream. It is no wonder Intel does not offer any dual core chips in servers. They would do serious harm to their reputation.

 

[clarkey01]

Originally posted by: Pabster

Originally posted by: Duvie
I always felt in my testing my HT was qute consistent...so it seems hard for me to understand this flopping...so one could see my conspiracy theory on this...

Keep in mind that the HT behavior might well be different on the Smithfield dual cores than it is on Prescott. It is entirely possible that Intel has made a few revisions to it.

That said, I'll hope your "conspiracy theory" is simply that Tom has made a fool of himself and that the results are invalid. Considering Tom has 99% AMD advertising now I find it hard to believe he has any real incentive to hand Intel anything.

Smithfield is no more then prescott x 2. I doubt Intel have touched it since there's been zero talk about it. When prescott came out there was mention of it's new and improved hyperthreading, no such changes have happend for smithfield.

 

[Keysplayr]

Originally posted by: AnandThenMan
Having a cheap cooler is one thing, sure the CPU will get hot. But a cheap cooler should not be responsible for frying a motherboard! I have never seen a processor fry a motherboard, never. Well that's not true, I did see a Xeon bake a motherboard but it was not the fault of the processor, something happened to the voltage and it skyrocketed and cooked the CPU which in turn burned the mobo.

Personally I see the PEE 840 as a desperation product that is not ready for the mainstream. It is no wonder Intel does not offer any dual core chips in servers. They would do serious harm to their reputation.

Nobody with any credibiltiy would say that a cheap cooler fried my motherboard. I did see a processor burn. It was a Dual Athlon MP 2000+ and the fan on one of them failed. Literally turned the processor purplish blue and black and blistered the packagage. This was on a Tyan S2466 mobo. The computer was dead until I removed the cooked processor from cpu socket 2. Then it ran fine.

 

[PetNorth]

Have you noticed X2's heatsink used by Tom's? It is Ajigo MF043-044 heatsink

This is the heatsink supplied with, for example, Newcastle 3000+ box (s754):

http://www.madshrimps.be/printart.php?articID=135

I doubt this one (a low-mid end one) will be the boxed heatsink included with X2's family. I guess it will be the heatsink supplied with FX-55 box, this:

http://www.hothardware.com/image_popup....mage=big_cooler2.JPG&articleid=592&t=a

 

[imported_Opteron]

Originally posted by: PetNorth
Have you noticed X2's heatsink used by Tom's? It is Ajigo MF043-044 heatsink

This is the heatsink supplied with, for example, Newcastle 3000+ box (s754):

http://www.madshrimps.be/printart.php?articID=135

I doubt this one (a low-mid end one) will be the boxed heatsink included with X2's family. I guess it will be the heatsink supplied with FX-55 box, this:

http://www.hothardware.com/image_popup....mage=big_cooler2.JPG&articleid=592&t=a

SHHHHH !!! They change it and then there is 1 reboot for AMD.

That's what they want, so lets not give them any opportunities to spread crap !

AMD X2 will do just fine even with Sempron coolers

 

[PetNorth]

Originally posted by: Opteron

Originally posted by: PetNorth
Have you noticed X2's heatsink used by Tom's? It is Ajigo MF043-044 heatsink

This is the heatsink supplied with, for example, Newcastle 3000+ box (s754):

http://www.madshrimps.be/printart.php?articID=135

I doubt this one (a low-mid end one) will be the boxed heatsink included with X2's family. I guess it will be the heatsink supplied with FX-55 box, this:

http://www.hothardware.com/image_popup....mage=big_cooler2.JPG&articleid=592&t=a

SHHHHH !!! They change it and then there is 1 reboot for AMD.

That's what they want, so lets not give them any opportunities to spread crap !

AMD X2 will do just fine even with Sempron coolers

LOL

Yeah.. probably this crappy Ajigo HSF is supplied with Sempron too

 

[Keysplayr]

Originally posted by: PetNorth

Originally posted by: Opteron

Originally posted by: PetNorth
Have you noticed X2's heatsink used by Tom's? It is Ajigo MF043-044 heatsink

This is the heatsink supplied with, for example, Newcastle 3000+ box (s754):

http://www.madshrimps.be/printart.php?articID=135

I doubt this one (a low-mid end one) will be the boxed heatsink included with X2's family. I guess it will be the heatsink supplied with FX-55 box, this:

http://www.hothardware.com/image_popup....mage=big_cooler2.JPG&articleid=592&t=a

SHHHHH !!! They change it and then there is 1 reboot for AMD.

That's what they want, so lets not give them any opportunities to spread crap !

AMD X2 will do just fine even with Sempron coolers

LOL

Yeah.. probably this crappy Ajigo HSF is supplied with Sempron too

Yeah, but that's more than enough to run the much cooler X2's.
The Pentium D's need good, high CFM cooling. Dont forget, both of these cases have the sides removed so expect a couple of degrees higher than current. The faster spinning Intel fan should be sufficient for the average user, but not for even a sliver of o/cing.
Need one of those Vanessa fans.

 

[michal1980]

on another note, only the highest end pentium d chip, has ht , it be interesting to see what happens in this 'multi' thread test.

 

[boran]

Originally posted by: clarkey01

Originally posted by: Pabster

Originally posted by: Duvie
I always felt in my testing my HT was qute consistent...so it seems hard for me to understand this flopping...so one could see my conspiracy theory on this...

Keep in mind that the HT behavior might well be different on the Smithfield dual cores than it is on Prescott. It is entirely possible that Intel has made a few revisions to it.

That said, I'll hope your "conspiracy theory" is simply that Tom has made a fool of himself and that the results are invalid. Considering Tom has 99% AMD advertising now I find it hard to believe he has any real incentive to hand Intel anything.

Smithfield is no more then prescott x 2. I doubt Intel have touched it since there's been zero talk about it. When prescott came out there was mention of it's new and improved hyperthreading, no such changes have happend for smithfield.

The point here is, does winXP support dual HT cores ? We all know that winXP has support for hyperthreading, but supporting two cores with HT is another thing. then you dont have virtual processor 1 and 2 but you have 1, 2, 3 and 4 of which 1 and 2 are one real core and 3 and 4 one real core. and the OS has to take all into account. assigning threads should go like this:
first: 1
second: 3
fourth: 2 or 4 if 1 or 3 are not 100 % loaded, etc (this is still from assumption the OS cannot ask how loaded a CPU is, anyone more knowledgable may jump in and clarify, but I think the OS has to deduct by its workload handed out how loaded a CPU is.) so it has to see if virtual core 1 and 3 have space left and then assign it to the one with mosts space left.

as you can see the HT makes assigning threads not that easy and would require some logic, which is quite taxing for such a very basic item of a multitask OS that I think it will impact performance.

in general i'll repeat my statement that the winXP thread sheduler sees the virtual cores as real one and acts like they are, because the concensus now is that the divX thread is set to low priority which means it should barely get any work done, the fact that it does get so much work done, and the big differences in other areas make me think that the divX thread is hampering intel in the other areas. which is essentially a bad thing, you dont want something set at idle priority to interfere with items you assigned a higher priority.

 

[Keysplayr]

Question: Is the Intel system literally 23 hours behind the AMD system? I mean, since they reset the reboots and stats, the AMD rig has been up just about 93 hours and the Intel rig, because of issues has been up about 70 hours now. So is it accurate to say that the Intel systems scores represent 23 hours of downtime? If so, I am impressed with the current scores. Yes, I know this would be different with only 2 apps running, (I think).
But who here thinks this Intel chip is a slouch? Yes, we know its hot, but it seems to be stable at around 70C at full load. So, who thinks it is a slouch and why?..

Thanks.

 

[boran]

that time is the uptime, so time between now and the last reboot. there is no measurement of the actual downtime. which is another flaw in the test imho. in fact they should put a big warning message on the results as in: warning these numbers are only here so you can see that these things actually do something. because each number on it's own is worthless, all 8 numbers + the downtime are one solid item. you cannot draw conclusions about any performance related items there.

 

[Keysplayr]

Originally posted by: boran
that time is the uptime, so time between now and the last reboot. there is no measurement of the actual downtime. which is another flaw in the test imho. in fact they should put a big warning message on the results as in: warning these numbers are only here so you can see that these things actually do something. because each number on it's own is worthless, all 8 numbers + the downtime are one solid item. you cannot draw conclusions about any performance related items there.

So, what does AMD systems 93 hour uptime mean?
And what does Intels 70 hour uptime mean?

If I recall correctly, both systems were restarted when toms thought they had the bugs worked out of the Intel system. The AMD system has been running 93 hours since that point correct? The Intel system as you know has had 3 reboots since then with a lot of downtime. Please explain why this does not mean anything. Thanks.

 

[PetNorth]

Originally posted by: keysplayr2003
who here thinks this Intel chip is a slouch? Yes, we know its hot, but it seems to be stable at around 70C at full load. So, who thinks it is a slouch and why?..

For many people (I'm included) this is a mayor and unsavable issue. I'm a silentPC fan (right now, I enjoy an A64 Newcastle 3000+ with Thermalright XP-120 & Papst 120mm 600-700rpm; passive PSU, passive GPU; and a Pentium M system with Papst 120mm 600-700rpm, PSU and GPU passive right now). These configurations are unimaginable with Pentium EE & D. In the other hand, I'm sure this configuration is possible with Athlon 64 X2. I'll check if it is soon

 

[Markfw]

Originally posted by: PetNorth

Originally posted by: keysplayr2003
who here thinks this Intel chip is a slouch? Yes, we know its hot, but it seems to be stable at around 70C at full load. So, who thinks it is a slouch and why?..

For many people (I'm included) this is a mayor and unsavable issue. I'm a silentPC fan (right now, I enjoy an A64 Newcastle 3000+ with Thermalright XP-120 & Papst 120mm 600-700rpm; passive PSU, passive GPU; and a Pentium M system with Papst 120mm 600-700rpm, PSU and GPU passive right now). These configurations are unimaginable with Pentium EE & D. In the other hand, I'm sure this configuration is possible with Athlon 64 X2. I'll check if it is soon

Exactly ! I am now in pusuit of the silent PC. Just got 2 XP90's yesterday, a 9600XT with a passive hsf, and a Sonata case for the two in my bedroom that I work on.

 

[Keysplayr]

PetNorth and Mark, I appreciate your input about silent PC's. But that's not what I asked. I asked if anyone here thinks the Intel chip is a slouch. Regardless of temperature and noise which matter to some but not all. I'm not asking about how noisy it is or how hot it gets. Stop dancing around my question and answer it. And don't say it can't be answered because there are too many variables. Is it performing well or is it not performing well?
JEEZ, I have to be an attorney to ask a question around here. I don't mean to get snippy, just tired of people incorrectly answering questions with answers to totally different questions.

 

[Markfw]

OK, The heat is SO bad and the power consumption SO high, that if it doesn't affect the motherboard in a week, in a year it will slowly melt it ! Yes, if you ignore that, the $1,000 chip is a close competetor to the 4800 when 4 heavy tasks are running. On two, the 4800 kills it ! And as a usable chip/platform, the Optern dual-core and X2 really win big time. This is why everybody is upset with tom (amoungst others) that he picked the absolute best scenario to show the EE duallie as a competetor, and it still fails to do nothing but tie AT BEST !

 

[Keysplayr]

CD Encoding

Intel : 494 CD's divided by 73 hours = 6.76 CD's per hour
AMD : 472 CD's divided by 95 hours = 4.96 CD's per hour

WINRAR

Intel : 2695 archives divided by 73 hours = 36.91 rar's per hour
AMD : 3665 archives divided by 95 hours = 38.57 rar's per hour

FarCry

Intel : 1719 runs divided by 73 hours = 23.54 runs per hour
AMD : 2056 runs divided by 95 hours = 21.64 runs per hour

Divx

Intel : 2240 minutes divided by 73 hours = 30.68 minutes encoded every hour
AMD : 170 minutes divided by 95 hours = 1.78 minutes encoded every hour

 

[imported_Opteron]

does winXP support dual HT cores ?

YES and NO..

Windows XP HOME supports 1 physical + 1 logical CPU's = 2 CPU's total.

Windows XP PRO supports 2 physical + 2 logical CPU's = 4 CPU's total.

But has Microsoft changed their policies after dual core was released ?

I don't think so !

XP HOME would become WORKSTATION OS if it supports 2 CPU's = M$ would loose revenues.

btw. Microsoft is selling software licences per socket and not by CPU.

 

[Markfw]

You forgot. They were BOTH running 95 hours, just uptime on the intel is 73....Try again The counts were not changed to 0, just the uptime.

 

[Keysplayr]

Originally posted by: Markfw900
You forgot. They were BOTH running 95 hours, just uptime on the intel is 73....Try again The counts were not changed to 0, just the uptime.

I don't know about you, but the opposite of uptime, to me anyway, is downtime. Or not running.

Maybe you can explain what you meant because that really made little sense. Thanks.

 

[Markfw]

The counts on the intel were over the same 95 hour period, with only an hour or two to change out hardware. Use the same divisor for both, 95 !
Intel : 494 CD's divided by 95 hours = 5.2 CD's per hour
AMD : 472 CD's divided by 95 hours = 4.968421053 CD's per hour 0.955465587 4% behind

WINRAR

Intel : 2695 archives divided by 95 hours = 28.36842105 rar's per hour
AMD : 3665 archives divided by 95 hours = 38.57894737 rar's per hour 1.359925788 36% ahead

FarCry

Intel : 1719 runs divided by 95 hours = 18.09473684 runs per hour
AMD : 2056 runs divided by 95 hours = 21.64210526 runs per hour 1.196044212 20% ahead

Divx

Intel : 2240 minutes divided by 95 hours = 23.57894737 minutes encoded every
AMD : 170 minutes divided by 95 hours = 1.789473684 minutes encoded every 0.075892857 93% behind

 

[Keysplayr]

Originally posted by: Markfw900
The counts on the intel were over the same 95 hour period, with only an hour or two to change out hardware. Use the same divisor for both, 95 !

I don't think so. You are talking about total stable time, not performance.
At least I think you are.

So when Toms reset the uptime to 0 on both systems, why now is there a 23 hour difference between the AMD and Intel rig? Because the intel rig was down? right? Offline?

What am I missing here that is so obvious to you? Please.

What your saying to me is, the INtel system, although says it only was up for 73 hours, was indeed processing full load for 95 hours? Or just as long as the AMD rig? How is that happening?

 

[Markfw]

like this up for 5 hours, down 1, up 1 hour , down 5 minutes, etc... The units started 95 hours ago, but the last re-boot for the Intel was 73 hours ago.

 

[Keysplayr]

Originally posted by: Markfw900
like this up for 5 hours, down 1, up 1 hour , down 5 minutes, etc... The units started 95 hours ago, but the last re-boot for the Intel was 73 hours ago.

Ahh ok. Now all we need to know is exactly how long the Intel system was down in the last 95 hours. I understand now. Thanks. Maybe we can tell from the charts.