Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: SURPRISING RESULTS P4 Xeon dual 2.8Ghz

Author: Vincent Diepeveen

Date: 07:48:49 12/17/02

Go up one level in this thread


On December 17, 2002 at 10:31:02, Robert Hyatt wrote:

>On December 17, 2002 at 10:10:46, Vincent Diepeveen wrote:
>
>>Hello,
>>
>>Some tests were performed in the USA, where some P4 Xeon dual 2.8Ghz
>>systems get delivered now. In Europe we can't get them yet and
>>most likely we don't want them either:
>>
>>Here are the results of DIEP at the Xeon 2.8Ghz dual ECC registered DDR ram.
>>
>>test 1: diep 4 processes. Of course HT enabled.
>>   181538 nps
>>
>>test 2: diep 2 processes. HT enabled.
>>   135924 nps
>>
>>test 3: diep 2 processes K7 1.6ghz (registered DDR ram all other settings
>>        identical to xeon dual setup):
>>   146555
>>
>>THE 2 TESTS NALIMOV DIDN'T OR COULDN'T WANT TO DO WITH CRAFTY
>>SOME WEEKS AGO REVEAL A BIG WEAKNESS OF HT/SMT:
>>
>>test 4: diep 2 processes. HT disabled.    171288 nps
>>
>>test 5 and 6: diep single cpu HT disabled and enabled were same speed
>>   92090  nps versus 92019 nps.
>>
>>First conclusion is that the system is profitting only from HT when you
>>use 4 processes at the same time, OTHERWISE IT IS A DISADVANTAGE IF
>>YOU MULTITHREAD, because see the big difference between 2 processes
>>running with HT turned on and off.
>
>So?  _some_ of us are smart enough to understand this even before trying
>to run it.  It has _already_ been discussed here.
>
>
>>
>>In itself when you have a program with just 2 threads which you
>>run on a dual it gets slower. My assumption is that the hardware reports
>>4 cpu's and that the software doesn't care at what cpu to schedule
>>the processes/threads. the result of that is that there is a 33% chance
>>that things get scheduled at a cpu which is already running a thread/process.
>
>
>
>It doesn't necessarily get slower.  And once you move to windows .new, or
>the new linux kernel gets fixed, this won't happen at all as both of these
>systems will understand that two threads need to run on two physical processors
>rather than on two logical processors on the same physical processor...
>
>what is your point?  This is a sudden revellation to you?  :)
>
>
>
>
>>
>>Resulting in a system where 1 cpu idles kind of shortly and 1 cpu is running
>>2 threads/processes.
>>
>>Actually the actual chance that the 2 processes are scheduled at
>>2 different processors (there is 4 processors for the OS
>>times 3 processors left for the second process is 12 different
>>schedulings) is: 8/12 = 2/3 = 66%. In short there is a disaster possibility
>>of 33%.
>
>Until you move to a newer operating system that understands hyper-threading.
>This is a known problem.  Windows .net supposedly has this fix already.  Linux
>is "in progress".
>
>_not_ a big deal.
>
>A _known_ issue for months now...
>
>
>>
>>Now the absolute speed from performance viewpoint. If the system idles
>>completely and then starts to run *exclusively* diep at 4 processors, then
>>the measured speedup as you can calculate is in the order of 11.4% for
>>SMT/HT.
>>
>>That's not so much actually. The loss by searching parallel is at most
>>parallel applications bigger than the win of 11.4%. In case of DIEP
>>i am on the lucky side and go for that 11.4% faster speed.
>
>Can't be.  If you get 11.4% speed increase, and you lose more than that in
>parallel search overhead, you have a problem on _normal_ SMP machines as well.

Indeed it is true that the first seconds the HT/SMT gives big problems
in speed. Only after a couple of minutes the speed shows. I see only
a speedup after a minute or 3 each position.

I need to add however that i could improve a few issues in this version
which could get that down to 1 minute but like you i doubt whether the 11.4%
of HT is worth it.

I prefer a dual AMD instead for the moment!

>
>>
>>Yet the sad confirmation is that the pessimistic expectation about the
>>absolute speed is completely confirmed. This system performs (assuming
>>lineair scaling) like a 1.98 Ghz dual K7.
>>
>>there are motherboards now which do not require registered memory and
>>the K7 runs already quite a while at 2.0Ghz in fact. Now i don't care
>>for XP at all here nor do i care for the P4 at all. I just care for
>>parallel search here.
>>
>>If we know that a 2.0Ghz dual K7 is identical to a dual 2.8Ghz Xeon
>>and that in the majority of cases the K7 is going to win, then considering
>>the huge price difference, the choice would be trivial for most who
>>are looking for a lot of computing power for little money.
>>
>>Doesn't take away the fact that the P4 is winning ground. I remember
>>the first dual AMD 1.2ghz test versus P4 dual 1.7Ghz and the AMD dual
>>being 20% faster. Meaning in short that the speed of a P4 was performing
>>about 1 : 1.7
>>
>>Now if i compare a dual Xeon 2.8Ghz with a 2Ghz K7 then it's equal
>>meaning the P4 is performing 1 : 1.4
>>
>>So that's a big step forward!
>
>This from someone that said these machines "do not exist" a week ago, too.
>
>So this is a _big_ step forward alright.  :)
>
>
>
>
>>
>>Whether the step is because of DDR ram versus the very bad performing
>>RDRAM (nearly 2 times slower latency) is a matter of open discussion.
>>
>>HT/SMT in itself is not so impressing now.
>>
>>It's trivial to say that it will get impressive when the P4 can split itself
>>into 2 real processors having little dependencies on each other.
>>
>>Right now the single cpu win on a P4 3.06Ghz HT (18%) is
>>clearly more than the older generation 2.8 Ghz HT/SMT. so it seems
>>also this technique is slowly winning in realism.
>>
>>Right now i can't take what's getting on the market now very serious.
>>
>>Best regards,
>>Vincent



This page took 0.01 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.