I understand most of the basics about hardware, like DIMMs and FSB, and
HD/RPM, caches, etc. and this is dangerous, especially when I decide to
build my own computers, lol. Anyways, here's my tale and questions.
I decided about six months ago to "upgrade" my computer. I thought I
was in a 89' buick compared to what I could be in. And the buick, aka,
my system, was actually pretty good. 2GB RAM, 3.2 GHz Pentium with HT
(< this is important later), and a few too many harddrives. So trying
to get in gear with teh times, I decided to get a new motherboard, new
harddrives (using SATA inplace of my IDE mess), and the crown jewel, a
2.33 GHz Dual Core Pentium. Now, I had done my research on HT versus
Dual Core, and I was convinced that I was going to see a nice biiiiiig
bump in performace, especially that my FSB was huge compared to the old
FSB, and my RAM had better ratings than before.
So I spend two weeks putting the thing together, working out the bugs,
this and that. And when its all said and done, it runs like molasses.
My 3.2GHz w/HT ran WAAAAAYYYY better than the dual core. I am soooo
ticked at myself.
So, here are my questions:
One, what did I do or think wrong..... and what should I know now that
I should have known before I bought the Dual Core?
Two, The Dual Core Pentiums come in three (or four, maybe) flavors. Ya
have your Dual Core, Dual Core Extreme, and then Quad Core (and possibly
Quad Core Extreme sometime soon). Is the quad core simply a dual core
with HT? or is Dual Come Extreme using HT along with its unlocked
ratio?
Three, should I simply forget that there ever was a thing as HT? or do
you think Intel is going to be using it, or is using it someplace in the
future?
"ldiaco" <ldiaco.33jgg4@no.email.invalid> wrote in message
news:ldiaco.33jgg4@no.email.invalid...
>
> Hey all,
>
> I understand most of the basics about hardware, like DIMMs and FSB, and
> HD/RPM, caches, etc. and this is dangerous, especially when I decide to
> build my own computers, lol. Anyways, here's my tale and questions.
<snip>
> So I spend two weeks putting the thing together, working out the bugs,
> this and that. And when its all said and done, it runs like molasses.
> My 3.2GHz w/HT ran WAAAAAYYYY better than the dual core. I am soooo
> ticked at myself.
>
> So, here are my questions:
>
> One, what did I do or think wrong..... and what should I know now that
> I should have known before I bought the Dual Core?
What operating system are you using? If your new pc runs Vista then it will
run "like molasses" regardless of the hardware. (Hopefully it will be better
when SP 1 comes out.)
> Two, The Dual Core Pentiums come in three (or four, maybe) flavors. Ya
> have your Dual Core, Dual Core Extreme, and then Quad Core (and possibly
> Quad Core Extreme sometime soon). Is the quad core simply a dual core
> with HT? or is Dual Come Extreme using HT along with its unlocked
> ratio?
My understanding is that Quad core is 4 cores, so its not dual core with HT.
HT is a technique for getting more out of each core by letting the
instruction pipeline work on instructions for other threads. Can't answer
the rest, sorry.
> Three, should I simply forget that there ever was a thing as HT? or do
> you think Intel is going to be using it, or is using it someplace in the
> future?
HT always seemed quite a clever idea to me, but it does seem to have
disappeared from the promotional literature. I've no idea what its future
is, but it would seem a shame if it has no place in the future.
--
Brian Cryer www.cryer.co.uk/brian
ldiaco wrote:
> Hey all,
>
> I understand most of the basics about hardware, like DIMMs and FSB, and
> HD/RPM, caches, etc. and this is dangerous, especially when I decide to
> build my own computers, lol. Anyways, here's my tale and questions.
>
> I decided about six months ago to "upgrade" my computer. I thought I
> was in a 89' buick compared to what I could be in. And the buick, aka,
> my system, was actually pretty good. 2GB RAM, 3.2 GHz Pentium with HT
> (< this is important later), and a few too many harddrives. So trying
> to get in gear with teh times, I decided to get a new motherboard, new
> harddrives (using SATA inplace of my IDE mess), and the crown jewel, a
> 2.33 GHz Dual Core Pentium. Now, I had done my research on HT versus
> Dual Core, and I was convinced that I was going to see a nice biiiiiig
> bump in performace, especially that my FSB was huge compared to the old
> FSB, and my RAM had better ratings than before.
>
> So I spend two weeks putting the thing together, working out the bugs,
> this and that. And when its all said and done, it runs like molasses.
> My 3.2GHz w/HT ran WAAAAAYYYY better than the dual core. I am soooo
> ticked at myself.
>
> So, here are my questions:
>
> One, what did I do or think wrong..... and what should I know now that
> I should have known before I bought the Dual Core?
>
> Two, The Dual Core Pentiums come in three (or four, maybe) flavors. Ya
> have your Dual Core, Dual Core Extreme, and then Quad Core (and possibly
> Quad Core Extreme sometime soon). Is the quad core simply a dual core
> with HT? or is Dual Come Extreme using HT along with its unlocked
> ratio?
>
> Three, should I simply forget that there ever was a thing as HT? or do
> you think Intel is going to be using it, or is using it someplace in the
> future?
To eliminate any confusion about what you bought, you could quote a
model number, like E6600, E2180, or even an SLxxx code (printed on the
processor box). The thing you bought, could be a Core2 family processor.
For example, it might be an E6550.
The Core2 2.33GHz, you'd multiply that by 1.5x to 2x, to get some idea of
its "Pentium 4 equivalent" speed. It should feel a little faster
than what you had. And not molasses. More like 3.5 to 4.5GHz or so
in P4 equivalents.
In the past, I've run into a couple people, whose systems seemed to be
stuck at the low multiplier. With EIST (equivalent in its way, to
Cool N' Quiet on AMD), the multiplier changes as a function of system
load. Now, I never did find a definitive step that was missing in their
build. The builders eventually concluded it was fixed, but since they
changed so many things at once, they lost track of what fixed it.
You can use RMClock, to monitor the system. RMClock is supposed to be
able to monitor for throttling, if it is present. I don't really like
the graphical output, but it is better than nothing.
"Molasses" can come from a number of things. For example, say your
hard drive was running in PIO mode. The processor would waste a lot
of cycles, any time a file is referenced (since it would use polled
transfer). The bandwidth would be limited (4MB/sec to 8MB/sec).
Programs would be slow to load, but fast once started. You can
test with HDTach or HDTune, and get some idea of the actual speed.
My disk speed ranges from 60MB/sec to 40MB/sec sustained, from the beginning
to the end of the disk. More modern drives manage 70MB/sec or a
little more, in sustained performance. The performance graph
should be a curve - a flat line means cable/interconnect trouble.
Now, if you were to test with SuperPI 1.5 or something similar,
that would spend a good deal of its time using nothing but
processor and memory. Comparing the old system to the new system,
would then give an idea if the CPU and memory was the limitation.
Just the "feel" of the system, may not be narrowing down where the
problem is. (On the machine I'm typing on, SuperPI 1.5 will do the
1 million digits test in 50 seconds. It became that slow, after
I added my antivirus software. I think I might have had maybe
45 seconds at one time, before the antivirus was added. Just if you
want to compare. SuperPI is single threaded AFAIK. My system is a
Northwood P4 3.15GHz/FSB900/DDR450 dual channel. A highly overclocked
Core2 can do the same benchmark in sub 10 seconds, to give some idea
of how good it can be. So my system is 5x out of date.)
The multiplier is one issue. Processor throttling (CPU instruction rate
drop, to try to stave off overheating), is another mechanism for
losing performance. You want enough cooling, to stay below 70C. About
65C to 70C, is where some of the newer processors start to throttle.
The cooling solution should stay clear of those kinds of temps, in
order to get 100% of the goodness.
Keep the RMClock window open, then run SuperPI, and watch the graph
while it runs.
You can also have a look through this thread, but first I'd prefer
to see you get a handle on exactly what is slow, before going
crazy with unrelated fixes.
"ldiaco" <ldiaco.33jgg4@no.email.invalid> schreef in bericht
news:ldiaco.33jgg4@no.email.invalid...
>
> Hey all,
>
> I understand most of the basics about hardware, like DIMMs and FSB, and
> HD/RPM, caches, etc. and this is dangerous, especially when I decide to
> build my own computers, lol. Anyways, here's my tale and questions.
>
> I decided about six months ago to "upgrade" my computer. I thought I
> was in a 89' buick compared to what I could be in. And the buick, aka,
> my system, was actually pretty good. 2GB RAM, 3.2 GHz Pentium with HT
> (< this is important later), and a few too many harddrives. So trying
> to get in gear with teh times, I decided to get a new motherboard, new
> harddrives (using SATA inplace of my IDE mess), and the crown jewel, a
> 2.33 GHz Dual Core Pentium. Now, I had done my research on HT versus
> Dual Core, and I was convinced that I was going to see a nice biiiiiig
> bump in performace, especially that my FSB was huge compared to the old
> FSB, and my RAM had better ratings than before.
>
> So I spend two weeks putting the thing together, working out the bugs,
> this and that. And when its all said and done, it runs like molasses.
> My 3.2GHz w/HT ran WAAAAAYYYY better than the dual core. I am soooo
> ticked at myself.
>
> So, here are my questions:
>
> One, what did I do or think wrong..... and what should I know now that
> I should have known before I bought the Dual Core?
>
Per core, I don't think there's much performance difference between your
old and new processor.
I'd running a duo core at 2.4Ghz and compared to a 3.2 HT, the duo core was
noticably faster but I didn't do a real one core cpu test.
I just want to say that it felt much faster right out of the box.
Most programs are not optimized for multi-threading.
Those programs are running on just one core so they don't get any advantage
from the other cores.
You should notice an improvent when running several somewhat cpu intensive
programs at once.
> Two, The Dual Core Pentiums come in three (or four, maybe) flavors. Ya
> have your Dual Core, Dual Core Extreme, and then Quad Core (and possibly
> Quad Core Extreme sometime soon). Is the quad core simply a dual core
> with HT? or is Dual Come Extreme using HT along with its unlocked
> ratio?
The dual core processors (both normal and extreme) are 'true' dual core,
meaning: an optimized design for two cores.
The quad core processors (again, normal and extreme) are two dual cores
merged together in one package.
Some say this isn't 'true' quad, other say: who cares!
Anyway, quad cores have four real cores while hyperthreading is just a
pipeline 'trick' to get a bit more processing power.
Extreme processors (both dualcore and quadcore) have an unlocked
multiplier.
All conroe processors can decrease the multiplier.
Only the extremes can increase the multiplier.
>
> Three, should I simply forget that there ever was a thing as HT? or do
> you think Intel is going to be using it, or is using it someplace in the
> future?
>
Yes, as far as I know, hyperthreading is gone.
Maybe a part of it is still used for optimization, don't really know.
One other advantage of conroe compared to pentium:
conroes can be hugely overclocked.
You can easily increase your FSB and get the processor running
at 2.6 or 2.8GHz.
Just make sure your memory can handle the FSB and your cpu
cooler is up to the task!
(With good cooling you can get much higher, 3.0 or even 3.2
is do-able)
I don't know why your system runs a lot slower than your previous
system.
My guess is there is something wrong with your setup.
How is your harddrive performance?
If you copy a file from/to your harddisk, is your cpu usage ramping
up to 40% or more? If that is true, you might have a problem with
an IDE device and since the harddisk is accessed a lot, your whole
system will suffer.
The same for graphics drivers. When you drag a window, do you get
a smooth update or is everything very sluggish?
"ldiaco" <ldiaco.33jgg4@no.email.invalid> wrote in message
news:ldiaco.33jgg4@no.email.invalid...
>
> Hey all,
>
> I understand most of the basics about hardware, like DIMMs and FSB, and
> HD/RPM, caches, etc. and this is dangerous, especially when I decide to
> build my own computers, lol. Anyways, here's my tale and questions.
>
> I decided about six months ago to "upgrade" my computer. I thought I
> was in a 89' buick compared to what I could be in. And the buick, aka,
> my system, was actually pretty good. 2GB RAM, 3.2 GHz Pentium with HT
> (< this is important later), and a few too many harddrives. So trying
> to get in gear with teh times, I decided to get a new motherboard, new
> harddrives (using SATA inplace of my IDE mess), and the crown jewel, a
> 2.33 GHz Dual Core Pentium. Now, I had done my research on HT versus
> Dual Core, and I was convinced that I was going to see a nice biiiiiig
> bump in performace, especially that my FSB was huge compared to the old
> FSB, and my RAM had better ratings than before.
I assume molasses is/are bad - never heard of it/them?
Either I am missing something, you have phrased your question badly, or you
really have upgraded from a 3.2 P4 to a 2.33 P4 with 2 cores and you are
wondering why it is slower??? 3.2GHz down to 2.33GHz = slower! This is not
an upgrade!
You don't tell us the model of the new CPU. I suspect you have actually
bought a 'Core 2 Duo' chip, but you state it to be a P4 dual core. A Core 2
Duo at 2.33 GHz (probably an 'e6600') should be faster than the 3.2GHz P4 at
most tasks.
> So I spend two weeks putting the thing together, working out the bugs,
> this and that. And when its all said and done, it runs like molasses.
> My 3.2GHz w/HT ran WAAAAAYYYY better than the dual core. I am soooo
> ticked at myself.
>
> So, here are my questions:
>
> One, what did I do or think wrong..... and what should I know now that
> I should have known before I bought the Dual Core?
You should have known what processor you were buying and read performance
tables comparing the old and new processors. TomsHardware.com has loads of
these.
> Two, The Dual Core Pentiums come in three (or four, maybe) flavors. Ya
> have your Dual Core, Dual Core Extreme, and then Quad Core (and possibly
> Quad Core Extreme sometime soon). Is the quad core simply a dual core
> with HT? or is Dual Come Extreme using HT along with its unlocked
> ratio?
P4s and Pentiums, then Dual Core P4/Pentiums went out ages ago. They were
replaced by Core Duo about 18 months ago, then superceeded by the Core 2 Duo
about 12+ months or so ago. The latest Intel processors are called Core 2
Duo and Core 2 Duo Quad. If you have really bought a Dual Core P4 at
2.33GHz, then you have bought a 3yr old processor. It will be slow compared
to the P4 at 3.2GHz that you had before.
> Three, should I simply forget that there ever was a thing as HT? or do
> you think Intel is going to be using it, or is using it someplace in the
> future?
Forget it - Dual and Quad cores are genuine multi-processor units, HT simply
used spare clock cycles on a single processor to 'make out' that there were
two.
"ldiaco" <ldiaco.33jgg4@no.email.invalid> wrote in message
news:ldiaco.33jgg4@no.email.invalid...
>
> Hey all,
>
> I understand most of the basics about hardware, like DIMMs and FSB, and
> HD/RPM, caches, etc. and this is dangerous, especially when I decide to
> build my own computers, lol. Anyways, here's my tale and questions.
>
> I decided about six months ago to "upgrade" my computer. I thought I
> was in a 89' buick compared to what I could be in. And the buick, aka,
> my system, was actually pretty good. 2GB RAM, 3.2 GHz Pentium with HT
> (< this is important later), and a few too many harddrives. So trying
> to get in gear with teh times, I decided to get a new motherboard, new
> harddrives (using SATA inplace of my IDE mess), and the crown jewel, a
> 2.33 GHz Dual Core Pentium. Now, I had done my research on HT versus
> Dual Core, and I was convinced that I was going to see a nice biiiiiig
> bump in performace, especially that my FSB was huge compared to the old
> FSB, and my RAM had better ratings than before.
>
> So I spend two weeks putting the thing together, working out the bugs,
> this and that. And when its all said and done, it runs like molasses.
> My 3.2GHz w/HT ran WAAAAAYYYY better than the dual core. I am soooo
> ticked at myself.
>
> So, here are my questions:
>
> One, what did I do or think wrong..... and what should I know now that
> I should have known before I bought the Dual Core?
>
> Two, The Dual Core Pentiums come in three (or four, maybe) flavors. Ya
> have your Dual Core, Dual Core Extreme, and then Quad Core (and possibly
> Quad Core Extreme sometime soon). Is the quad core simply a dual core
> with HT? or is Dual Come Extreme using HT along with its unlocked
> ratio?
>
> Three, should I simply forget that there ever was a thing as HT? or do
> you think Intel is going to be using it, or is using it someplace in the
> future?
>
>
>
A Dual Core Pentium is just a P4 with 2 cores. What you did was the
equivalent of going from a 3.2 GHz P4 to a 2.33 GHz P4. The dual core
would only show a performance gain in a multithreaded application.
On the other hand a 2.33 GHz Core 2 Duo even running single core
applications would leave the 3.2 GHz P4 in the dust. The reason is
that the Core 2 Duo performs more instructions per clock cycle than
the P4.
l> So I spend two weeks putting the thing together, working out the bugs,
l> this and that. And when its all said and done, it runs like molasses.
l> My 3.2GHz w/HT ran WAAAAAYYYY better than the dual core. I am soooo
l> ticked at myself.
i'm pretty sure you wouldn't notice CPU speed in "normal" applications
unless it's really *very* slow.
1 GHz is enough.
so, can you please tell us what you do and why do you think it's slow?
probably you have problems with other components, or problems with CPU --
it's better to run diagnostic utilities to see what happens
l> One, what did I do or think wrong..... and what should I know now that
l> I should have known before I bought the Dual Core?
differences should be very subtle -- you will notice them only on heavy
tasks such as video/audio encoding, rendering etc. probably in some games..
i've myself switched from Pentium4 3 GHz w/HT to Intel Core 2 Duo E6400 2.1
GHz, and it gave me a huge benefit in calculations i was doing.
as for normal usage, i don't think i can notice a difference..
G> P4s and Pentiums, then Dual Core P4/Pentiums went out ages ago. They
G> were replaced by Core Duo about 18 months ago, then superceeded by the
G> Core 2 Duo about 12+ months or so ago. The latest Intel processors are
G> called Core 2 Duo and Core 2 Duo Quad.
wrong. Intel sells low-end core2s having"Pentium" brand on them.
it has clock speed similar to Core 2 Duo E6300, but twice less cache and 800
MHz instead of 1066 MHz FSB.
i had no experience with such chips, but people say they are fine and can be
really good with overclocking. cache and FSB are not always useful, indeed..
there are also cheapest models called "Celeron" -- they have only 512 KB
cache and one core.
"ldiaco" wrote in message news:ldiaco.33jgg4@no.email.invalid...
>
> ... 3.2 GHz Pentium with HT (< this is important later), ...
> ... 2.33 GHz Dual Core Pentium ...
You expected faster processing with a slower clock?
> So I spend two weeks putting the thing together, working out the
> bugs,
> this and that. And when its all said and done, it runs like
> molasses.
At the POST screen, it should mention your CPU. Check what it states
for speed. Could be you don't have the BIOS setup correctly for
multiplier.
While the OS might make some use of the CPU, and it will probably be
handy with virtual machine programs where you can assign a CPU to a
particular VM, your applications won't get much boost. Don't know why
you went slower for the CPU clock. Multiple cores do not make up for
clock rate.
??>> ... 3.2 GHz Pentium with HT (< this is important later), ...
??>> ... 2.33 GHz Dual Core Pentium ...
V> You expected faster processing with a slower clock?
there is also such parameter as Instructions Per Cycle. and pentium4 is
known for very low IPC because of insanely long pipeline.
it works on full speed only when doing some boring loop (such as
cryptography or some kinds of processing).
for most workloads it's pipeline stalls due to branch mispredictions etc.
i'd say Pentium4 3 GHz compares to 1.3-1.4 OtherCPU (AMD ones, Intel Core..)
on typical workloads, and it can be even worse on some workloads..
i've got huge 4x boost for calculations i was doing when moved from 3 GHz
Pentium4 to 2.1 GHz Core2 Duo.
There never was a 2.33GHz dual-core Pentium D (Pentium 4 Family) the
slowest Pentium D was 2.8GHz if memory serves me correctly.
There is also no 2.33GHz family Pentium-Dual Core E2xxx, the fastest
is the 2.0GHz Pentium Dual-core E2200.
So the CPU in question is most likely a E6540 or E6550, both with 4MB
L@ cache and on the 1333MHz FSB - this chip, with good DDR2-667 or 800
RAM and a modern SATA hard drive should perform over 2-3 times better
than a Pentium 4 3.2GHz, even in single-threaded apps.
Likely causes of molasses effect? Hard drive issue (test speeds?)
speedstep issue (stuck at low speed?) or maybe overheating - is the
cooler correctly installed? Maybe the CPU is thermal throttling. I
doubt Vista would be the issue, as it performs flawlessly even on an
old Pentium 4 or Pentium D at 3GHz.
On Sun, 20 Jan 2008 23:13:45 -0600, ldiaco
<ldiaco.33jgg4@no.email.invalid> wrote:
>
>Hey all,
>
>I understand most of the basics about hardware, like DIMMs and FSB, and
>HD/RPM, caches, etc. and this is dangerous, especially when I decide to
>build my own computers, lol. Anyways, here's my tale and questions.
>
>I decided about six months ago to "upgrade" my computer. I thought I
>was in a 89' buick compared to what I could be in. And the buick, aka,
>my system, was actually pretty good. 2GB RAM, 3.2 GHz Pentium with HT
>(< this is important later), and a few too many harddrives. So trying
>to get in gear with teh times, I decided to get a new motherboard, new
>harddrives (using SATA inplace of my IDE mess), and the crown jewel, a
>2.33 GHz Dual Core Pentium. Now, I had done my research on HT versus
>Dual Core, and I was convinced that I was going to see a nice biiiiiig
>bump in performace, especially that my FSB was huge compared to the old
>FSB, and my RAM had better ratings than before.
Big bump in performance doing what (apps or windows GUI or
??)?
>
>So I spend two weeks putting the thing together, working out the bugs,
>this and that. And when its all said and done, it runs like molasses.
>My 3.2GHz w/HT ran WAAAAAYYYY better than the dual core. I am soooo
>ticked at myself.
What bugs?
Are you certain the heatsink is on good, that processor temp
reports look acceptibly low?
What is it doing when it runs like molasses?
For many tasks, the CPU and FSB is not the limitation, the
HDD is. Since you had "a few too many harddrives"
previously, have you changed their logical arrangement as to
what files are stored where, or not put these in the new
system, meaning you're using only one drive now?
Since you have focused on CPU and FSB, run a benchmark
(Google will find many) to establish if your CPU is
performing at the expected level. You can compare your
result to online results. Next do the same for the hard
drive.
>
>So, here are my questions:
>
>One, what did I do or think wrong..... and what should I know now that
>I should have known before I bought the Dual Core?
Hard to say since you haven't detailed when you see lower
performance.
>
>Two, The Dual Core Pentiums come in three (or four, maybe) flavors. Ya
>have your Dual Core, Dual Core Extreme, and then Quad Core (and possibly
>Quad Core Extreme sometime soon). Is the quad core simply a dual core
>with HT? or is Dual Come Extreme using HT along with its unlocked
>ratio?
Quad core is two dual-cores on the same die. Chosing dual,
dual extreme or quad should not have accounted for the issue
you have.
>
>Three, should I simply forget that there ever was a thing as HT? or do
>you think Intel is going to be using it, or is using it someplace in the
>future?
You might've already benefitted from it, but you haven't
even told us exactly what processor and motherboard you
have, nor the OS (relevant since Vista will run slower than
XP).
It would have been nice for you to reply earlier after your
first post so people aren't speculating so much, rather
focusing more on the specifics of the situation.
Wow, ..... I know I mentioned that I wasn't the smartest guy around with
hardware, but some people are touchy, lol. So to answer the question a
few people had asked me "Did you expect to get more speed with a slower
clock (3.2 to 2.33)" The answer is, yes, I did expect a faster
platform. From the digging around I did, I found out that the HT was
only helpful in certain tasks as far as speed goes, while it did nothing
for the CPU intensive tasks I normally perform.
My Hard drives, I tested them just now, and here's what came out.
Only one of them. Transfer Rates:
Minimum: 32.7MB/s
Max: 58.6MB/s
Average: 50.6MB/s
Access time: 18.6 ms
Burst Rate 135.7 MB/s
CPU usage: 14.2%
The other is exactly the same, only its hard to benchmark right now
cause I am accessing the disk as we speak. Its going all over the
place.
My graphics card is a Radeon X1650 512MB. Don't know what else you
might wanna know about it.
I'm running "RightMark Multi-Threading Memory Test" right now, and
Threads 0, and 1 are averaging around 14,750 MB/s - 14,950 MB/s, with a
total average BW of ~29,700 MB/s.
I am running some tasks that are made for multiple cores. One of them
is a graphics intensive game, and it is extremely choppy.
And now, I am running "RightMark CPU Clock Utility". Genuine Intel,
blah blah, Conroe, Rev. G0, Core Clock...... constantly switching
between 1999.8 MHz, and 2,333.12 MHz. Cpu Load is between 16 and 24%.
Throttle is 2300MHz or over. OS Load is reaching as high as 34.38%, and
as low at 6%. Core temp is 21 C. Current Multiplier (FID) is changing
between 6x & 7x. 6x is min, 7x is max.
I probably forgot some stuff, but i'll be back soon. Thanks for the
replies all.
In message <huGdncsTf5srNAnaRVnyiwA@pipex.net> "Brian Cryer"
<brian.cryer@127.0.0.1.ntlworld.com> wrote:
>HT always seemed quite a clever idea to me, but it does seem to have
>disappeared from the promotional literature. I've no idea what its future
>is, but it would seem a shame if it has no place in the future.
Not at all -- Hyperthreading was a hack for an insanely poor processor
design in the first place.
Unless we ever see a stupidly long pipeline like the P4 in the future,
HyperThreading will stay buried as a bad memory.
"Alex Mizrahi" <udodenko@users.sourceforge.net> wrote in message
news:4794ca66$0$90272$14726298@news.sunsite.dk...
> G> P4s and Pentiums, then Dual Core P4/Pentiums went out ages ago. They
> G> were replaced by Core Duo about 18 months ago, then superceeded by the
> G> Core 2 Duo about 12+ months or so ago. The latest Intel processors are
> G> called Core 2 Duo and Core 2 Duo Quad.
>
> wrong. Intel sells low-end core2s having"Pentium" brand on them.
>
> e.g.:
> Pentium Dual Core; E2160; 1.8GHz; 1M cache; 800MHz; LGA-775 box; INTEL;
> PN: E2160
>
> it has clock speed similar to Core 2 Duo E6300, but twice less cache and
> 800 MHz instead of 1066 MHz FSB.
> i had no experience with such chips, but people say they are fine and can
> be really good with overclocking. cache and FSB are not always useful,
> indeed..
>
> there are also cheapest models called "Celeron" -- they have only 512 KB
> cache and one core.
Fair enough, but is there a 2.33GHz Dual Core Pentium for sale at the
moment? This is the chip he has put in his system. I still think he means a
Core 2 Duo and probably an e6600, but until he comes back with more info we
can't help!
??>> HT always seemed quite a clever idea to me, but it does seem to have
??>> disappeared from the promotional literature. I've no idea what its
??>> future is, but it would seem a shame if it has no place in the future.
D> Not at all -- Hyperthreading was a hack for an insanely poor processor
D> design in the first place.
D> Unless we ever see a stupidly long pipeline like the P4 in the future,
D> HyperThreading will stay buried as a bad memory.
SMT is widely used to hide memory latencies -- when one thread is blocked
waiting for memory operation (cache miss), another thread gets executed.
this gives significant performance boost on some workloads, and is used in
high-end processors like POWER, Itanium, UltraSPARC. especially in
UltraSPARC T2 -- they run 8 threads per core, have very small cache, and
beat any other processor in performance-per-mhz factor on webserver
workload.
HyperThreading is an extended form of SMT -- it works not only on cache
misses, but in general course, allowing to use execution units
simultaneously. probably other processor architectures doesn't need this
much -- they don't have lots of spare execution units, or it's not worth
complexity. but it's not totally different -- on UltraSPARC T2 one thread
can run expensive FPU operations while other do simple arithmetic.
there are rumours
(http://www.intel.com/pressroom/archi...070328fact.htm) that
Intel is planning to return some form of SMT on desktop processors, but it's
not clear if that would be mere SMT on cache misses, or HT sharing execution
units..
> Fair enough, but is there a 2.33GHz Dual Core Pentium for sale at the
> moment? This is the chip he has put in his system. I still think he means a
> Core 2 Duo and probably an e6600, but until he comes back with more info we
> can't help!
No, the fastest Pentium Dual-Core is the E2200 at 2.2GHz (not 2.0GHz
like I said earlier..)
The only current 2.33GHz Intel Dual-Core CPU's are the Core 2 Duo
E6540 and E6550
"AdenOne" <pacific-one@hotmail.com> wrote in message
news:cb320c59-35ff-4b96-be3c-7f5aae757180@h11g2000prf.googlegroups.com...
> There never was a 2.33GHz dual-core Pentium D (Pentium 4 Family) the
> slowest Pentium D was 2.8GHz if memory serves me correctly.
>
> There is also no 2.33GHz family Pentium-Dual Core E2xxx, the fastest
> is the 2.0GHz Pentium Dual-core E2200.
>
> So the CPU in question is most likely a E6540 or E6550, both with 4MB
> L@ cache and on the 1333MHz FSB - this chip, with good DDR2-667 or 800
> RAM and a modern SATA hard drive should perform over 2-3 times better
> than a Pentium 4 3.2GHz, even in single-threaded apps.
I have tested a P4 3.2 and a Core 2 e6400 2.13GHz in the same PC.
Performance is about the same. Obviously multi-threaded apps are a different
story, but I don't use any. Neither do most users of multiple core machines!
> I have tested a P4 3.2 and a Core 2 e6400 2.13GHz in the same PC.
> Performance is about the same. Obviously multi-threaded apps are a
> different story, but I don't use any. Neither do most users of multiple
> core machines!
That isn't important, really. There are always quite a bit of processes
running simultaneously and any regular user performing regular tasks
usually ends up running more than a single application at a time. If that
user is using a multi-processor system running a decent operating system
then he will experience a nice performance boost.
On Tue, 22 Jan 2008 18:45:32 +0000, Rui Maciel
<rui.maciel@gmail.com> wrote:
>GT wrote:
>
>> I have tested a P4 3.2 and a Core 2 e6400 2.13GHz in the same PC.
>> Performance is about the same. Obviously multi-threaded apps are a
>> different story, but I don't use any. Neither do most users of multiple
>> core machines!
>
>That isn't important, really.
Yes it's very important.
>There are always quite a bit of processes
>running simultaneously and any regular user performing regular tasks
>usually ends up running more than a single application at a time.
Which means very little because there are always far more
processes than cores either way, and when apps are sitting
in the background idling they tend to be at lower priority
and consuming very little CPU time at all.
>If that
>user is using a multi-processor system running a decent operating system
>then he will experience a nice performance boost.
False. There are two scenarios where it makes a significant
difference, well actually 3 but I discount the first as
being bad software that should be replaced.
1) Applications running at inproper priority level taking
processor time when they should not. Manually reduce the
priority level or replace the app.
2) Linearly processed jobs that run in the background will
finish sooner. For example video encoding.
3) Well written, multithreaded application.
For general system use like office, email, internet, etc,
the difference is very little except a perceived increase in
"snappiness" which comes from the OS being in a poorly
tiered priority state as well as some applications can be.
>>That isn't important, really.
>
> Yes it's very important.
It only appears that it's very important to those who fail to understand how
any competent operating system manages it's processes. For a brief
explanation please read below.
>>There are always quite a bit of processes
>>running simultaneously and any regular user performing regular tasks
>>usually ends up running more than a single application at a time.
>
> Which means very little because there are always far more
> processes than cores either way, and when apps are sitting
> in the background idling they tend to be at lower priority
> and consuming very little CPU time at all.
That's perfectly irrelevant. The operating system manages the processing
time allocated to each process, which also includes allocating a certain
process to a certain processor. If you have a set of processes running,
including a process that demands a considerable amount of processor time,
the operating system will allocates CPU time slices so that every process
runs according to its needs and according to the available resources. That
means that the operating system can dedicate one core to run the heavy
process and run all the other processes on the remaining processors. In
certain operating systems like Solaris you can even force the OS to bind a
process to a certain processor.
That is not possible in a single processor system. In that case, the only
thing that the operating system can possibly do is keep on executing all
the processes in that single CPU, constantly switching between them, one
after the other and putting the remaining processes on hold for a given
time until their time is up and they can get CPU time once again. That
waiting time is drastically reduced in multi-processor systems, specially
if the operating system levels the work load through all processors.
Switching a process also comes with a performance penalty. As it is easy to
understand, if your process does not have to wait in line for all the other
process to run to get it's processor time and yours processors don't have
to spend as much time switching between processes, that processor will end
up spending less time doing managerial tasks and ends up doing more work in
less time. That translates to a nice performance boost.
>>If that
>>user is using a multi-processor system running a decent operating system
>>then he will experience a nice performance boost.
>
> False. Â*There are two scenarios where it makes a significant
> difference, well actually 3 but I discount the first as
> being bad software that should be replaced.
>
> 1) Â*Applications running at inproper priority level taking
> processor time when they should not. Â*Manually reduce the
> priority level or replace the app.
>
> 2) Â*Linearly processed jobs that run in the background will
> finish sooner. Â*For example video encoding.
>
> 3) Â*Well written, multithreaded application.
>
> For general system use like office, email, internet, etc,
> the difference is very little except a perceived increase in
> "snappiness" which comes from the OS being in a poorly
> tiered priority state as well as some applications can be.
It appears that you don't quite grasp how the operating system manages the
process' processor time, which explains why you don't understand the impact
that a multi-processor system has on the system's performance.
First of all, you fail to understand that all running processes demand
processor time. All of them. Then you fail to understand that the operating
system manages the system's processors according to a number of criteria,
which include levelling the system's work load and the application's
performance. Another thing that you fail to understand is that multiple
threads/processes are also used to deal with blocking calls, which means
that applications that read files and have to deal with network connections
do end up using multiple threads/processes. When you start holding those
concepts in consideration, you start to understand the advantages that come
with having more than one processor.
Then you need to understand another thing. You may already know that all
regular, every day applications like email, word processor, spreadsheet and
even web browser aren't constantly very demanding in terms of processor
time. As normal GUI applications, you end up with isolated peaks of
processor demand. That means that, as I already tried, you don't even
notice the performance differences of one of those regular applications
when you run one of those on a P3 1GHz or on a new Athlon 64 X2 system.
On the other hand, when those isolated peaks start flowing in, you do notice
a difference. That difference gets bigger and bigger once you start firing
up more and more instances of those processes. So if you run an email
client, open a bunch of different sites and start up a word processor, then
those peaks start dragging the system performance to the ground. If you
happen to run a multi-processor system, that will not affect you as much as
when you run a single-processor system. There are less processes that pile
up to get those precious processor time slices, there are even fewer
processor switches and, even better, your operating system can level down
the processor demands. Once you take in consideration that browser plugins
like java and flash (which powers youtube) are ran as different processes,
then there is no question.
But hey, don't take my word for it. There is plenty of operating system
books out there that can help you understand this issue. One which is a
good read, is very accessible and easily explains quite a few complex
concepts is Andrew Tanenbaum's Modern Operating Systems. Do give it a try.
You won't regret it.
On Wed, 23 Jan 2008 01:38:25 +0000, Rui Maciel
<rui.maciel@gmail.com> wrote:
>> Which means very little because there are always far more
>> processes than cores either way, and when apps are sitting
>> in the background idling they tend to be at lower priority
>> and consuming very little CPU time at all.
>
>That's perfectly irrelevant. The operating system manages the processing
>time allocated to each process, which also includes allocating a certain
>process to a certain processor. If you have a set of processes running,
>including a process that demands a considerable amount of processor time,
>the operating system will allocates CPU time slices so that every process
>runs according to its needs and according to the available resources.
You don't have any operating system processes that take a
considerable amount of processor time. Even a single core
sits at the desktop using a mere 1% or so before you install
the 3rd party applications.
>That
>means that the operating system can dedicate one core to run the heavy
>process and run all the other processes on the remaining processors. In
>certain operating systems like Solaris you can even force the OS to bind a
>process to a certain processor.
Which doesn't necessarily matter unless multiple processes
have realtime processing needs, one or more are linear in
nature, and the processor priority is set wrong.
>
>That is not possible in a single processor system.
It doesn't need to be possible.
>In that case, the only
>thing that the operating system can possibly do is keep on executing all
>the processes in that single CPU, constantly switching between them, one
>after the other and putting the remaining processes on hold for a given
>time until their time is up and they can get CPU time once again. That
>waiting time is drastically reduced in multi-processor systems, specially
>if the operating system levels the work load through all processors.
Processes can be put on hold and are all the time in both
single core and multi-core CPU systems. The distinction
you're trying to make does not exist, except if the linear
bound processes, not the OS processes, are too high in
priority which is a software flaw.
>Switching a process also comes with a performance penalty. As it is easy to
>understand, if your process does not have to wait in line for all the other
>process to run to get it's processor time and yours processors don't have
>to spend as much time switching between processes, that processor will end
>up spending less time doing managerial tasks and ends up doing more work in
>less time. That translates to a nice performance boost.
.... only a trivial amount which is born out by anyone who
benchmarks single threaded apps while their OS (itself with
several threads) is running them, on single core versus dual
core systems. That alone is proof positive that the
difference is negligible.
>But hey, don't take my word for it. There is plenty of operating system
>books out there that can help you understand this issue. One which is a
>good read, is very accessible and easily explains quite a few complex
>concepts is Andrew Tanenbaum's Modern Operating Systems. Do give it a try.
>You won't regret it.
It's a shame you don't stop long enough to think about it
instead of thinking someone else needs to be brought up to
speed. Try proof instead of theory. Do as I mentioned
above, benchmark single threaded apps. Keep in mind that
you need not have only the one app running, that merely
putting an app in the background reduces priority so it can
"wait in line", because that's exactly what we want to have
happen, while the other app is being benchmarked.
I'm not suggesting there aren't particular scenarios where
multiple cores won't help, just not all the ones you're
claiming.
??>> So the CPU in question is most likely a E6540 or E6550, both with 4MB
??>> L@ cache and on the 1333MHz FSB - this chip, with good DDR2-667 or 800
??>> RAM and a modern SATA hard drive should perform over 2-3 times better
??>> than a Pentium 4 3.2GHz, even in single-threaded apps.
G> I have tested a P4 3.2 and a Core 2 e6400 2.13GHz in the same PC.
G> Performance is about the same.
no, core2 is about 1.4 times faster on most workloads.
G> Obviously multi-threaded apps are a different story, but I don't use
G> any. Neither do most users of multiple core machines!
most likely P4 CPU had HT feature, which doesn't give much acceleration, but
increases system responsiveness.
you know, many user actually use heavy computational tasks -- virus scan,
file compression/decompression, backup, video/audio coding/editing, games
etc -- and even if they don't get accelerated by multiple cores (but many
modern ones do!), you'll get more responsiveness -- you won't notice
slowdown. even if you're not engaging into such uses all the time, it's
pretty annoying that your normal activities are disturbed by tasks like
virus scan from time to time.
with HT you get something like half of processor guaranteed. this is enough
for normal tasks, but with dual core CPU you can play games while scanning
for viruses, for example.
with modern freaking and unrealiable software this "heavy computations"
situations happen more frequently than you think -- some application can
hang in infinite loop (i had it pretty freaquently for winword and
iexplore), or some web site using JavaScript/Flash in erroneous maner can
eat all CPU resources without actually hanging. with one-core CPU you'll
instantly notice this: "OMG, my system got so slow, something got hang.."
and will have to take down abusing process, while on dual-core system you
won't even notice this.
and, one more thing, there are real-time applications like watching videos
(large, HD ones, that consume lots of CPU power) and playing games. as i've
said, lots of applications are badly written, and some of 500 threads
running on your system wake from time to time and do some tasks. on
single-core CPU this can degrade quality of your realtime applications --
there will be glitches in video or you'll be killed in the game -- and you
will be annoyed. with multi-core one it should be fun.
of course, multi-core CPUs eliminate only one bottleneck, but HDDs still
have only one head that has to move on the platter to read data, and this
becomes bottleneck when using multiple applications at once. partly you can
fix this with RAID. but soon large SSDs will get pretty cheap, and that can
solve IO bottleneck.
"Rui Maciel" <rui.maciel@gmail.com> wrote in message
news:47969a8f$0$27426$a729d347@news.telepac.pt...
> kony wrote:
>
>>>That isn't important, really.
>>
>> Yes it's very important.
>
> It only appears that it's very important to those who fail to understand
> how
> any competent operating system manages it's processes. For a brief
> explanation please read below.
Please don't lecture us (especially Kony) on architecture and OS operations.
Open any process browser and you will see that the 30 or so system processes
consume, at most, 3-5% CPU time as a peak value and typically less than 1%
average, so dual core will really make no difference at all to the
foreground app - hence usage profiles are important when selecting a CPU. A
dual core CPU for a user of single core applications is a waste of time as
system and other processes do not consume enough CPU time to make dual core
worth while! Besides, you would need around 32 processors to erradicate the
background CPU times that we are talking about as they don't neatly organise
themselves into serial operation on the second core!
"Alex Mizrahi" <udodenko@users.sourceforge.net> wrote in message
news:479704bf$0$90271$14726298@news.sunsite.dk...
>
> ??>> So the CPU in question is most likely a E6540 or E6550, both with 4MB
> ??>> L@ cache and on the 1333MHz FSB - this chip, with good DDR2-667 or
> 800
> ??>> RAM and a modern SATA hard drive should perform over 2-3 times better
> ??>> than a Pentium 4 3.2GHz, even in single-threaded apps.
>
> G> I have tested a P4 3.2 and a Core 2 e6400 2.13GHz in the same PC.
> G> Performance is about the same.
>
> no, core2 is about 1.4 times faster on most workloads.
Not in the real world tests here it wasn't! However it is more overclockable
and has 2 cores for future multi-threaded apps.
G>>> I have tested a P4 3.2 and a Core 2 e6400 2.13GHz in the same PC.
G>>> Performance is about the same.
??>>
??>> no, core2 is about 1.4 times faster on most workloads.
G> Not in the real world tests here it wasn't!
what do you mean by "real world tests"?
is file compression "real world" enough for you?
Core2 E6400 is 27% faster than Pentium D 940 (3.2 GHz) in bzip2 compression.
(i believe you'll get comparable results with any type of compression).
E6400 is faster in all tests. SPEC CPU2006 is probably most official
benchmarking suit, however it's oriented on professional uses, i do not know
if professional uses are "real world" enough for you.
if winword and mspaint is world reality for you then yes, they are pretty
same. 0.0001 and 0.00013 seconds of CPU time used are barely noticeable.
i also have another set of benchmarks here, but they compare E6400 to
Pentium XE 965 (2 core with HT, 3.73 GHz) -- E6400 beats even that fastest
Pentium in various tasks including games, video compression, file
compression, OCR, professional CAD/CAE -- by 15-22%.
On Wed, 23 Jan 2008 12:33:30 +0200, "Alex Mizrahi"
<udodenko@users.sourceforge.net> wrote:
> G>>> I have tested a P4 3.2 and a Core 2 e6400 2.13GHz in the same PC.
> G>>> Performance is about the same.
> ??>>
> ??>> no, core2 is about 1.4 times faster on most workloads.
>
> G> Not in the real world tests here it wasn't!
>
>what do you mean by "real world tests"?
>
>is file compression "real world" enough for you?
>
>Core2 E6400 is 27% faster than Pentium D 940 (3.2 GHz) in bzip2 compression.
>(i believe you'll get comparable results with any type of compression).
>
>video compression? E6400 is 21% faster.
>
>and GNU C Compiler runs 30% faster on E6400.
>
>you can take SPEC CPU2006 results yourself:
>http://www.spec.org/cpu2006/results/...205-00376.html
>http://www.spec.org/cpu2006/results/...205-00388.html
>
>E6400 is faster in all tests. SPEC CPU2006 is probably most official
>benchmarking suit, however it's oriented on professional uses, i do not know
>if professional uses are "real world" enough for you.
The key thing is a benchmark is only as valid as it mirrors
the real use of the system. Even benchmarking using the
same application but a different version of it may not be
valid.
In general, benchmarkers tend to use the newer benchmark
suites which are more likely optimized for SMP, while plenty
of "real world" professional users do not rebuy new software
when they buy their new system, particularly when much of
the software can cost quite a lot more than the CPU does.
> Please don't lecture us (especially Kony) on architecture and OS operations.
> Open any process browser and you will see that the 30 or so system processes
> consume, at most, 3-5% CPU time as a peak value and typically less than 1%
> average, so dual core will really make no difference at all to the
> foreground app - hence usage profiles are important when selecting a CPU. A
> dual core CPU for a user of single core applications is a waste of time as
> system and other processes do not consume enough CPU time to make dual core
> worth while! Besides, you would need around 32 processors to erradicate the
> background CPU times that we are talking about as they don't neatly organise
> themselves into serial operation on the second core!
Without sounding like a lecturer, I see about 50 processes, but 658
threads, in Windows Vista at idle. Now dual-core is not process-
orientated, but thread-orientated, so with 2 cores, each core has only
329 threads to deal with on my computer (irrespective of how many
processes these are attached to), so leaves each core with 50% more
available cycles than before? Also, although many older programs are
not dual-core optemized, they still have several threads, and a good
OS like XP or better, Vista, allocates the threads to different cores,
weather the program likes it or not. I know Vista allocates the
threads, as my dual-core shows average 10-15% on BOTH cores at idle,
the graphs look very similar, meaning that at least half the running
threads are on the second core.
I think what you are saying is a single-threaded app has no benefit
itself from dual-core, but the system as a whole does, as when one
core is maxed out with a single-thread game or heavy duty app, the
other core can take over system tasks? correct?
> Now, if you were to test with SuperPI 1.5 or something similar,
> that would spend a good deal of its time using nothing but
> processor and memory. Comparing the old system to the new system,
> would then give an idea if the CPU and memory was the limitation.
> Just the "feel" of the system, may not be narrowing down where the
> problem is. (On the machine I'm typing on, SuperPI 1.5 will do the
> 1 million digits test in 50 seconds. It became that slow, after
> I added my antivirus software. I think I might have had maybe
> 45 seconds at one time, before the antivirus was added. Just if you
> want to compare. SuperPI is single threaded AFAIK. My system is a
> Northwood P4 3.15GHz/FSB900/DDR450 dual channel. A highly overclocked
> Core2 can do the same benchmark in sub 10 seconds, to give some idea
> of how good it can be. So my system is 5x out of date.)
Really? Sub 10 seconds? Wicked!
My E4500 @ 3.3GHz takes 17 seconds to calculate 1M.
Quad/Dual Core Duo Questions 
Linear Mode
