Is this CPU 4 x 3.60GHz ?

[Desmond]

I have had a amd dual core 4.2G and was told that they add up all the speads which is not exactly the same. This is the CPU

http://www.amazon.co.uk/AMD-Bulldoz...1_1?s=computers&ie=UTF8&qid=1354022685&sr=1-1


I want to buy also
http://www.amazon.co.uk/Asus-M5A78L...1_1?s=computers&ie=UTF8&qid=1354022295&sr=1-1


AND

http://www.amazon.co.uk/Corsair-CMZ...1_1?s=computers&ie=UTF8&qid=1354022847&sr=1-1

Will this be OK please.

Desmond.

 

[SC Tom]

I have had a amd dual core 4.2G and was told that they add up all the
speads which is not exactly the same. This is the CPU

http://www.amazon.co.uk/AMD-Bulldoz...1_1?s=computers&ie=UTF8&qid=1354022685&sr=1-1


I want to buy also
http://www.amazon.co.uk/Asus-M5A78L...1_1?s=computers&ie=UTF8&qid=1354022295&sr=1-1


AND

http://www.amazon.co.uk/Corsair-CMZ...1_1?s=computers&ie=UTF8&qid=1354022847&sr=1-1

Will this be OK please.

Desmond.

If you're asking if they'll all work together, yes they will. The 4100 is a
quad-core CPU running at 3.6GHz, which would be faster in some applications
than your dual-core 4.2GHz (this is assuming that the app can address
multiple cores). Sounds like a pretty decent setup
The 4100 is on the "supported CPU" list
<http://www.asus.com/Motherboards/AMD_AM3Plus/M5A78LMUSB3/#CPUS>

 

[Paul]

If you're asking if they'll all work together, yes they will. The 4100
is a quad-core CPU running at 3.6GHz, which would be faster in some
applications than your dual-core 4.2GHz (this is assuming that the app
can address multiple cores). Sounds like a pretty decent setup
The 4100 is on the "supported CPU" list
<http://www.asus.com/Motherboards/AMD_AM3Plus/M5A78LMUSB3/#CPUS>

It's pretty hard to find good benchmarks, and all in one place.
But this site is a start.

http://www.cpubenchmark.net/high_end_cpus.html

AMD FX-4100 Quad-Core 4,020

Since we don't know what your original dual core processor is,
we can just try Athlon II X2 260 Regor at 3.2GHz, then use
linear scaling for estimation purposes to 4.2GHz.

http://www.cpubenchmark.net/cpu_list.php

AMD Athlon II X2 260 1810 x 4.2/3.2 = 2,376

There is a difference between single-threaded benchmarks (SuperPI)
and multi-threaded benchmarks (Passmark). The single-threaded ones
are basically testing how well a super-high clock rate on one
core helps. Some applications in practice are single-threaded
as well, and so a single-threaded benchmark has some predictive
value. But it's harder to dig up results like that. While
hwbot.org has such results, actually extracting them from
the website is a real chore (not much fun). HWBot concentrates
on overclocking contests, and it's pretty hard to get "stock"
speed numbers. You'd think the membership there, would at
least be interested in the ratio of their overclocked system,
to a "stock" result, for bragging rights. (The site owners
seem to spend their days re-coding the HTML, and taking pride
on making "pretty" web pages. I'd rather just have the data
thanks )

http://www.hwbot.org/hardware/processors

Paul

 

[RayLopez99]

It's pretty hard to find good benchmarks, and all in one place.

But this site is a start.



http://www.cpubenchmark.net/high_end_cpus.html

Thanks Paul. I've added this site to my benchmark sites. I do notice whentesting my chips the benchmarks relative to one another (faster score chip/ slower score chip) are a bit different, since it seems cpubenchmark is testing raw performance rather than real world performance. I also like this site:

http://www.anandtech.com/bench/CPU/2

RL

 

[Paul]

Thanks Paul. I've added this site to my benchmark sites. I do notice when testing my chips the benchmarks relative to one another (faster score chip / slower score chip) are a bit different, since it seems cpubenchmark is testing raw performance rather than real world performance. I also like this site:

http://www.anandtech.com/bench/CPU/2

RL

I'm still looking for a good source of SuperPI though.
That's the missing ingredient. It's amazing they can
include so many other test, and not that one. And
they do know about SuperPI. It has a long history.

Paul

 

[Desmond]

RayLopez99 wrote: > On Wednesday, November 28, 2012 10:27:36 PM UTC+2, Paul wrote: >> It's pretty hard to find good benchmarks, and all in one place.. >> >> But this site is a start. >> >> >> >> http://www.cpubenchmark.net/high_end_cpus.html >> > > Thanks Paul. I've added this site to my benchmark sites. I do notice when testing my chips the benchmarks relative to one another (faster score chip / slower score chip) are a bit different, since it seems cpubenchmark is testing raw performance rather than real world performance. I also like this site: > > http://www.anandtech.com/bench/CPU/2 > > RL I'm still looking for a good source of SuperPI though. That's the missing ingredient. It's amazing they can include so many other test, and not that one. And they do know about SuperPI. It has a long history. Paul

Hi Thanks for this. I upgraded my PC in 2006
asus m2v sx motherboard
The AMD Athlon II X2 4.2G. I was told that it was a bit of a con as I only had 2 2.1G cores and they add them all up to make it look like you are getting a faster processor than you are. If I bought 10 average family cars andbolted them all together have I got a ferrari?

This means that
AMD Bulldozer FX-4100 Socket AM3+ 4 Core Processor - 3.60GHz, 3.80GHz Turbo
Is 4 x 0.9 G cores. I thought the cores were to get over the limitations onhow fast a core could be under the current technology. If I am using a very basic app that only has one thread it would run toooo slow.

 

[Desmond]

I have had a amd dual core 4.2G and was told that they add up all the speads which is not exactly the same. This is the CPU http://www.amazon.co.uk/AMD-Bulldoz...1_1?s=computers&ie=UTF8&qid=1354022685&sr=1-1 I want to buy also http://www.amazon.co.uk/Asus-M5A78L...1_1?s=computers&ie=UTF8&qid=1354022295&sr=1-1 AND http://www.amazon.co.uk/Corsair-CMZ...1_1?s=computers&ie=UTF8&qid=1354022847&sr=1-1 Will this be OK please. Desmond.

Sorry to be a pain. If all these work together that is fine. It was the 4 *0.9G that concerns me. In addition I came across this by accident.
AM3+ 700
AM3+ 900

The 900 seemes to be a difrent proccessor Bulldozer. Will some AM3+ mobos work with one and not the other? Just a simple clarification. I am checking everywhere to resurch this.

TIA

Desmond.

 

[SC Tom]

Hi Thanks for this. I upgraded my PC in 2006
asus m2v sx motherboard
The AMD Athlon II X2 4.2G. I was told that it was a bit of a con as I only
had 2 2.1G cores and they add them all up to make it look like you are
getting a faster processor than you are. If I bought 10 average family
cars and bolted them all together have I got a ferrari?

This means that
AMD Bulldozer FX-4100 Socket AM3+ 4 Core Processor - 3.60GHz, 3.80GHz
Turbo
Is 4 x 0.9 G cores. I thought the cores were to get over the limitations
on how fast a core could be under the current technology. If I am using a
very basic app that only has one thread it would run toooo slow.

No, you have 4 cores running at 3.6GHz each. If you use a program like SIW
<http://www.gtopala.com/> you can see the frequency of each core. For
example, I have an AMD Phenom II X4 965 Black Edition Processor 3.4GHz, and
each core shows that speed at 100%. Usually, when I'm not doing anything
that requires heavy CPU usage or heavy graphics, it idles at 1.8 to 2.2GHz.

FWIW, you can get the 965BE for ~5 Pounds cheaper, and it has better
benchmarks than the FX-4100 on
<http://www.cpubenchmark.net/high_end_cpus.html>. The FX-4100 isn't even
listed on the AnandTech site. It's also on the list of supported CPUs on the
Asus link I posted earlier.

 

[Geoff Pearson]

No, you have 4 cores running at 3.6GHz each. If you use a program like SIW
<http://www.gtopala.com/> you can see the frequency of each core. For
example, I have an AMD Phenom II X4 965 Black Edition Processor 3.4GHz,
and each core shows that speed at 100%. Usually, when I'm not doing
anything that requires heavy CPU usage or heavy graphics, it idles at 1.8
to 2.2GHz.

FWIW, you can get the 965BE for ~5 Pounds cheaper, and it has better
benchmarks than the FX-4100 on
<http://www.cpubenchmark.net/high_end_cpus.html>. The FX-4100 isn't even
listed on the AnandTech site. It's also on the list of supported CPUs on
the Asus link I posted earlier.

But don't download from the gtopala site without watching for Babylon -
which will take over your machine. It is a menace.

 

[SC Tom]

But don't download from the gtopala site without watching for Babylon -
which will take over your machine. It is a menace.

No trouble here with that, or anything else unwanted. There was a prompt for
some other "free" software, but all I had to do was uncheck the Accept box.

 

[Yousuf Khan]

Sorry to be a pain. If all these work together that is fine. It was
the 4 * 0.9G that concerns me. In addition I came across this by
accident. AM3+ 700 AM3+ 900

No, it's not 4 * 0.9GHz, it is 4 * 3.6GHz, each core does run at 3.6GHz,
and in fact that processor has an automatic overclocking feature which
lets some of its cores run as high as 3.8GHz.

The 900 seemes to be a difrent proccessor Bulldozer. Will some AM3+
mobos work with one and not the other? Just a simple clarification. I
am checking everywhere to resurch this.

The FX-4100 is an AM3+ processor, meaning it fits into specific socket
which is the same for all processors that are AM3+. The AM3+ socket is
also downwardly compatible with processors that were AM3 compatible (the
earlier generation socket). So an older processor that is AM3 will work
on an AM3+ motherboard.

Yousuf Khan

 

[Paul]

No, you have 4 cores running at 3.6GHz each. If you use a program like
SIW <http://www.gtopala.com/> you can see the frequency of each core.
For example, I have an AMD Phenom II X4 965 Black Edition Processor
3.4GHz, and each core shows that speed at 100%. Usually, when I'm not
doing anything that requires heavy CPU usage or heavy graphics, it idles
at 1.8 to 2.2GHz.

FWIW, you can get the 965BE for ~5 Pounds cheaper, and it has better
benchmarks than the FX-4100 on
<http://www.cpubenchmark.net/high_end_cpus.html>. The FX-4100 isn't even
listed on the AnandTech site. It's also on the list of supported CPUs on
the Asus link I posted earlier.

http://support.asus.com/Cpusupport/List.aspx?SLanguage=en&m=M5A78L-M/USB3&p=1&s=24

The FX-8350 is supported as well as the FX-4100.

Looking at the chart, in the price bracket, I might want
a FX-4170 instead of the FX-4100. Higher clock for
single threaded stuff. FX-4170 is $120, FX-4100 is $110.

AMD FX-4170 Zambezi 4.2GHz (4.3GHz Turbo) Socket AM3+ 125W Quad-Core $120
AMD FX-4100 Zambezi 3.6GHz (3.8GHz Turbo) Socket AM3+ 95W Quad-Core $110

The FX-8350 would look similar to this, while an
FX-4170 would use half of this. Each quadrant supports
two threads of execution. They count those threads
of execution as "cores", but the definition is a bit mushy.

http://images.anandtech.com/doci/4955/FXCPU_Die_575px.jpg

A quadrant looks like this. Shared fetch and decode, followed
by two execution sections. A little different than how Intel
does it. The 965 doesn't work like this at all, and has
a more conventional ratio of fetch:decode:execution.

http://images.anandtech.com/galleries/754/BulldozerHotChips_August24_8pmET_NDA-6_575px.jpg

The 965 looks like this. It has four cores, compared to the
two quadrants the FX-4100 type chip has.

http://images.anandtech.com/reviews/cpu/amd/phenom2/810/900die.jpg

I might be tempted to classify the 965 as "4C 4T" and the
FX-4100 as "2C 4T". Implying the two threads of execution
in the FX-4100 are similar to a heavy-weight Hyperthreading
implementation. But what do I know. Marketing wants to sell
it as "4C 4T" for the FX-4100, and who are we to argue.

We can always let the benchmarks decide. That's the best way
to guide people wanting to buy one.

*******

Programs that are multi-threaded (like Photoshop), can
use more than one core at a time. That's where a
multi-core processor provides its advantage. If Photoshop
chops an image into eight pieces, the FX-8350 uses
one thread of execution of the eight its got, to process
each piece. This reduces the time to finish the job. With
the FX-4170, Photoshop chops the image into four
pieces, with one thread running on each of the
four threads of execution. The job takes twice as long
as it would on the FX-8350.

Not all Photoshop filters are multi-threaded. About
half of the filters are single threaded. The single threaded
Photoshop filters run as fast on the FX-4170 as they do on
the FX-8350. Because only one execution unit in a single
quadrant gets used for the single thread of execution.

And other application mixes will show the same thing.
Applications that use multiple threads of execution,
will get a boost from FX-8350 over FX-4170. But
things that are written with a single thread of
execution (like SuperPI), then only one core is
running that in either the FX-8350 or FX-4170 case.

So only some applications get a big boost from the
more expensive processor. If you edit movies, play
certain games, the fat processor wins. For a lot
of older programs and utilities, there isn't that
much of a difference. And this is why, sometimes
users are disappointed in their upgrade. They
don't understand how the applications are split.
If you do nothing but movie editing, then the
more cores, the better. (Buy the fat processor.)
On Photoshop, it depends on what filters you use,
because they're mixed. Photoshop uses single-threaded
filters, for things requiring the most mathematically
accurate methods. In Photoshop, if you rotate an image
by 5 degrees, 72 times, the developers at Adobe want
the image quality to be preserved and look like the
original unrotated picture. And sometimes, that means
dispensing with multi-threaded methods. Adobe prefers
accuracy, over speed. At least that's the claim I've read.

Paul

 

[Flasherly]

The AMD Athlon II X2 4.2G. I was told that it was a bit of a con as I only had 2 2.1G cores and they add them all up to make it look like you are getting a faster processor than you are. If I bought 10 average family cars and bolted them all together have I got a ferrari?

More along a "Power Wall" design implementation. A justification of
increasing power consumption to diminishing overall return factors,
shrinking or 'bolting them together' in some better engineering sense
mitigates. In Intel's implementation, perhaps for one way I'm reading
it, is their dual-core of a quad permits, as an entity, singular
instances of one dual core working, (presuming the other dual core is
not in a sense needlessly consuming processing clock cycles), whereas
within a quad framework of un-unified, disparate processors,
logically, requiring all four to be running in a constant order of
qualifying the quad for always a quad.