XP3000+ compared?

[Jess Fertudei]

Hello,

As I noted in another post, I am going to bump up the processor on my ASRock
K7S8XE+ that I use for rendering plug-ins. I keep the machine very lean as
to background processes and have only recently found a few plug-ins that
slow down it's Duron 1.3G. But it is time to upgrade it a little.

How much real-life drop off is there from an XP3000+ @ 400 to one @ 333FSB?

Then again, how much drop off in everyday performance is there from an
XP3000+ @333 to a Sempron3000+ @333?


The Sempron3000+ 333 sure seems attractive from the price point, but best
bang for buck is only applicable if the drop off is only marginal.




Thanks as usual.

 

[jaster]

Hello,

As I noted in another post, I am going to bump up the processor on my
ASRock K7S8XE+ that I use for rendering plug-ins. I keep the machine very
lean as to background processes and have only recently found a few
plug-ins that slow down it's Duron 1.3G. But it is time to upgrade it a
little.

How much real-life drop off is there from an XP3000+ @ 400 to one @
333FSB?

Then again, how much drop off in everyday performance is there from an
XP3000+ @333 to a Sempron3000+ @333?


The Sempron3000+ 333 sure seems attractive from the price point, but best
bang for buck is only applicable if the drop off is only marginal.

Tomshardware wrote an article about the benefits of the Sempron 3100 oc'd
which might help ewith your dilema.

The Sempron 3000 is the XP 3000. AMD rolled the XP line into the Sempron
line.

 

[David Maynard]

Hello,

As I noted in another post, I am going to bump up the processor on my ASRock
K7S8XE+ that I use for rendering plug-ins. I keep the machine very lean as
to background processes and have only recently found a few plug-ins that
slow down it's Duron 1.3G. But it is time to upgrade it a little.

How much real-life drop off is there from an XP3000+ @ 400 to one @ 333FSB?

It depends on what one is doing. In some things the 333 FSB version is
actually faster by a percent, or so, because it increases clock speed to
compensate for the lower bus speed and if the app isn't memory intensive
the extra speed can over compensate. One the other hand, with memory
intensive applications, like video editing, it comes out 3 to 4% slower
because of the slower FSB.

It's supposed to be roughly a wash, on average, which is why they have the
same rating but, in practice, I'd say the 400 FSB is a percent or two faster.

Then again, how much drop off in everyday performance is there from an
XP3000+ @333 to a Sempron3000+ @333?

The Sempron 3000+ is a Barton core at 2 GHz whereas the XP 3000+ is a
Barton core at 2.167 GHz. So, scaling it to the XP, a Sempron 3000+ comes
out about a 2700+. Since you're looking at the same FSB for both processors
the XP should be roughly 11% faster.

 

[David Maynard]

Tomshardware wrote an article about the benefits of the Sempron 3100 oc'd
which might help ewith your dilema.

The Sempron 3000 is the XP 3000. AMD rolled the XP line into the Sempron
line.

It's true that Semprons are XPs by another name but the 'performance
rating' numbers are not the same as they changed what they're 'comparing
to'. XPs were to compete against the P4. Semprons compete with the Celeron.

XPs of the same number are faster.

 

[jaster]

It's true that Semprons are XPs by another name but the 'performance
rating' numbers are not the same as they changed what they're 'comparing
to'. XPs were to compete against the P4. Semprons compete with the
Celeron.

XPs of the same number are faster.

Yes slightly faster by 2-5% right?

I thought AMD decided hey we have this product line that's still pretty
good so why not just change it's name and competitor? Go down a weight
class.

 

[David Maynard]

Yes slightly faster by 2-5% right?

I think the number is closer to 11% (for the Sempron 3000+ vs the XP 3000+)
but I haven't done a comparison across the entire line.

I thought AMD decided hey we have this product line that's still pretty
good so why not just change it's name and competitor? Go down a weight
class.

Basically, yes. Of course, they had to change the 'name' in order to change
the speed rating because it would be bizarre, indeed, to have two different
speeds with exactly the same name/rating designation.

Anyone remember that the reason they gave for the speed 'rating' was to
make it *easier* to compare processors? But now we're right back to "3000"
not being "3000," depending on what processor you're looking at, in just
their own lineup. Even ignoring minor issues like a 333 FSB vs 400 FSB
version of the 'same' rating, an XP 3000+ does not perform like a socket A
Sempron 3000+, which does not perform like a socket 754 Sempron 3000+,
which does not perform like an Athlon 64 3000+.

And then Intel went and fouled the whole thing up by dropping 'MHz' from
the name entirely.

 

[jaster]

I think the number is closer to 11% (for the Sempron 3000+ vs the XP
3000+) but I haven't done a comparison across the entire line.

Tomshardware.com has a pretty complete chart comparing processor
performance from 64 3400 and 3.4E to Durons and Cels in their processor
section.

Basically, yes. Of course, they had to change the 'name' in order to
change the speed rating because it would be bizarre, indeed, to have two
different speeds with exactly the same name/rating designation.

Anyone remember that the reason they gave for the speed 'rating' was to
make it *easier* to compare processors? But now we're right back to
"3000" not being "3000," depending on what processor you're looking at,
in just their own lineup. Even ignoring minor issues like a 333 FSB vs
400 FSB version of the 'same' rating, an XP 3000+ does not perform like
a socket A Sempron 3000+, which does not perform like a socket 754
Sempron 3000+, which does not perform like an Athlon 64 3000+.

On the 939 socket.

And then Intel went and fouled the whole thing up by dropping 'MHz' from
the name entirely.

Yeah, I totally give up on the Intel naming. Unless you have benchmarks
there's no way to determine processor speed or features. I think its
silly at this point to even market new cpu unless there are significant
changes in the architecture. Like AMD I think the next generation cpus
should have all current features and speeds but quieter, cooler, less
voltage and case fans can provide enough cooling to the cpu.

 

[David Maynard]

Tomshardware.com has a pretty complete chart comparing processor
performance from 64 3400 and 3.4E to Durons and Cels in their processor
section.

Yep. I've seen it, but I didn't do a ratio for every one of them

 

[Conor]

The Sempron 3000 is the XP 3000.

No it isn't. The Sempron only has 256k L2 cache.

 

[David Maynard]

No it isn't.

You're right there.

The Sempron only has 256k L2 cache.

And you're right there too for all the Socket A Semprons, except for the
socket A Sempron 3000+ that has 512K L2 cache.

 

[Conor]

You're right there.


And you're right there too for all the Socket A Semprons, except for the
socket A Sempron 3000+ that has 512K L2 cache.

I thought only the Socket 754 versions had 512k?

 

[David Maynard]

I thought only the Socket 754 versions had 512k?

Understandable since the Sempron 3000+ is the only socket A version that
has 512K L2 (at least as of the last datasheet I downloaded).

It's the same clock speed as the 256K cache Sempron 2800+, 2 GHz, and gets
the extra '200' from the 512K cache.

A 7% bump sounds a tad low to me but that's what they came up with.

 

[Jess Fertudei]

Thanks to all involved for the discussion.

Understandable since the Sempron 3000+ is the only socket A version that
has 512K L2 (at least as of the last datasheet I downloaded).

It's the same clock speed as the 256K cache Sempron 2800+, 2 GHz, and gets
the extra '200' from the 512K cache.

A 7% bump sounds a tad low to me but that's what they came up with.

 

[jaster]

You're right there.


And you're right there too for all the Socket A Semprons, except for the
socket A Sempron 3000+ that has 512K L2 cache.

Question:
Which is the better Fry's combo deal then?

Sempron 3100 for $159

or

XP 64 3000 for $199

motherboard unknown probably the same ECS with the same features boxed
cpus.

 

[David Maynard]

Question:
Which is the better Fry's combo deal then?

Sempron 3100 for $159

or

XP 64 3000 for $199

motherboard unknown probably the same ECS with the same features boxed
cpus.

Well, that depends entirely on how you define "better deal."

I'd go for the XP 64 because it's 64 bit whereas the Sempron isn't and 40
bucks isn't all that much compared to the total cost of a system. And the
Athlon 64 3000+ is faster than the Sempron 3100+ by quite a bit, depending
on what apps you're running.

 

[ian lincoln]

i bought an amd barton core 2800 its supposed to be faster than the
semprons. It actually clocks at 2.07ghz.