Performance question re new build

[Hackworth]

My wife needed a new computer, so I decided to give her my system and use
that as an excuse to build myself a new one.

Her system is now my old Pentium 4 (Northwood) 2.6 GHz running on an Asus
P4B266 (400-MHz FSB) w/ 1GB Mushkin Black memory and an AGP Radeon 9600, all
running on Win XP Pro SP2.

My new system is an Athlon64 3200+ running on an Asus A8V Deluxe (Rev. 2) w/
latest BIOS. It has 2 GB dual-channel PC3200 memory. It too has an AGP
Radeon 9600. The new system also runs Win XP Pro SP2.

According to my benchmarks (Passmark Performance Test 5.0 and PCMark05), my
new system is only about 10 percent faster than my old one.

I'm running the latest A8V Deluxe BIOS (v. 1013). I've checked and rechecked
the BIOS settings to make sure that I didn't accidentally screw anything up.
(I've put together scores of machines over the past decade and have
repaired/upgraded countless others, so I know enough to check the obvious
stuff first.) I have Asus' Cool-n-Quiet feature disabled because my power
supply (Antec NeoPower 480) and CPU fan all adjust fan speeds independently.
Also, I'm not overclocking yet because I want to make sure that everything
runs smoothly at stock speeds and voltages.

Any ideas as to what else I may need to examine? Or is this pretty much the
performance increase that I should be expecting?

 

[Fred]

My wife needed a new computer, so I decided to give her my system and
use that as an excuse to build myself a new one.

Ah, your wife. You mean the town bike.

[IDIOTICON] Her system is now my old Pentium 4 (Northwood) 2,6 GHz
running on an Asus P4B266 (400-MHz FSB) w/ 1GB Mushkin Black memory and an
AGP Radeon 9600, all running on Win XP Pro SP2.

How old?

My new system is an Athlon64 3200+ running on an Asus A8V Deluxe (Rev.

Gibberish and gobbledygook. All false.

I worry, worry, worry and...

Crikey. I'm not buying into that one.

2) w/ latest BIOS.

Classic inappropriate analogy.

I just called my partner the wrong name in bed. Anyone got any
bandages?

It could be worse. You might have had to go to the doctor and get a huge,
stinking, rotting lump of flesh removed from the neck up.

It has 2 GB dual-channel PC3200 memory.

Denial of the antecedent. "A implies B, A is false, therefore B is false."

I'm falling apart. These problems are driving me mad. First my
Korsakov's syndrome returned, and then, would you believe, my doctor
thought it was herpes! I'm barely able to walk and I don't know if I can
keep drinking the way I have been. I just pray that next week I prosper on
the new diet.

If you insist, Hackworth.

It too has an AGP Radeon 9600.

Classic disregard for the truth, as evidenced by your remark.

hi all, this may sound weird but is it possible for women to have wet
dreams or am i wrong? i went to bed last night and awakened this
morning and not to be rude but my underwear was soaking wet, with what
i have no idea, and don’t really want to know, and i had not peed
myself!

Don't start up any new relationships, Hackworth.

The new system also runs Win XP Pro SP2.

**** it off. Install linux. It's worse. You'll love XP after that.

According to my benchmarks (Passmark Performance Test 5.

Classic hypocrisy, as evidenced by the posting of the same tripe over and
over.

Im 12 and ive known this guy for all my life (well since i was 2) and i
really fancy him. But i dont no if he fancys me cuz were good m8s and
we do stuff 2gether. He always shows off in front of me and stuff, but
how do i no for sure that he fancies me.

You are at a friend's pool party and see their incredibly gorgeous 16-year
old daughter sunbathing. Do you fantasise about her?

0 and PCMark05), my new system is only about 10 percent faster than my
old one.

Do you worry about getting old?

I'm running the latest A8V Deluxe BIOS (v.

Why not try a Tu quoque lame, Hackworth?

i know you might have the excitement of a secret relationship but i
hardly even get to see her and when i do her friends are always there
so i cant d anything with her she wont tell anyone so is she ashamed to
be with me what do i do?

Have you ever hung up the phone because the person taking your order did not
speak English?

I've checked and rechecked the BIOS settings to make sure that I didn't
accidentally screw anything up.

You didn't accidentally screw anything up? Bullshit. I don't believe you.
Liar.

(I've put together scores of machines over the past decade and have
repaired/upgraded countless others, so I know enough to check the obvious
stuff first)

Does anyone else know enough to check the obvious stuff first?

I have Asus' Cool-n-Quiet feature disabled because my power supply
(Antec NeoPower 480) and CPU fan all adjust fan speeds independently.

Really? You have asus' cool-n-quiet feature disabled because your power
supply?

Also, I'm not overclocking yet because I want to make sure that
everything runs smoothly at stock speeds and voltages.

Is that your main goal in life, to make sure that everything runs smoothly
at stock speeds?

Any ideas as to what else I may need to examine?

Why bother?

Or is this pretty much the performance increase that I should be
expecting?

How would I know if that is pretty much the performance increase that you
should be expecting?

 

[Ruel Smith]

My wife needed a new computer, so I decided to give her my system and use
that as an excuse to build myself a new one.

Her system is now my old Pentium 4 (Northwood) 2.6 GHz running on an Asus
P4B266 (400-MHz FSB) w/ 1GB Mushkin Black memory and an AGP Radeon 9600,
all running on Win XP Pro SP2.

My new system is an Athlon64 3200+ running on an Asus A8V Deluxe (Rev. 2)
w/ latest BIOS. It has 2 GB dual-channel PC3200 memory. It too has an AGP
Radeon 9600. The new system also runs Win XP Pro SP2.

According to my benchmarks (Passmark Performance Test 5.0 and PCMark05),
my new system is only about 10 percent faster than my old one.

I'm running the latest A8V Deluxe BIOS (v. 1013). I've checked and
rechecked the BIOS settings to make sure that I didn't accidentally screw
anything up. (I've put together scores of machines over the past decade
and have repaired/upgraded countless others, so I know enough to check the
obvious stuff first.) I have Asus' Cool-n-Quiet feature disabled because
my power supply (Antec NeoPower 480) and CPU fan all adjust fan speeds
independently. Also, I'm not overclocking yet because I want to make sure
that everything runs smoothly at stock speeds and voltages.

Any ideas as to what else I may need to examine? Or is this pretty much
the performance increase that I should be expecting?

Yeah, put them both in Safe Mode and rerun the benchmark. Sometimes
background processes can seriously skew benchmarks. The different
applications installed on each system make a difference. The only way to get
a true benchmark from both is if they both had a fresh install of Windows XP
and drivers only. You might try some other benchmarks too, like Sandra.

Also, I really don't expect a really large increase in performance. A 3200+
theoretically should be about 20+% faster than a 2.6 GHz, but we're
comparing apples and oranges. The 3200 actually runs at 2 GHz, uses slightly
different instruction sets, and is a different beast altogether, besting
Intel processors in some benchmarks while getting waxed in others. Also,
even if you had a 3.2 GHz Intel vs. a 2.6 GHz Intel, we're now at the stages
of diminishing returns on clock speed increases, so a 23% faster clock
rating wouldn't yield a 23% faster machine. However, I, too, would expect
better than 10%. I'm thinking around 18 - 20% or so.

These diminishing returns is what's stopping me from building a new system.
I also have a 2.6 GHz P4 system based on a Gigabyte GA-8IRXP motherboard and
I haven't seen enough improvement in processors I can actually afford to
warrant a new build. Maybe the new Athlon 64 X2 will give me a reason when
the prices settle down.

 

[Hackworth]

Yeah, put them both in Safe Mode and rerun the benchmark. Sometimes
background processes can seriously skew benchmarks. The different
applications installed on each system make a difference. The only way to
get a true benchmark from both is if they both had a fresh install of
Windows XP and drivers only. You might try some other benchmarks too, like
Sandra.

Also, I really don't expect a really large increase in performance. A
3200+ theoretically should be about 20+% faster than a 2.6 GHz, but we're
comparing apples and oranges. The 3200 actually runs at 2 GHz, uses
slightly different instruction sets, and is a different beast altogether,
besting Intel processors in some benchmarks while getting waxed in others.
Also, even if you had a 3.2 GHz Intel vs. a 2.6 GHz Intel, we're now at
the stages of diminishing returns on clock speed increases, so a 23%
faster clock rating wouldn't yield a 23% faster machine. However, I, too,
would expect better than 10%. I'm thinking around 18 - 20% or so.

These diminishing returns is what's stopping me from building a new
system. I also have a 2.6 GHz P4 system based on a Gigabyte GA-8IRXP
motherboard and I haven't seen enough improvement in processors I can
actually afford to warrant a new build. Maybe the new Athlon 64 X2 will
give me a reason when the prices settle down.

My version of the A64 3200+ has the Venice core (with only 512MB cache), so
that's probably at least one reason for the new system's lackluster
performance.

 

[Ruel Smith]

My version of the A64 3200+ has the Venice core (with only 512MB cache),

so that's probably at least one reason for the new system's lackluster
performance.

I haven't seen any Venice vs. Winchester performance comparisons, but they
seem to be almost the same chip, only with the Venice having the SSE3
instruction set. Both run at 2.0 GHz, and both have the same amount of
cache. The Winchester benchmarks 20 - 30% better than does the 2.6 GHz P4,
depending on the benchmark. However, like I said, background processes can
affect the benchmarks, so you'r mileage can vary.

http://www.tomshardware.com/cpu/20041221/index.html