Seeking upgrade advice

[Albuquerque]

well no.....
if a e6600 used the same rating as the 3500 what would it be ?


EDIT :
found some benchmarks for the fx62 and it appears to be very slightly better than a 5000+ or about 30% faster than a pentium 4 d840 3.2 ghz

also found some some benchmarks for the fx62 vs e6700 and they seem roughly similar

Really? Id like a link to that review, because essentially every review I've read (and I've read a LOT) say that the E6600 beats the FX62 in almost every benchmark, usually with a pretty decent margin, with exception to one or two in certain cases. And try as you may, double core is always going to be used unless you're on an old Win9x (or older) operating system. Graphics drivers are multithreaded, sound drivers are multithreaded, and the OS itself is multithreaded. You cant even build a contrived case on a modern OS where a second core goes wholly unutilized.

And, I clearly stated "it depends on what you're doing", but there's no question that an E6600 would be considerably faster than an A64/3500. If the E6600 is besting an X2/5000 by 10+% in gaming cases, then how much would it be utterly spanking an aging 3500? It's a rhetorical question, because it doesn't even need an answer.

I'm really not sure what you'd be trying to prove to suggest that an E6600 is somehow only ~25% faster than a 3500, other than you're just not thinking straight.



EDIT
I felt the need to post some benchmark links, even though I'm on an utterly crappy ~120kbps DSL link and it's taking me forever to surf through all this.

First, Tom's Hardware. WHy Tom's? Because this is the only place I know where you can compare basically every processor made in the last three years...

Quake 4: 1024x768. E6600 ~90% faster than 3500, 5% faster than FX-62
FEAR: 1280x960. E6600 is ~20% faster than 3500, 3% faster than FX-62
CoD2: 1024x768. E6600 is 80% faster than 3500, 15% faster than FX-62

I can find more when I get back on Sunday morning, but that should get you started.

 

[Davros]

post messed up
see reply below

 

[Davros]

I'm really not sure what you'd be trying to prove to suggest that an E6600 is somehow only ~25% faster than a 3500, other than you're just not thinking straight."

here's where i got my fx62 is roughly as fast as a e6700 ( notice i said roughly )

http://www.tomshardware.co.uk/2006/06/05/first_benchmarks_conroe_vs_fx-62_uk/page5.html
http://www.tomshardware.co.uk/2006/06/05/first_benchmarks_conroe_vs_fx-62_uk/page6.html
http://www.tomshardware.co.uk/2006/06/05/first_benchmarks_conroe_vs_fx-62_uk/page7.html

i also found a benchmark where the fx62 was about as fast as a 5000+ and 5000+ is about 35% faster than 3500+

so 5000+ = a fx62 or 35% faster than a 3500+
also fx62 = e6700 (roughly)
so e6700 must be 35% faster than a 3500+
a e6600 im guessing is about 10% slower than a e6700

so in wow (which only uses a single core) an e6600 would be 25% faster than a 3500+

and thats how i worked it out - as you can see I am surely a genius

 

[Albuquerque]

So, you figure an E6700 is fairly equal to an FX62, but you're using numbers that are video-bottlenecked?? Or did you happen to notice that every case where graphics are not the bottleneck the E6700 is definitely faster than the FX62?

Your math is laughable, to be sure. But next time, try being more serious...

 

[Davros]

im not using numbers that are video bottlenecked - you notice that when they overclock the cpu by about 5% the scores go up by a similar amount

 

[Albuquerque]

im not using numbers that are video bottlenecked - you notice that when they overclock the cpu by about 5% the scores go up by a similar amount

What benchmark are you reading? Because it's not the one you linked -- the one that you specifically linked above shows two things:

Number 1: An fx62 running SLI video cards that is being overclocked and gaining some performance.

Number 2: No overclocking of the E6700

So, howabout again you point me to the benchmark that you are referring to? Because now that I'm on a much faster internet connection, I can throw you half a dozen different benchmarks from a large variety of sites showing the E6600 surpassing the FX62 in essentially every popular game out there.

PCStats: Quake4 and Doom3 benchies E6600 is ~5 to 15% faster than FX62
PCStats: FarCry and Fear benchies E6600 is ~5 to 25% faster than FX62
Hexus: Q4, FarCry, SplinterCell, 3DM05 E6600 is ~1 to 15% faster than FX62
AnandTech: Q4, HL2 ep1, BF2142 E6600 is ~5 to 20% faster than FX62
AnandTech: FEAR and Rise of Legends E6600 is ~5 to 30% faster than FX62
Anandtech: Oblivion in several locations E6600 is ~10% faster than FX62
Xbit Labs: Farcry, HL2, FEAR, Q4 E6600 is ~2 to 20% faster than FX62

 

[Davros]

here http://www.tomshardware.co.uk/2006/06/05/first_benchmarks_conroe_vs_fx-62_uk/page5.html

notice the diference between the fx @2.8 + the fx @3.0

the score goes up by a percentage similar to the cpu overclock which wouldnt happen if there was a gfx bottleneck

 

[Albuquerque]

here http://www.tomshardware.co.uk/2006/06/05/first_benchmarks_conroe_vs_fx-62_uk/page5.html

notice the diference between the fx @2.8 + the fx @3.0

the score goes up by a percentage similar to the cpu overclock which wouldnt happen if there was a gfx bottleneck

Ah, I thought you were saying the E6700 was being overclocked. No matter, even in that benchmark it's still quite possible that the extra FSB and memory speed was responsible for the very slight increase in benchmark performance on the overclocked system.

Further, since the E6700 trounced the FX62 even in your own benchmarks in every stock-vs-stock test except one, what does this tell you? This was on an Intel-provded closed rig at the time of benchmarking too. And since I was able to dredge up such a huge handful of post-release links showing you that the E6600 was able to surpass the FX62 equally well, what else does this tell you?

Let me nudge you in the right direction: As I stated quite obviously before, E6600 > FX62 in essentially every benchmark. "Essentially" doesn't mean every single one, but awfully close -- and that's exactly where we are.

 

[Davros]

well i had a look at the game tests and the e6700 hardly "trounces" the fx62 its about 15% ahead which is bugger all

so my arguement that e6600 = fx62 = just over a amd 5000+ = 35% faster than a 3500+
and that would equate to less than 20% increase in fps in wow + therefore isnt worth spending $400 on - wasnt far out

(infact if he did upgrade his cpu he would end up with onboard video instead of a Geforce 6800GT so far from gain frame rate he's lose it)

 

[Albuquerque]

well i had a look at the game tests and the e6700 hardly "trounces" the fx62 its about 15% ahead which is bugger all

so my arguement that e6600 = fx62 = just over a amd 5000+ = 35% faster than a 3500+
and that would equate to less than 20% increase in fps in wow + therefore isnt worth spending $400 on - wasnt far out

(infact if he did upgrade his cpu he would end up with onboard video instead of a Geforce 6800GT so far from gain frame rate he's lose it)

The E6700 certainly "trounces" the FX62 by my definition: cheaper by double-digit percentage, faster by double-digit percentage and more power efficient by double-digit percentage. If I had said "annhilate" then I might concede...

And why are you repeating exactly what I said? Need a reminder?

The E6600 would be significantly faster than the 3500, without question. I'm also sure a processor upgrade wouldn't hurt, but again, if you're using a 6800GT to play at 1920x1200 then the processor probably isn't your biggest bottleneck.

I said the E6600 would be significantly faster than the AMD 3500+ that he's using, but I didn't suggest that as the option he should take. And I was right on both accounts, thank-you-very-much.