4400 x2, vs SD 3700.

i recently purchased an x2 4400, and have been reading lots about advantages and disadvantages of the proc.
i've also been getting guff from my friends about spending so much on a chip. so i felt a personalized
comparison, between my old proc and this one, was in order. i mainly play counter strike 1.6, and since i got my gtx, CS source, and far cry.
but i've always been a huge multitasker.

initially i decided to run 3dmark03, 05, and the CS:S video "stress" test for this comparison. but after
running 3dmark03 and 05, i decided against them. simply because they aren't why i bought the chips, and 3d
marks are not what im concerned with. so i limited the testing to the video stress test, and super pi.
i ran the stress test 10 times at each 200MHz interval, and took the average. and what i discovered (for myself)
is little less than distressing.
but first the process for each chip.

3700 San Diego
CPU Speed..multi...."HTT"........divider..mem freq......V.......load temp
2200.........11........200..........1:1.......200.............1.30...39
2406.........11........219..........9:10.....197.............1.42...43
2596.........11........236..........5:6.......196.............1.52...46
2749.........11........250..........5:6.......208.............1.58...48 not stable

4400 Toledo
CPU Speed..multi...."HTT"........divider...mem freq.......V.......load temp
2200.........11........200..........1:1.......200.0............1.30...38
2365.........11........215..........9:10......193.5............1.35...41
2409.........11........219..........9:10......197.1............1.35...46
2475.........11........225..........9:10......202.5............1.40...46
2530.........11........230..........9:10......207.0............1.45...49
2596.........11........236..........5:6.......196.7.............1.55...55

all 12 hrs prime stable except for the 2.75GHz on the sandy. i did the 2.75 test, just to see if the frames would go even higher, further implying cpu bottlenecking.

now for the distress. it's known that the x2's aren't as well equipped for gaming as the singles, but i feel
the advantage of multitasking far outweighs what i thought to be "a few frames" loss in game.

results, (stock video clocks, all same components etc. etc.)
(CS:S fps...1024*768, 4AA, 8AF, all high settings, water reflect all)
3700
Stock
1.PI
1 mil. 0 min 38.86 sec... 4 mil. 3 min 20.33 sec... 32 mil. 33 min 54.86sec.
2.CS:S fps: 182.02

@2.4
1.PI
1 mil. 0 min 36.27 sec...4 mil. 3 min 3.05 sec...32 mil. 33 min 2.73 sec.
2.CS:S fps: 195.86

@2.6
1.PI
1 mil. 0 min 33.69 sec...4 mil. 2 min 51.13 sec...32 mil. 32 min 43.77 sec.
2.CS:S fps: 199.24

@2.75
1.PI
not stable
2.CS:S fps: 208.81

4400
Stock
1.PI
1 mil. 0 min 38 sec...4 mil. 3 min 18sec...32 mil. 34 min 37 sec.
2.CS:S fps: 150.44!!!

@2.4
1.PI
1 mil. 0 min 35 sec...4 mil. 3 min 21 sec...32 mil. 32min 5 sec.
2.CS:S fps: 188.47

@2.6
1.PI
1 mil. 0 min 34 sec...4 mil. 2 min 53 sec...32 mil. 30 min 41 sec.
2.CS:S fps: 187.41

ok, now to play the 4400's game. while multitasking the 4400 performed pretty much with identical
results as when it's not multitasking.

the 3700 while multitasking.
@2.6 w/ 2 browsers open, and winamp.
CS:S fps: 198.01
one word comes to mind. marginal.
@2.6 while ripping a cd.
CS:S fps: 58.91
@2.6 while scanning for viruses.
CS:S fps: 182.49

the distressing part is how at stock speeds, the 4400 gets 35 fewer frames than the 3700.
also, the 3700 gets progressively better results per overclock, while the 4400 tops at around 188 frames.
my intentions for this build were 2 high end GPU's (gtx's now, 6800gt's a few months ago). but it's fairly obvious
that the 4400 is a bottleneck. this completely changes my priorities for this build. i suppose now my
priorities lay in a raptor hdd, perhaps water cooling, and 2gigs of better ram, so i can get rid of that dang divider,
and play BF2. im still going to get a second gtx (i know one of my old buddies from high school must have one),
just to run some tests. but the way things are looking, no way am i purchasing a second gtx.


conclusions.
the multitasking capabilities of the 4400 are absolutely astounding. when i think of a high end cpu, i think of
a proc. that can take what i can throw at it. and this proc is exactly that. on a new install, i can be listening
to winamp, browsing, and be running 3 installers at once, and there isn't even a hicup. i can't keep up with this proc.
no matter how fast i click. and for me, that is testament enough, to buying this cpu. i suppose this thread was simply
to justify spending too much money on a system. i really need to evaluate my priorities in life here. i spend a way
too high % of my paycheck on this machine... and i think it's time to chiiiiiiill.

Wow, the X2s do perform pretty close to single cores in gaming, I'd say an average of 10fps off? Great review showing how X2s aren't too bad in gaming.

well... when overclocked, heck yeah, basically the same. negligible differences to the eye. but the whole 150 fps at stock really bothered me.

I'd say the hit is fps is well worth the multi-tasking benefits.

Thank you for posting. You said it, if you are willing to pay a disproportionate premium to get rid of 'hick ups,' dual core (at this time) is for you.

There are people who have to stop what they are doing just to let the CPU complete one big huge (business) task. X2s are worth the $ for these people unquestionably.


So if you are not one of these people, you need to note that the single core is pretty damn fast & those hick ups are not going to be that noticeable so it's not a good idea to sacrifice other parts of the system just to get dual core.


But if you have the $, dual core will sure 'feel' sweeter for a whole range of reasons.


I don't know, people with money find it difficult to understand how that big price difference affects the pockets of teens /minimum wage people so it's difficult to see the 'hick up' advantages being justified with such a high disproportionate price difference.

Let me also add to this.

If you are coming from a P4 with Hyperthreading (Northwood, Prescott) you are DEFINATELY going to want to go to an X2. Although Hyperthreading isn't the be all, end all there is a noticeable difference going to a single core system with no multi thread management.

Nexus Realized said:

I'd say the hit is fps is well worth the multi-tasking benefits.

Click to expand...

Yeah but is the cost difference worth it along with the hit in fps? I wouldnt think so if all you plan on doing with the machine is hardcore gaming. But for an all around system its looking like these dual cores are king.

Yeah but is the cost difference worth it along with the hit in fps?

Click to expand...

i'd say, for the 4400; no. i mean, $550 is pretty unreasonable when i look at the big picture. especially when considering the 3800's. i'd love to see some similar reviews on the 3800. i can't imagine it performing that much worse (if any), at the same clocks. not to say this proc isn't absolutely killer, but i'd assume the 3800 would be a better investment (considering what forum this is).

This looks like one of those science fair projects where you have to post conclusion, method, intro, etc.!

p|astic said:

but the whole 150 fps at stock really bothered me.

Click to expand...

I would say you had some other issues affecting those scores. As long as the cpu's are real close in speeds, you are NOT going to get 35fps difference. Not to mention at STOCK speeds the single core 3700+ has a 200mhz advantage over the dual core 4400+. You have something else going on. did you accidentally run the single core at 500mhz more than the X2? If you run BOTH cpu's at the IDENTICAL speeds, you are NOT going to have 35fps difference. Now go retest properly and report your results again.

4400 Toledo x2 San Diego 3700
both run stock core speed at 2.2 GHz. as stated before, each overclock had identical RAM and HTT timings, dividers, and clocks.
i too was shocked, and sceptical of errors. until i checked, rechecked, and ran the test 10 times.

Was an attack warranted? Or could you merely have inquired into the accuracy of his results?

Was an attack warranted? Or could you merely have inquired into the accuracy of his results?

Click to expand...

dude, it's all good. i think i didn't state well enough that it was a san diego. i think the old 754's are clocked at 2.4ghz. i <3 every one.

I get a bit....overzealous. Nothing personal intended. Sorry it seemed that way....but I still question the validity of those scores....no way those 2 cpu's are 35fps apart, something else interfered...

therapture said:

I get a bit....overzealous. Nothing personal intended. Sorry it seemed that way....but I still question the validity of those scores....no way those 2 cpu's are 35fps apart, something else interfered...

Click to expand...

You do realize that there has been many site that shows the difference between a single core and x2 right ? The only way for an x2 to out perform a single core cpu is if the game is coded for multiprocessing. (that is with current technology.)

It could easily be windows that interfered, or the game itself, i know some of my games arent fond of my x2, lineage2 for example, will blue screen if i try to set affinity to either one cpu, and critically error if i try to open two instances of it. From a hardware standpoint theres no reason why an x2 shouldnt be a little bit faster in every single thing you do on your pc, because its always runing multiple threads, but from the software side of things, there is still some catching up to do.

Hc000 said:

You do realize that there has been many site that shows the difference between a single core and x2 right ? The only way for an x2 to out perform a single core cpu is if the game is coded for multiprocessing. (that is with current technology.)

Click to expand...

Yes, I know that. I never said the X2 would outperform a single core in a single threaded app.... Just that given equal clock speeds and equipment, the X2 gives up basically nothing in game speed. And depending onwhat you are running in the background, an X2 CAN be FASTER than a single core because one of the cores is dedicated to the game while the other handles the OS and what not...

From his results the X2 does give up some in game speed when running just a game. And yeah the X2 will always be faster when your multitasking. But I really dont think that the difference would be worth the extra cost involved. At least at this point in time. After all we have the M2 socket coming out soon and 65nm process is on the horizon. IMO it just isnt worth the cost to buy a dual core now unless you are into some serious multitasking and need that extra computing power. Other than that its just frills that dont do a whole lot for you.

CarRamrod said:

From his results the X2 does give up some in game speed when running just a game. And yeah the X2 will always be faster when your multitasking. But I really dont think that the difference would be worth the extra cost involved. At least at this point in time. After all we have the M2 socket coming out soon and 65nm process is on the horizon. IMO it just isnt worth the cost to buy a dual core now unless you are into some serious multitasking and need that extra computing power. Other than that its just frills that dont do a whole lot for you.

Click to expand...

Waiting for something "better", you would never get anything new....if you wait for M2, why not go ahead and wait for M3, or the 15nm process after that, or the quad core coming out in 2009 or....you get the point...

And if someone is buying a new cpu TODAY, I think the X2 is almost foolproof and warrants the extra cost, like I said before, XP itself IS multithreaded and the dual core offers incredibly smooth and silky usage even while ripping/encoding dvd, or whatever...and gets those types of jobs done in half the time.

Plus my machines Fold@home about 16 hours a day and I can get double the units done. Even now I am running a dedicated Doom3 server on one core nad using the net and whatnot on the other, and I can even join my own game if I need to

Well as you said it all depends on how you use it. If its just your gaming machine you wont see any real benifits worth the cost of an X2 because the games arent set to fully utilize the dual cores. So in just running a game your not going to see a performance increase. I do see your pint about the waiting for the next thing new but the M2 is JUST around the corner, Q2 of 06 I believe and that is not long to wait at all especially if you want to get the longest life out of a new rig IMHO. Or even just wait until the M2 comes out and the price of the X2's will be lower than now so it would be more cost benificial.