New AMD build - *budget* high performance gamer for 19x12

[richierich1212]

My best Cinebench run so far

 

[evolucion8]

Originally posted by: richierich1212
My best Cinebench run so far

Is a nice score, my CPU at 3.82GHz scores about 16,294, which proves that Phenom II is competitive at a nice price.

 

[soccerballtux]

Originally posted by: Rhoxed

Originally posted by: soccerballtux

Originally posted by: Rhoxed
synthetic tests and gaming performance are both relevant IMO

not all enthusiasts are gamers, many buy this hardware for folding/image editing/file management/video work/modeling etc.

if you were comparing video cards i would stick to games, but seeing as you are comparing CPU's and different architecture even, i would opt to add in a few synthetic benches, maybe even some x264 work (grayskys comes to mind)

-----

as for your CPU dilemma i would say go ahead and jump on the 720, you would be satisfied even if it didn't unlock (on the other hand the 740 is about to come out IIRC @ 3ghz)

if you feel you NEED a quad, grab the 940BE and use K10STAT to downcore to a triple core for comparisons on x2 > x3 > x4 performance gains.

Synthetic benchmarks are a farce:

My my. Swap CentaurHauls for AuthenticAMD, and Nano's performance magically jumps about 10 percent. Swap for GenuineIntel, and memory performance goes up no less than 47.4 percent. This is not a test error or random occurrence; I benchmarked each CPUID multiple times across multiple reboots on completely clean Windows XP installations. The gains themselves are not confined to a small group of tests within the memory subsystem evaluation, but stretch across the entire series of read/write tests. Only the memory latency results remain unchanged between the two CPUIDs.

At the very least, this suggests some incredibly sloppy coding on Futuremark's part, as the company may be enabling or disabling CPU optimizations based on a processor's vendor name in CPUID instead of actually checking CPUID for SIMD support. In this case, PCMark 2005's memory subsystem test doesn't appear to be aware that Nano supports SSE2 and SSE3, and is instead running a decidedly less-optimized code path. There are two factors, however, that make this explanation a bit difficult to swallow.

First, there's the issue of timing. PCMark 2005 was released (obviously) in 2005, and was obviously coded with an eye towards supporting current and future processors. This is standard operating procedure for Futuremark, which always builds benchmarks designed to last for at least a year, and often two. VIA's C5N-T (Nehemiah) core may have only supported MMX and 3DNow!, but the C7 launched in 2005, and that processor supported SSE2 and SSE3 from day one. Even if proper extension support wasn't built into the first version of PCM2K5, we tested version 1.2.0, and that patch was released on or around 11-29-2006.

Second, there's the issue of performance when Nano is identified as AuthenticAMD. If performance between the AMD and Intel CPUIDs was identical, there wouldn't really be a story here, but it isn't, and that's curious. Futuremark could plausibly argue that VIA's C3/C7 processors weren't exactly on the radar back in 2004-2005, but AMD and K8 certainly were, and K8 launched with full SSE and SSE2 support, with SSE3 added in 2005.

None of this constitutes proof of wrongdoing, but it flies in the face of Futuremark's neutrality claims. Bad code is a fact of life, but companies that write benchmarks for a living and sell those benchmarks as evaluation tools have a responsibility to ensure that their software delivers the neutral framework that it promises. Based on the information I've gathered thus far, it seems Futuremark may have created three paths?one for Intel, one for AMD, and one generic "other" path. There's nothing wrong with optimized code paths, but our results would seem to indicate that some paths are decidedly more optimized than others.

i understand the disadvantages in some of these tests, i was more referring to tests such as
winrar
everest
grayskys x264
cinebench

but as for comparison of AMD vs AMD you can see improvements from the hardware reguardless if the test runs faster using an intel or not.

if you are comparing intel in this, i still do not see a problem, as the gaming performance will not rely on the results of these synthetic tests either. Maybe i am just used to seeing these skewed results and understand intel just performs some things faster.

but of all people you should know this coming from an intel e8*00 series to an AMD 720BE @ 4core (by the way, what improvements/decreses did you notice from this switch?)

Oh ok. With the exception of Everest I would consider all of those real world.

 

[richierich1212]

Originally posted by: evolucion8

Originally posted by: richierich1212
My best Cinebench run so far

Is a nice score, my CPU at 3.82GHz scores about 16,294, which proves that Phenom II is competitive at a nice price.

I'd say so, since the 550 is 1/3rd the cost of the Q9650 hehe.

 

[apoppin]

well, actually the 550 X2 Phenom II kicks ass over the Q9550S in a few games - at the same clock = 3.1 GHz - using 4870-X2 at 19x12 and 16x10 resolutions [4xAA/16xF - ultra details and DX10 when available]
- very nice!!

i am almost done with the game benchmarking at stock 3.1GHz and have a few synthetic tests to run
- then i overclock her as far as she will go and i will see what i get against Q9550S at 4.0GHz

My fellow editor is sending me his Athlon 250 X2 this week to ALSO compare with the 550 and the 720-X3 i am getting

so far, i am very impressed with a cheap Dual core Phenom II in gaming
:thumbsup:

 

[Rhoxed]

sorry it took so long to reply, been very busy with work.

here is a set of everest tests with NB ranging from 2200 to 2800 clockspeed was kept between 3700-3800 for all tests, except 2800NB done at 3.9ghz (this is also done all on multiplier, so it also shows advantages of ram speed scaling and increase)

All tests done on 720BE @ 4core Gigabyte 790X-UD4P 4GB G.Skill-1066

3.757Ghz NB @ 2.2Ghz RAM = DDR2-884
http://img505.imageshack.us/img505/444/2200nb.jpg

3.8Ghz NB @ 2.3Ghz RAM = DDR2-924
http://img191.imageshack.us/img191/5688/2300nb.jpg

3.735Ghz NB@ 2.4Ghz RAM = DDR2-964
http://img191.imageshack.us/img191/8787/2400nb.jpg

3.765Ghz NB@ 2.5Ghz RAM = DDR2-1000
http://img114.imageshack.us/img114/2840/2500nb.jpg

3.784Ghz NB@ 2.6Ghz RAM = DDR2-1044
http://img4.imageshack.us/img4/9088/2600nb.jpg

3.794Ghz NB@ 2.7Ghz RAM = DDR2-1084
http://img191.imageshack.us/img191/3823/2700nb.jpg

3.934 NB@ 2.8Ghz RAM = DDR2-1124
http://img4.imageshack.us/img4/1986/2800nb.jpg

As soon as i get more time i will run tests with RAM set to ddr-800 and 1066 and JUST scale NB speeds, but this just shows what a little OC tweaking can do (especially comparing NB at 2200 and almost stock ram to NB at 2800 OC'd ram and only ~100 mhz core increase)

 

[apoppin]

thanks

i have decided not to run any synthetics beyond futuremarks 3Dmark06 and Vantage [which i desperately hate except to track changes in the same system; but everyone else does them]

This is a gaming PC .. who cares about encoding

Lame

[lame .. ...
]

anyway, guess what i just ordered for $ 119 shipped
- oh .. and Athlon 250 X2 will also be here Friday
:Q

fun .. fun .. fun

i am also NOT going to bother with the stock AMD HS
- it is a frickin' toy - who *cares* about it .. no gamers do .. they get a real cooler
... i am just going to compare cooler master hyper n212 with thermalright .. and what is cool is that my Scythe 120 MM fan appears interchangeable with the Cooler Masters' - so i can run TWO fans at once in "push-pull" !!
:evil:

 

[Rhoxed]

in that case a small comparison in 3dmark06 of 720BE (3core and 4core) vs 940BE both using triple xfire 3870's and 4850's to show the 4th core advantage (best of luck unlocking that 720BE)

720BE(X4) @ 3.85ghz + 3x HD4850
http://service.futuremark.com/compare?3dm06=11295979

940BE @ 3.937ghz + 3x HD4850
http://service.futuremark.com/compare?3dm06=10372826

720BE(X3) @ 3.85ghz + 3x HD3870
http://service.futuremark.com/compare?3dm06=11092172

940BE @ 3.937ghz + 3x HD3870
http://service.futuremark.com/compare?3dm06=11218327

 

[apoppin]

Originally posted by: Rhoxed
in that case a small comparison in 3dmark06 of 720BE (3core and 4core) vs 940BE both using triple xfire 3870's and 4850's to show the 4th core advantage (best of luck unlocking that 720BE)

720BE(X4) @ 3.85ghz + 3x HD4850
http://service.futuremark.com/compare?3dm06=11295979

940BE @ 3.937ghz + 3x HD4850
http://service.futuremark.com/compare?3dm06=10372826

720BE(X3) @ 3.85ghz + 3x HD3870
http://service.futuremark.com/compare?3dm06=11092172

940BE @ 3.937ghz + 3x HD3870
http://service.futuremark.com/compare?3dm06=11218327

thanks for posting!

3DMark06 doesn't really show much as it is pretty dated
- Vantage is better but it is not free

Have you tried the free World-in-Conflict demo stand-alone benchmark?
- as a game bench, it will definitely show both scaling from overclocking and from more cores than 2

Also the Stalker Clear Sky benchmark is free and pretty intensive; it is 12 minutes long;

The Phenom II 720 just DROPPED in price to $119 - still free shipping !!
- it is SO ordered

 

[Sylvanas]

@ Rhoxed: Are you on a 32bit OS?

 

[Rhoxed]

Originally posted by: Sylvanas
@ Rhoxed: Are you on a 32bit OS?

Yes, the only OS i had laying around ATM other then vista x64 (which i am not fond of)

 

[apoppin]

Originally posted by: Sylvanas
@ Rhoxed: Are you on a 32bit OS?

You keep mentioning it as a big deal
- what is is about 64-bit vs. 32-bit with AMD processors that does not affect Intel ?


i switched to Vista 64 - not for gaming performance [as there is NO improvement with Vista 64 for *most* games] - but for more system RAM later on.

Phenom II 720 X3 should be here tomorrow. Not bad for $119 shipped
- let's see if i am "lucky" with unlocking .. or not
. . . this time i DON'T CARE

tonight i try to find 550 x2's max oc with my Hyper n212 and two 120 mm cooling fans
- for the first time in years, i have a quality CM Gladiator 600 case with built-in excellent airflow and cooling*
- i will study the 'tips' thread and post some questions if i have any - there and/or here


* Quick question:

i have the option to turn my CM Silent Power 600w PSU "upside down" in my case:

1. the "regular" way sucks the air in from the bottom below the case [thru a dust filter supplied] and blows it out the back of the PSU out the case. But the PSU fan is effectively NOT affecting case air flow
2. The way i have it - "upside down" - the PSU fan draws warm air from the bottom of the inside of my case and blows it out the back of the PSU

- the DIFFERENCE is that NOW the PSU is part of the case air flow [and AFaiK the PSU is not affected badly by the warmer air coming from inside the case]

comments, please

 

[soccerballtux]

Hm that's a good idea, I think I might do that myself...would be fun to see how if any that affects case temps...

I was same way about the 720be-- I realized if I couldn't get it to unlock, I could still OC it to at least 3.6Ghz and cpu-nb some, at which frequency it would be plenty fast enough to be a sidegrade from my e8400 @ 3.6ghz.

BTW-- we have the same case, heh.
Consideration for me is then I would have 2 fans blowing in and 2 fans blowing out. Currently I have 2 in and 1 out; the positive pressure helps push hot air from my cpu out the top (my Ultra-120 can only mount vertically on AM2/3 sockets) (with one fan sucking the rest out the back).

 

[Sylvanas]

You keep mentioning it as a big deal
- what is is about 64-bit vs. 32-bit with AMD processors that does not affect Intel ?

It is a known occurrence that a 64bit OS limits OC's in Phenoms compared to a 32bit- the difference can be as much as 300mhz.

Yes, the only OS i had laying around ATM other then vista x64 (which i am not fond of)

Fair enough, I was just wondering because those clocks are very good and I have not seen anyone get near 2.8ghz NB on 64bit. Even at 1.45v NB VID my 955 will not boot at 2.8ghz. Sweet OC, enjoy .

Hope this one goes better for you apoppin.

 

[apoppin]

Hope this one goes better for you apoppin.

i *GUARANTEE* IT WILL :Q

This time i have 3 cores and don't care about the fourth except as a possible bonus

X-3 is scheduled to arrive today. Today i take off the stock HS/Fan on 550- stick on a Cooler Master Hyper n212 - with 2 120 mm fans - and i go for it
:evil:

Friday i should get my 220 X2 athlon ... lots of benching going on at apoppin's this week




frankly i don't see *why* a 64-bit OS limits O/C'ing


AND .. last question before i head for my (immediately) paying job []
Since this is an "all AMD" platform, should i use Overdrive to OC it?

 

[soccerballtux]

Originally posted by: apoppin

Hope this one goes better for you apoppin.

i *GUARANTEE* IT WILL :Q

This time i have 3 cores and don't care about the fourth except as a possible bonus

X-3 is scheduled to arrive today. Today i take off the stock HS/Fan on 550- stick on a Cooler Master Hyper n212 - with 2 120 mm fans - and i go for it
:evil:

Friday i should get my 220 X2 athlon ... lots of benching going on at apoppin's this week




frankly i don't see *why* a 64-bit OS limits O/C'ing


AND .. last question before i head for my (immediately) paying job []
Since this is an "all AMD" platform, should i use Overdrive to OC it?

People say it's buggy; I found it too busy-- I just need multiplier and cpu-nb.
I OC the NB through BIOS; the multiplier with PhenomMsrTweaker.

 

[Rhoxed]

AMD Over Drive is nice to find max OC without having to restart many times

i find it very useful when first testing the OC on a chip, then start dialing it into bios

really its all preference, do what feels more comfortable to you

 

[evolucion8]

Originally posted by: apoppin
frankly i don't see *why* a 64-bit OS limits O/C'ing


AND .. last question before i head for my (immediately) paying job []
Since this is an "all AMD" platform, should i use Overdrive to OC it?

At least it does when an Intel CPU is used, my QX6850 at 3.62GHz was able to pass Prime95 in 25 hours with no issues, but when I moved to 64 bit, it would crash between 45 minutes or up to 2 hours, probably the longer words in the 64 bit environment are able to spot up easier inconsistences in the CPU or cache.

Yes, use Overdrive, it's a nice feature.

 

[soccerballtux]

update us

 

[apoppin]

OK

720 X3 is here and the batch number is: 0904EPMW

i am done with testing 550 x2 with 4870-X2 at stock 3.1 GHz with 14 games [including ARMA2] and now i am running AMD's Overdrive to overclock my CPU [with the stock HS/fan - just for fun ]

Now, it takes a long (long) long time and it went totally unresponsive so i had to reboot even though the temps did not go very high (35C up from 24C according to Everest) nor did i get much of an OC

 

[soccerballtux]

High chances of unlocking success with the 0904's.

 

[apoppin]

Originally posted by: soccerballtux
High chances of unlocking success with the 0904's.

Don't do this to me

i am dying to find out if it unlocks; however, i am stuck O/C'ing with Overdrive and it is like watching your fingernails grow kind of "exciting"
:roll:

slooooow .. and i am just on the stock AMD cooler because i want to see where the temps go up
- i might toss X3 in just to see if it unlocks when i put the Hyper n212 in .. hopefully tonight

EDIT: i am at a point where the LCD is black and the fan is spinning up and down


i think i am going to skip OD and OC from the BIOS
- i know O/C'ing with intel and now we'll see how far i get with AMD

EDIT:

So i upped the FSB and added +.01 v to stabilize it; running Everest's stress test and, so far, it is fine at 3.50 GHz - temps with the stock cooler are mid 40s - so i will not go too much higher before switching to a much better CPU cooler
- i will let Everest run and then try for 3.60 GHz to see where the stock cooler becomes inadequate

2nd EDIT:

3.6 GHz and still with the stock cooler. Minimal voltage applied everywhere - except to the CPU +.1 [1.4 v]
- temps are not up over 3.5 GHz' mid-40s

all i did was raise the FSB [and mini voltage boosts; took everything off 'auto']

 

[Sylvanas]

Keep in mind raising the HTT frequency (or 'FSB' even though it isn't a Front side bus) up from stock you will also be increasing the memory frequency. I find it much easier to just use multipliers, since that's what so great about a Black Edition CPU. Keep going Apoppin .

 

[apoppin]

Originally posted by: Sylvanas
Keep in mind raising the HTT frequency (or 'FSB' even though it isn't a Front side bus) up from stock you will also be increasing the memory frequency. I find it much easier to just use multipliers, since that's what so great about a Black Edition CPU. Keep going Apoppin .

Keep going?
- of course

i ran Everest all night just as a stability test and i watched the temps at 3.6 GHz; it is fine and it passed

This AM i also raised the HTT frequency some more to put my 550 X2 at 3.7 GHz and it is also stable
- this is with the STOCK CPU cooler btw :Q
- AND still using the stock multi- btw - she might get well over 3.8 GHz !!

Now the temperatures are starting to rise in the high 40sC and the next step will be with a proper cooler [hyper n212] tonight as i find the max OC

This Fatal1ty RAM is decent; at least it is going way over 950MHz and it is not giving errors [so far]

Vantage:
550 X2 @ 3.1 GHz/4870-X2 = 9890 Overall/13838 GPU
550 X2 @ 3.7 GHz/4870-X2 = 11019 Overall/14762 GPU

Vangtage mini game frame rates:
550 X2 @ 3.1 GHz/4870-X2 = 38.4/42.75
550 X2 @ 3.7 GHz/4870-X2 = 41.82/44.71

 

[MegaWorks]

So did successfully unlocked the fourth core apoppin?

I might build a similar setup, I have 8GB OCZ Blade PC2-6400 lying around. My current setup is a DDR3 4GB, and I don't want to buy a new set of DDR3 8GB.

This thread's having a bad influence on me to make a purchase..must hold..on...