My final setup check it

after long debate among myself i came up with this:

CODEGEN BRIZA-63-CA Black/Silver Steel ATX Mid Tower Computer Case 350W P4 ready w/ 12cm fan 20+4Pin (775Ready) Power Supply - Retail

Gigabyte GA-8I915PL-G Socket T (LGA 775) Intel 915PL ATX Intel Motherboard - Retail

Intel Pentium 4 630 Prescott 800MHz FSB LGA 775 EM64T Processor Model BX80547PG3000F - Retail

Kingston HyperX 512MB 184-Pin DDR SDRAM DDR 400 (PC 3200) Unbuffered System Memory Model KHX3200UL/512 - Retail

eVGA 128-P2-N368-TX Geforce 6600GT 128MB GDDR3 PCI Express x16 Video Card - Retail

LG Black IDE DVD Burner With 5X DVD-RAM Write Model GSA-4167B BK - OEM

final price is:606 (old price 656) with shipping
please review and give ur opinion or correct any incompatibility

my main problem was that i have a p4c 533mhz, 2.8ghz cpu but itsa socket478 $ht just wouldnt work out:
if i go with ddr i cant use ddr400 and if i go with ddr2 theres no motherboards that support both pci-e and ddr2 for socket478 so im kinda stuck on that one all day
anyone got better solution than this? or anyone wants my 2.8 cpu? :iagree:


350watts isn’t enough power

need dual channel ram with a socket 939 amd 64

  1. I’ve built several computers using good PSUs (Codegen, Q-Tec is not good) and they’ve all been fine (similar setups, 6600GT instead of a X800GTO for instance). The ones In-Win includes are decent, not great but they’ll do fine. If you want to run Crossfire (ATi’s type of SLI) you’ll need to get a better PSU. If doubt feel free to inform why it wouldn’t be sufficient.

  2. No, you dont. Socket 939 CPUs does although have two memory controllers which increases performance but we’re not talking about 10% or so, rather 3% and getting two 256Mb memory sticks would not be a great idea since most CPUs still are being picky about populating 3 or more DIMM-slots. Dual Channel (matched RAM) is a marketing bullshit, non matched will do fine as long as they have the same specifications (perferably same manufactorer and model). In fact I’m using two non matched Corsair Value Select memory sticks (512MB if you’re wondering) right now, one was purchased half a year ago and the other one three weeks from now and on top of the from different retailers.

Post your results from everest latency! or reads or writes!
mine top the charts, course i have 2 ballistic ddr 500(cas 2.5) 256 megs
One memory controller can handle dual channel well, and even a 10% performance increase is significant. My sticks weren’t a matched pair, but they sell for 60$ total!

rate PS’s by the +5V and +3.3+V and +12V amperage rails
“Second, the new processor socket offers dual-channel memory access to the owners of the Socket 939 platform. I can’t say that the two channels give the Athlon 64 a great advantage in speed (the performance gain from enabling the second memory channel is 3-5% in average).”

06/01/2004 | 07:59 AM
your link

a newer analysis and YMMV depending upon use

welli broke down y problem with ddr2 in socket type confusion thread
my main thing in this is to spend less money and get a good setup
power supply will def go as soon as i find good and quiet one any suggestions?(pref with disconnecting wires whatever its called)
i dont know if the mobo i got here is good overclocker or not i heard abits are good

I know Antec make a quiet PSU, with connecting cables (where you only connect the power cables you need to the PSU, I believe that’s what you mean). I’ve seen them online, they’re called NeoHE.

Edit: no personal experience of these, I’m yet to receive mine.

And I’ve had very good experiences with Abit mobos…

@ rbrtpl
So… Lemme ask you this, did you even have a look at the rest of the test?
It still shows as I said ~3% performance boost except for the times when 2Gb RAM is used… If you’re going to benchmark at last use the same about of memory all the time.

@ raverx3m
The AMD setup is noticeably faster than the Intel setup…

While speaking of Antec they’ve been reports of incompatibility and I wondn’t really recommend them since you can get better PSUs (read more quiet) for the same money. Have a look at the Seasonic S12 PSU-family for instance, you can find a review at no modular cables though.

The CPU has an integrated dual-channel 128 bit memory controller, which we use to drive Corsair’s memory sticks to the limits of their performance. It is vital in this kind of comparison to use high-quality RAM, to ensure that the memory controller has room to stretch its legs, so to speak.

if you look at my signature you will see I am familiar with good ram in dual channel mode and the potential of venice cpu’s

Point being? No one ever said anything about overclocking.

no the point is, the better the cpu, the more ram becomes the bottleneck,
on a crappy system dual channel has little impact

Eh what? As you’ve already seen dual channel doesn’t give much of a performance boost, that’s memory bandwidth. Overclocking affects a lot of things, not only memory bandwidth.

explain this

Could be that Half Life 2 likes more than 1Gb memory?
I do find Techconnect’s “test/benchmark” odd since they’re more or less comparing 1Gb to 2Gb rather than single to dual channel.
Anandtech seems to have come to the same conclusion as Xbitlabs…

that’s testing L2 channels

Oh I misread, I read it as additional 2 sticks for dual mode…
Oh well, no one else has come up with Techconnects results so I’ll go with the others.

Until we got the amd64 939 on board memory controller, the cpu was often the bottleneck in any operation, at higher clock speeds memory starts to impact
the tests more. That’s why tests with a P4 show little difference.
If you take everest home edition and just look at memory reads and writes,
you see a clearer picture of the potential of dual channel memory. Latency wise there’s no real advantage, just bandwidth.

noone is benching single channel vs dual channel any more, why?
Dual channel won! And 2 sticks of smaller memory aren’t necessarily
more expensive than a single stick of 2 times the size, often they are less.

I bought 2 sticks of this for 65$ and I blow away everest mem benchs
with my socket 939.
I tried 2 other sets before getting these.

I think you should try something real instead of some synthetic benchmark, Everest doesn’t say a shit about overall performance (which I’m talking about). Both Anandtech and Xbitlabs ended up with ~3% performance increase (real world applications) despite the godly dual channel mode.