**Is paying for the e6400 worth it, or should i buy a e6300 or even e4300?
the e6300 and e6400 both have 266mhz bus, but the e6400 has an 8x multiplyer.
the e4300 has a 9x multiplier but im not that experienced in OCing and i dont like the idea of over clocking a chip 1.14Ghz above its rated level with stock cooling (which is the 50 bucks difference to a 6400). I wouldn't have to raise the bus speed as high on the other chips either, which is better for stability. right?
I had to make the same decision myself.
Personally I went with the 6400 because I am not a pro overclocker.
I saw 400mhz x 2(800) x 8(cpu multiplier) = 3.2 Ghz.
With 800 mhz ram(400x2) putting fsb400=ram(400x2) at a nice balance.
Also come April there are price-cuts on all of the C2D's so you might wait a month or 2 if you can.
I'm not sure of the differance between the S3/DS3.
Ok I think the only differance is raid 0/1 controller on the DS3 meaning I should have gotten S3.(I notice after checking specs for 15 mins.)
Also I got the X1900 Pro because it beat the 7900gs but the 7900 might be cheaper/is more ocable if your looking at that.
the only difference between s3 and ds3 is that ds3 has solid state capacitors. i work at a shop that fixes computers and the number one problem we have with motherboards (specifically dell mobos) is that their capacitors are blowing up. obviously gigabyte makes much better mobos but thats the only difference.
Regarding DX 10 yes DX 10 might/should be coming out later this year.
Nvidea's 8800 series is only cards supporting DX10 thus far.
Personally I wouldn't buy a DX10 card yet.
If I was I'd buy the 8800GS which is the cheapest.
If ATI(AMDTI) ever releases their RD600(codename) cards they should all go down in price.
I'd buy a lower end card(Or mid to high such as the 7900gs/x1900pro) and still run XP until DX10 comes out.
Then once it's out/working and you have a game which supports it(DX10) I'd upgrade to vista/DX10/a new video card.
The Amount of the OC is not as important as much as how much over design the OC is doing.
Example - The E4300 is built using the same basic process as the E6800.
The E6800 runs at 2.9Ghz and the E4300 at 1.8.
If you OC the E4300 from 1.8 to 3.0 Ghz, it would be a 1.2Ghz OC which likely can run fine on air cooling. This is because the E4300 is now running basically within Design specs despite having a huge OC. It may not run at these speeds since it was not tested there, but the majority of chips should do so since it is in design specs.
The E6400 would have the same top design limits, but the chance of it failing at 3.0 is lower since it was tested at higher speeds already.
If you are going for a very final CPU speed I would go with the E6400 since it was tested higher. If you are going for a moderate OC such as 3.0 (in design limits so I consider moderate) then I would do the E4300 and pocke the $45 or put that into another component such as a faster GPU or better PSU.
**Is paying for the e6400 worth it, or should i buy a e6300 or even e4300?
the e6300 and e6400 both have 266mhz bus, but the e6400 has an 8x multiplyer.
the e4300 has a 9x multiplier but im not that experienced in OCing and i dont like the idea of over clocking a chip 1.14Ghz above its rated level with stock cooling (which is the 50 bucks difference to a 6400). I wouldn't have to raise the bus speed as high on the other chips either, which is better for stability. right?
please offer advice
thanks
What zenmaster said. Put another way:
The e4300's 200Mhz FSB gives you lots of breathing room. You might be raising the FSB more on an e4300, but you're not exceeding the designed speed of the mobo, RAM etc. You can take an e4300 up to 266Mhz (2.4Ghz) without putting any strain on mobo or RAM whatsoever. Taking an e6400 to 2.4 would require a 33 Mhz overclock, which wouldn't be a problem, but requires better than cheapo DDR2-533 and may have issues with flaky onboard video controllers and the like. You could stick an e4300 in the cheapest C2D mobo that has adjustable FSB and be guaranteed to get to 2.4 Ghz because even the cheapest possible C2D mobo is designed for 266 Mhz.
Also, as zenmaster said, all chips in the family come from the same manufacturing process. You might get the rare 'bad' chip that won't handle 3.5+Ghz, but that could happen with an e4300 or an x6800. There are times when processors are being produced near the limits of the process and a >10% OC requires a 'lucky' chip but C2D's are so far from design limits that almost any chip in the C2D family will OC dramatically.
The e4300 also requires lower speed RAM. You could take your e4300 to 3 Ghz with DDR2-667 at stock speeds with 1:1 memory multiplier. Your e6400 would require DDR2-750, and an e6300 would require DDR2-857.
That might shave a few dollars off the build, in addition to the $50 on the processor.
An e6400 has advantages over an e6300 due to multiplier. At stock speeds there are also good reasons to buy the e6400.
Generally the processors aren't the limiting factor in overclocking, the RAM, MB and onboard components are. Or, I should say, the processors give out at 3-4 Ghz, while in most setups the mobo or onboard video fails before that.
Allendale is a different die than Conroe, but the two are almost identical with the exception of cache size and virtualization, esp. compared to, say, the Netburst family. They all appear to overclock fairly similarly. I haven't seen anyone claiming 5Ghz with an e4300 yet, but I have seen postings at over 200% stock. Of course, just because it is on a message board doesn't make it true, but successful mid-3Ghz e4300 postings are very common.
yeah "Q2 2007" which is April-June is when intel is releasing the new numbers
so really the only advantage of buying a 6300 or 6400 is if there at stock speeds, and the performance increase for overclocking cant be seen because of limitations on the mobo and memory...right?
how well will i be able to overclock with those components?
yeah "Q2 2007" which is April-June is when intel is releasing the new numbers
so really the only advantage of buying a 6300 or 6400 is if there at stock speeds, and the performance increase for overclocking cant be seen because of limitations on the mobo and memory...right?
how well will i be able to overclock with those components?
and i need a PSU (random)
The mobo should be good to 3.2+
Processor and memory are bigger questions I think.
Then again you can always get a finicky board.
As I said before(I think) each part is different from the other 100 they came off the line with so no promises with anything but
I would say somewhere around 2.6 on stock voltages and probably 3.2 with slight voltage increases.
Also won't necessarily be cpu/mobo/memory for your OC's bottleneck.
(meaning could be any of them that bottleneck it)
yeah "Q2 2007" which is April-June is when intel is releasing the new numbers
so really the only advantage of buying a 6300 or 6400 is if there at stock speeds, and the performance increase for overclocking cant be seen because of limitations on the mobo and memory...right?
how well will i be able to overclock with those components?
and i need a PSU (random)
There are lots of reports of the S3 being good to low 400's, which means:
e4300 3.6Ghz
e6300 2.8Ghz
e6400 3.2Ghz
The Asus P5B Deluxe has lots of stories of successful 500+Mhz FSB overclocking, which would be 4.5Ghz on an Allendale, or 3.5-4Ghz on a Conroe. I saw one site using a P5B listing an FSB over 600 with an undermultiplied CPU.
The DDR2-800 will go to 400 Mhz FSB with a 1:1 multiplier and not exceed stock speeds. 4-4-4-12 DDR2-800 should be good for higher speeds if you open up the timings to 5-5-5-15 or so. I think wusy's guide lists it at DDR2-900 or so. Also, if you look at the THG review from a few days ago, you'll see some possible results at extreme voltages. Doesn't mean your RAM will handle the voltage, but it is possible.
The CPU is hard to say, lots of reports of mid-3's, some reports of over 4 Ghz.
PSU's typically work or they don't. If the PSU provides enough power and is made with good quality parts that won't fail in a few months, it is fine. Lots of people recommend $200 PSU's, which is fine if you have the money and at that price they should never fail, but I personally have a problem with a power transformer being the single most expensive piece of my computer.
Actually there is a lot more to "work or don't work" for a PSU.
One key factor is clean stable power.
This becomes more critical for a serious OC since the greater the power fluctuations in your PSU the higher your voltage will need to be to handle the lower fluctuations.
Unless on a tight budget, I would not get anything less than a Tier 3.
[quote="
PSU's typically work or they don't. If the PSU provides enough power and is made with good quality parts that won't fail in a few months, it is fine. Lots of people recommend $200 PSU's, which is fine if you have the money and at that price they should never fail, but I personally have a problem with a power transformer being the single most expensive piece of my computer.[/quote]
And who designates the tiers? I've heard a lot of crap spouted about PSU tiers and never seen the results of a lab that has tested 100 units of a single model, much less dozens of models.
Having several years experience in Quality Control at a large manufacturer, it really pisses me off when people talk about PSU's differently from other electrical devices in terms of quality.
I had a disclaimer in my original post that I removed, and it was: "and 'work' doesn't just mean power on, it means provide enough clean power for the system." If you do have fluctuations in power quality from a PSU, it generally indicates a failing component or joint. It might take months for that solder point to fail, or for that capacitor to finally blow out, but the PSU is still in a death spiral.
The "Tiers" in the previous link are defined by the author of the link.
I have examined her reviews and seen reviews of her reviews.
I choose to trust her judgement.
These "Tiers" are generally based upon both testing and analysis of the components used within the system. The reports are also based upon feedback which could require further analysis.
Example - If I test one set of car tires made of rubber and another set made of paper mache and find that the rubber ones last much longer you are correct that I can't make a case that the rubber ones will always last longer. But I can use analysis of the mateials use to construct the tires to see of it explains the difference. In addition, many of the PSUs built by different manufacturers share common components which in essence get tested many many times to help affirm the reliability of the subsystems.
And fluctuation can occur even when not failing straight from the box.
If you read a large number of indepth reviews of different power supplies ranging for poor to good you will see this.
Under normal operating conditions this is not an issue because the CPU and other components are operating well within engineering guidelines.
As you OC the system, then small fluctuations can cause more and more of an issue.
Now, Would I recommend a Tier 2 PSU for a $300 build. Of Course not.
But If I was looking at en E6600 and an E6700 whith a $100+ price difference despite being manufactured Identically, it may make more sense to get the E6600 and put some of that $100 to a higher quality PSU.
The "Tiers" in the previous link are defined by the author of the link.
Where? If you read the thread, you'll see that there are no guidelines for each tier, just 'gut feeling'.
Quote :
I have examined her reviews and seen reviews of her reviews.
I choose to trust her judgement.
Yeay for you. Not saying that this list of PSU's is bad, but people talk about PSU tiers as if there is some official industry testing for them.
Quote :
These "Tiers" are generally based upon both testing and analysis of the components used within the system. The reports are also based upon feedback which could require further analysis.
I don't disqualify this person as a valid reviewer, but people other than the creator of that site use these terms with no real evidence or scientific backing. Hell, if you read the thread you'll see posts like, "I don't know why I made that one a tier 2, they have as good components and test as well as the tier 1's." I think proves pretty definitively that there are no clear metrics for determining "Tier".
Manufacturers also change components frequently. When one supplier changes rates or is out of stock, a new supplier is found for a part that meets the same requirements. So, an eval model or the unit you get might have high quality capacitors from Fuso, but the model I get has lower quality ones. They are probably pulled out of the same bin, as they are treated as 'like' components, so you could even get a unit with mixed components.
Quote :
Example - If I test one set of car tires made of rubber and another set made of paper mache and find that the rubber ones last much longer you are correct that I can't make a case that the rubber ones will always last longer. But I can use analysis of the mateials use to construct the tires to see of it explains the difference. In addition, many of the PSUs built by different manufacturers share common components which in essence get tested many many times to help affirm the reliability of the subsystems.
And if your rubber tires have holes in them due to poor manufacturing control, it doesn't mean mine will. Unless we go and review the purchasing, manufacturing, and QC processes for an entire plant, a single good review does not make a quality line. A well run shop using all tested generic components and a stolen design can produce better knockoffs than the real company if they have tighter processes.
Quote :
And fluctuation can occur even when not failing straight from the box.
If you read a large number of indepth reviews of different power supplies ranging for poor to good you will see this.
Under normal operating conditions this is not an issue because the CPU and other components are operating well within engineering guidelines.
As you OC the system, then small fluctuations can cause more and more of an issue.
Two types of fluctuation are common:
1)Voltage drop under load, typically at the very high end, ie in the last 10-15% of a rail's rated output.
2)Poor quality. Part is failing, it will just take a while.
The first happens on all PSU's, some more than others. This is why sizing recommendations usually include a 10% buffer. Overclocking increases draw, there are some fairly good calculators out there, but the rule of thumb is the same die at same volts increases consumption linearly with Hz, and ditto for volts.
The second means you got a crap unit, less likely with a high dollar unit, but there is nothing stopping it from happening from any supplier.
Quote :
Now, Would I recommend a Tier 2 PSU for a $300 build. Of Course not.
But If I was looking at en E6600 and an E6700 whith a $100+ price difference despite being manufactured Identically, it may make more sense to get the E6600 and put some of that $100 to a higher quality PSU.
Is price a determining factor in tiers? It always seems to play a role. Maybe price plays into the quality of components, or maybe it goes to pay a marketing team to develop a fancy flash website and a cool logo.
If you size your PSU correctly, and a C2D, even overclocked, is not a huge drain like P4's were, and you avoid marketing gimmicks (like 60A on the 5v rail), and you don't buy the absolute cheapest one, you will probably do alright.
You can generally trust lists like the one linked to, but take them with a grain of salt. Your mileage will vary. There are hundreds of manufacturers and reboxers out there and just because yours isn't on the list doesn't mean it is of bad quality. And take lists that say you have to have a PSU rated at 1000w and costing $200 to have a quality unit for overclocking with a grain of salt as well.
