INTEL - Synchronous is not important

[Slappy00]

whee a pissing contest... I ran Futuremark 02 yeserday and got like 5900 or something, lousy video card... I PUNCH THEEE!!!....

as for memory bandwidth I paid a premium to run 500DDR at 1:1 2.5-8-4-4 (hell its rated for 533 but I need better cooling to get the OL' p4 past 3.0Ghertz) and I really think that when you need the bandwidth for games or apps that demand it your performance will slow down if you run a divider you just wait.... mwahahahahah

/rubs hands evilly

 

[stevejst]

Yes man, memory is holding me back else I'd be rocking at 3.3 GHz on 1:1, I am sure it would smoke through my power supply. Don't you like your Pentium, there is no way in hell I would buy AMD processor the way they are now. I just ordered my IC7, we'll see if I can go any higher than 285 I am doing now.

 

[Technonut]

Originally posted by: stevejst

What timings? With Street Racer (2-3-2-5) settings @ 1:1 450MHz, My Sandra unbuffered scores come in @ 3443/3431

I am sure you have better memory than me. Mine 228 runs at 2.5-4-4-8, just auto for GAT.

Then how are you pulling unbuffered Sandra scores of 4102/4063???

 

[stevejst]

Then how are you pulling unbuffered Sandra scores of 4102/4063???

It is FSB, running at 5:4, so the processor is 285x12 = 3.42 GHz. Yours is 1:1 so your processor works slower.
That is the whole point of this thread, synchronicity gives you 2-6 MHz power but since asynchonuous setup gives you more room for the processor, you can go higher. Of course if you have DDR 4000 or DDR 4200 you should stick to 1:1.

 

[Technonut]

Originally posted by: stevejst

Then how are you pulling unbuffered Sandra scores of 4102/4063???

It is FSB, running at 5:4, so the processor is 285x12 = 3.42 GHz. Yours is 1:1 so your processor works slower.
That is the whole point of this thread, synchronicity gives you 2-6 MHz power but since asynchonuous setup gives you more room for the processor, you can go higher, unless you have DDR 4000 or DDR 4200.

I understand this perfectly, but I was under the impression that you were running a FSB of 225 and 228....

Another one, using 5:4 at 225.6x2 memory, unbuffered 4088/3983, you cannot touch that!
And another one, 5:4 at 228x2 memory, unbuffered 4102/4063

 

[WaTaGuMp]

Ok lets see then if I can run my 2.8C at 235 FSB 5:4 2.5 3 3 6 and get Sandra scores of 5032 5017

Or 220 FSB 1:1 2.5 3 3 6 and get 5074 5053 the top would be better?

At 235 it is async and below is sync so this whole sync thing confuses me.

 

[stevejst]

Here is the explanation:
Assume you are happy with 230:230 scores (1:1 for FSB:Memory). I am telling you that synchronicity is worth for 2-6 MHz. Meaning add say 5 MHz and you have 235:235. Now run this asyncronuous, meaning add another 40 to FSB, that 275, and subtract 40 from memory, that is 195.
I am telling you that 275:195 will be the same as 230:230 on the performance level.
Since it is easy to find memory that runs 195 and easy to find good motherboard that goes 275 FSB (Abit IS7 for example) it is much cheaper to run 275:195 than 230:230.

I understand this perfectly, but I was under the impression that you were running a FSB of 225 and 228....

That was just memory, FSB was at 282 and 285.

 

[stevejst]

I keep on repeating this: there is a certain loss due to asynchronicity, but that loss is not big.
If you have DDR 4000 or 4200, by all means run it 1:1, that is the best you can do. But if you do not have that, say you have DDR 3200 or 3500, don't use 1:1. You have to use 5:4, because that will give you the best performance. If you only have DDR 2700, then use 3:2, and you can still have excellent performance due to increase in processor speed.

If you used AMD processor before free yourself of thinking that you must run 1:1, because Pentium is way more potent and because of amazing processor speed, it can deal with the fast memory, or even with the memory that is not as fast. You will still have a bandwidth AMD users can only dream about.

 

[THUGSROOK]

Technonut ~ give it up. these guys dont have a clue as to what they are talking about. they dont even know what sandra unbuffered is

the only reason i chimed in was for Jeff7181, cause hes understands this stuff.

 

[stevejst]

No, thugsrock, I posted unbuffered scores that are beating your unbuffered scores handily. You don't have to be a genious to churn that test in Sandra. In fact my buffered scores are more close to 6000/6000, you cannot touch that either.

In fact buffered scores are more close to reality of PC since while there will be no perfect buffering but it will still happen. So I posted unbuffered score only because you asked. As far as I am concerned I would count in only buffered scores.

 

[THUGSROOK]

do you even know how to run an unbuffered test?

 

[stevejst]

do you even know how to run an unbuffered test?

What a crap talk! You only have to uncheck that option in Sandra. How hard is that? Do you think only you know that???

Besides, update that Northwood B to C and you'll see with your own eyes.

 

[THUGSROOK]

thats what i thought

you have uncheck 9 options!

link to pic

there ya go

 

[stevejst]

No, you have to uncheck only one. I am using HT processor unlike you, and I am using Sandra Max 3. Update yourself, you are outdated.

 

[DAPUNISHER]

Originally posted by: THUGSROOK
thats what i thought

you have uncheck 9 options!

link to pic

there ya go

LMAO! owned! Thugsrook smacks down another troll :beer:

 

[stevejst]

What a bunch of morons. Why would anybody want to disable SSE/SSE2 instruction bench???

AMD idiot dapunisher in help to another idiot proving what? Neither one of you even have a CPU you are so eager to debate about.

 

[DAPUNISHER]

Dance for us fool!

 

[stevejst]

Well, your primate icon says the best how smart you are.
Didn't know you have a "Northwood project". What a loser you are?

 

[DAPUNISHER]

Take a vote wether the forum members here put stock in what Thugsrook posts or what you post, and see what the results are troll

 

[TerryMathews]

Originally posted by: THUGSROOK
actually,
those latencies are user adjustable on springdale/canterwood motherboards.
but like i said, you lose that ability when running async.

thats right ~ you can actually OC the northbridge latencies! :evil:

HTH

Actually , I was referring to actual latency and not 'latency' which is a setting of the memory controller. Any time you've got seperate clocks, you've got to wait for the recieving clock to go into it's open state, whether that's high or low. This is not easy to explain without a clock graph...

---\ /-----
\ /
\---/
CPU clock
--\ /--
\ /
\-/
memory clock

EDIT: I don't know how to make FuseTalk not delete spaces...

CPU would have to wait for memory and vice versa. A lot of this is negated by fairly large memory buffers in memory controllers these days, but the latency inherent in waiting for the clock signal to change states cannot be overcome. (Without clockless tech, which is still many years away from usefulness)

 

[stevejst]

@dapunisher
No need. I can see what a loser you are just by looking your "constantly evolving LAN."
About benchmarks and overclocking knowledge you apparently know only by "theory."

 

[THUGSROOK]

hey dont blame me for sandra unbuffered, i didnt invent the POS.

you will find that on forums talk means nothing ~ post your scores and shutitup.

youre type of attitude is why many ppl no longer bother to post information here.

 

[DAPUNISHER]

Ad Hominen attacks won't win you any credibility in these forums troll Now, you've danced enough, eat the scraps we thrown you fool!

 

[stevejst]

Ad Hominen attacks won't win you any credibility in these forums troll

Apparently it holds you in a cage though I guess you did not get your banana today. Cool down gorilla.

 

[Shimmishim]

oh gosh i read this whole thread...

funny stuff i tell ya funny stuff...


i dunno.. at least in my opinion... i'd rather run 3.5 ghz asynch 292/234 than run 250/250 synch at 3.0 ghz

3.5 looks more impressive than 3.0

but in terms of benching stuff... who cares!!!

it's all about the real life performance such as games and video encoding...

i say use the settings that make you happiest...

Thugs - SCREW YOU AND YOUR STINKIN' 2.4B processor... (in the nicest way possible)