What to do...

M

Matt_Hirst

New member
In my main work/office/home rig I have a Opty 170 modestly clocked to 2500, in the spare I have a winchester clocked to 2500.

However having read the thread about benchmarks needed am feeling the need to build a benching rig, to that end I have been offered a FX57 at £380.

Now bearing in mind Conroe is about also am2, is the FX worth getting and building a rig round or should I do abit of juggling drop winny into my ofice rig, the opty into my spare/benching rig to see what that can do.

Or should I go the FX57 route for the benching rig - but it'll mean selling either the Winny or the opty.

Any thoughts?

regards,

Matt

P.S. What benching ram would you reccomend, preferably 2gb?
 
Tcasemax on 3200 is

tcasemax_64_3200.jpg


Will do the opty tonight.

regards,

Matt
 
Hi there can you do this again as it is not showing your TDP? but as for the Tcase this chip should do well if cooled nicely
 
name='sacha35' said:
Hi there can you do this again as it is not showing your TDP? but as for the Tcase this chip should do well if cooled nicely
Click to expand...

Sacha,

Even downloading and running ver 1.19 it doesn't show TDP, does my chip support this/have this function?

regards,

Matt
 
I never understand why people bother using software to decide if the chip will OC well. Put it into a mobo, overclock it and see!! Use the one that works best in the bench rig and the other in everyday use while waiting for Conroe hehe :D

Sacha your trust in this software worries me just slightly. You seem to go off what it tells you is a good chip or not, but software will only know that because it's being told by another bit of software which is being told by another bit, and thats being told by the CPU. There is SO MUCH space for error it's silly.

Overclock the chips, see which is best, use that for benching. Simple, dnt bother with this tCaseMax bolox.

Boardy
 
Not really mate the Tcase and TDP have been set in the chip by AMD of which this is how chips are graded, hence Higher Tcase and TDP higher code I.E FX60- this is a good way to give a rough idea of a chips clocking capabilities before buying or trying, look at 4Qman's chip high Tcase and TDP and he has got a very good clock out of this chip, now look at my Opty Bad Tcase and TDP and not such a good clocker.

So i think to say the software is no good is not correct because from what we have seen from these readings is showing us that this is true as for the overclocks we have seen on these chips, so then we can look at the stepping codes and say this should be a good chip to buy.
 
name='boardy' said:
Overclock the chips, see which is best, use that for benching. Simple, dnt bother with this tCaseMax bolox.
Click to expand...

Nice :D

I somewhat agree, the software can only tell you so much. It should be used as a guide. If you got two chips which the software reads as the same - I bet that they wouldn't clock the same ;)
 
Yes stocky it is only a guide agreed, but this is a good guide , think about it, if you see a chips with a bad Tcase and TDP, the chances are it wont clock well, then by finding out the steeping of the chips allows us all to stay away from that stepping, or if the chips has a good Tcase and TDP then we look at the steeping codes of the chip and there is a good chance we will get a good overclocker, yes i agree all chips are different you know that you can by two chips with the same steeping codes on and one might clock just that bit better but they are normally within a few MHZ of each other.
 
In my experience and personal opinion, there is nothing more proving than overclocking the CPU itself. I have never used software to tell me how well a CPU will OC, thats just not possible no matter how much anyone tries to say it is. I proved this with my Barton Mobile 2500+ hitting 3.5Ghz when everyone told me it wouldn't get past 2.7Ghz. It took me a hell of a long time getting there and it wasn't stable but I proved every1 wrong. It's the only way of deciding, NEVER go by what software says, as it is only as good as the dev who wrote it!

Personally I think that an FX57 now is not worth buying unless it is Week 16. even then you still are not guarenteed a nice chip. I personally would juggle what you have and wait out Conroe or AM2, save for those as you can bet they won't be the cheapest things around.

I bought a Presler 930 yesterday, bottom of the 9 series and cheapest dual core with 800Mhz bus for the exact reason, I want Conroe, so I'm saving now.

Cheers

Boardy
 
I am not disagreeing with you mate but it is a good reference point to start with, we don't know how a chip will clock until we try, but like said if we can compair this software to chip steeping then this gives us a good starting point.
 
Matt_Hirst said:
Boardy,

So any thoughts on wether to go the FX route or not?

regards,

Mat
Click to expand...

£380 sounds like alot of money for an FX-57, I'd be much more inclined to buy a decent stepping Opteron or a nice clocking 3800x for about half that.
 
Watch out for the coldbugs on Optys tho. I think it was CABNE stepping without the bug and if u can find 1 of them get it under the ice, or just keep wat u have and wait as I say :)

The choice is yours at the end of the day :)

Boardy
 
I was not disputing what boardy has said as the only way we know how much we can get out of a chip is overclocking it, but like i said right from the start this software shows us the rated Tcase and TDP set by AMD, so when we see high settings in the TDP and Tcase we know that it could be a good clocker and when we see low settings in the Tcase and TDP we will stay away from these chips, we can then make a list of all the steepings of the good and bad chips to buy.
 
Would say stepping is a better indicator of how well a chips going to overclock over its TcaseMax... Also gives a much better indication as to how its going to react to cooling etc too....
 
steeping is better because this has been tried and often lodged, but does not guaranty the overclock you might get with a chip as they are all different but does give you a rough idea of how well it might clock, but when you do not have the steeping then the Tcase and TDP is a good example because this has shown when the TCASE and TDP are high these chips normally clock well but as said some chips do clock better than others.
 
boardy said:
I never understand why people bother using software to decide if the chip will OC well.

Overclock the chips, see which is best, use that for benching. Simple, dnt bother with this tCaseMax bolox.

Boardy
Click to expand...

Well consistency is a wonderful thing :D

Coz for me the two Fx60's I put on it showed exact same TCaseMax figures and overclocked the same also.

The question you have to ask yourself is can you discount it? Would you buy a wk16 FX57 knowing that it's TCaseMax is a lot lower than Fatties?

I will be asking people for these values in future when purchasing a CPU that I will use for benching - thats for sure :)
 
Yes mate your right, if we can put the Tcase to the stepping of the chips then that will give us the best info on what should clock best.
 
You must log in or register to reply here.

Similar threads

R
Backup Rig
Replies
0
Views
462
robbiec
R
Excalabur50
Need to pick peoples brains please
Replies
5
Views
479
Excalabur50
Excalabur50
P
Do some people never learn from the past?
Replies
2
Views
771
Pugsport
P
g0ggles1994
Kitchen Flooring
Replies
8
Views
388
g0ggles1994
g0ggles1994
C
Advise on my proposed build please
Replies
4
Views
458
AlienALX
AlienALX
Back
Top