Pifast Benchmark

Printable View

Show 100 post(s) from this thread on one page

10-15-2004, 02:42 AM
blinky

its an error: if u compare the score side by side, ull see that his series computing time is FAR below all the others, but his Division time, InvSqrt time, and final huge multiplication time are high than supposedly low clocks, and higher pifast scores. that does not make sense

look Pedro Rocha's:
http://ricky.home.sapo.pt/fx53/pifast3550.JPG

and yuri's
http://24.158.154.102/images/31.97.jpg

so yuri series computing time must have a bug in it, and that affected his scores by making them go waaay down, even with higher calculation and division, and invSqrt times. BUG!!!
10-15-2004, 02:50 AM
Crownie

Nice compare Blinky :) .

Defiantly a bug.
10-15-2004, 02:51 AM
RocKer

New pifast ,but cpu or mobo is dead now,this was with vdimm at 3.5v.
10-15-2004, 10:23 AM
PMM

Can I ask one thing here and that is feel free to question results but
please do not accuse people of doing the impossible!!!

As someone myself heavy into Voltmods and 1st person to Crack open some specials
to the overclocking community please do not think there are not ways around Over/under Volt & current protection Tom i'm sure like myself will withhold information if it gains him an advantage, i've done it myself and for anyone who has a decent
understanding of electronics they could do away with any on-board Voltage control anyway.

If I think something is a miss i'll either withhold a result pending a request to the user
for additional info, or i'll put up a result if its borderline to my checks in the view that
if I think its duff at a later date it will be removed.

A submitted result for a A64 3800mhz is not going to go up unless proved as its way
beyond the norm and alarm bells ring it's then down to the user to prove it and pass
back information that 100% meets my expectations.

There are alot of people who attempt to slip something in, since its not real time
results its a wasted effort and submitting a time of 1sec ain't ever going to happen
but funny enough there mainly people with Hotmail addies.

There are alot of people here I trust and have proven to me over time a great
honesty for the information they submit, its thanks to these people that I can
weed out the fakes as there result give a trusted trend that can be followed
and compared against.

Its all a bit of fun and more fun when played fare when you know your result
is genuine, I know I have great pride in what I achieve knowing I did it and did
it honestly.

There are some really mad people out there distroying there computers to gain
what many regard impossible, its not impossible its just not having the knowledge
to know how to over come, however there will always be someone who finds out
how.... but also might wan't to keep it to themselves for a bit.

So my moral is ask for evidence but never accuse a liar always trips up and for me
if it happens they have every result submitted stripped and will be named and shamed ;)
10-15-2004, 03:38 PM
SAE

PMM, that's the right way you're going :thumbsup:

I definately know what it takes to decrease the pifast time about 1second!!! It's truely incredible, there's a cpu/cooling/board doing that shown time actually.

He should prove it with a lot more benches... at least superpi should be cake as pifast is way more cpu dependant :D
10-15-2004, 04:52 PM
hipro5

Relax guys........None called anynone a lier........We all "play" in here.......One plays fair.....other not......not a big deal.......Me for example was in the first place of Pi-Fast a couple of moons........Now I'm 15 - 20 - I don't remember....So what?.....Life is going on..... :D
10-15-2004, 08:49 PM
blinky

no one called anyone a liar,

just sayin its so fantastic i want more specs
10-15-2004, 10:12 PM
CodeRed

Quote:

Originally Posted by blinky

no one called anyone a liar,

just sayin its so fantastic i want more specs

It might be a good idea for people to post a CPU-Z shot (showing the codes at the bottom left) together with their pifast screenie so that overclocks can be verified ;)

IMO this should be mandatory for all top 20 players.
10-15-2004, 10:21 PM
blinky

Quote:

Originally Posted by CodeRed

It might be a good idea for people to post a CPU-Z shot (showing the codes at the bottom left) together with their pifast screenie so that overclocks can be verified ;)

IMO this should be mandatory for all top 20 players.

i concur, and he hasnt posted anything besides that one screenie

SPEAK YURI!
10-16-2004, 12:59 AM
PMM

For info i've not had that result from Yuri Submitted ;)

I Always side on belief that some can be true, I did get caught out once
as I had a result from a German person that I just deleted this was a 3500mhz
speed some 300mhz+ higher than the lads here at the time, I realise now that
the result was genuine but since I deleted it I can't put it up I never contacted
the person because I dis-believed it.

So its is very much down to giving all the proof you can if your up there in the
dizzy heights of the Top20 ;)
10-16-2004, 09:25 AM
SAE

Quote:

Originally Posted by PMM

For info i've not had that result from Yuri Submitted ;)

I Always side on belief that some can be true, I did get caught out once
as I had a result from a German person that I just deleted this was a 3500mhz
speed some 300mhz+ higher than the lads here at the time, I realise now that
the result was genuine but since I deleted it I can't put it up I never contacted
the person because I dis-believed it.

So its is very much down to giving all the proof you can if your up there in the
dizzy heights of the Top20 ;)

Yep. True spoken. I would never post such a bench without verification through cpuz at least :)

So Yuri, please don't let us wait. Or maybe you just wanna yell: JJJOOOOKE!!! :lol: ;)
10-17-2004, 12:25 AM
LardArse

Anyone with LGA prescott tried running this program? Its very weird it gives an error before running the program on my 3.6E with Abit AS8.
Edit: It works but it sux on the prescotts :p
10-17-2004, 03:15 AM
hipro5

Quote:

Originally Posted by LardArse

Anyone with LGA prescott tried running this program? Its very weird it gives an error before running the program on my 3.6E with Abit AS8.
Edit: It works but it sux on the prescotts :p

Yes indeed........S-Sandra Arithmetic and Pi-Fast - any version - sucks as for the Prescott CPU's......There is "something" missing as for these benchies I think...... ;)
10-17-2004, 04:27 AM
LardArse

Hey Hipro whats the missing link?? :D
27s superpi but 41.39s pifast with P4EE
25s superpi but 42s pifast with prescott
:p:
10-17-2004, 06:32 AM
hipro5

Quote:

Originally Posted by LardArse

Hey Hipro whats the missing link?? :D
27s superpi but 41.39s pifast with P4EE
25s superpi but 42s pifast with prescott
:p:

Yes I've noticed that too.....No go for Pi-Fast as for Prescott CPU's......Pi-Fast maybe an old version BUT newer versions does the same thing.......I think that it must be some new instractions of the Prescott that the Pi-Fast progie isn't capable to "follow".........who knows...... :D
10-17-2004, 03:12 PM
Epsilon

My score:

http://members.home.nl/epsilon/temp/...293_3700EB.jpg

Specs:
A64 3400+ Newcastle @ 10x293 (1.845v)
Prom Mach1@ r507
Lanparty NF250 UT
OCZ PC3700EB @ 293mhz 1:1 (3.5v :D )
10-17-2004, 04:00 PM
yuri

Ok i have finally found the max of this board (Asus A8V), here is a 14x303 (FX53@ 4242Mhz) 2/3/2/5/11/16 4.2 vdimm / 2.3vcore

http://24.158.154.102/images/26.84.jpg
10-17-2004, 04:08 PM
cpulloverclock

I dont believe u
sorry...

I think u are using pifast 4.3 :stick:

try superpi 1M please at 4200MHz --> 15-20s
10-17-2004, 04:34 PM
SAE

Quote:

Originally Posted by cpulloverclock

I dont believe u
sorry...

I think u are using pifast 4.3 :stick:

try superpi 1M please at 4200MHz --> 15-20s

:lol:

Me neither :D

That's too unrealistic... no comments on cooling, no cpuz, no other benches. :(
10-17-2004, 04:35 PM
aoc007

Quote:

Originally Posted by yuri

Ok i have finally found the max of this board (Asus A8V), here is a 14x303 (FX53@ 4242Mhz) 2/3/2/5/11/16 4.2 vdimm / 2.3vcore

http://24.158.154.102/images/26.84.jpg

:eek: Whats the cooling? Got a Superpi? Also any pictures of the setup and more evidence, thats way faster than anyone before!
10-17-2004, 10:34 PM
Zeus

Yeah right!
Why do you use a danish version of PiFast to start with?
Are you danish?
2.3Vcore and 4.2Vdimm? uh huh :stick:
Tell us another one...
10-17-2004, 11:29 PM
CodeRed

Quote:

Originally Posted by yuri

Ok i have finally found the max of this board (Asus A8V), here is a 14x303 (FX53@ 4242Mhz) 2/3/2/5/11/16 4.2 vdimm / 2.3vcore

Oh come on Yuri, surely you can do better than that ... 5GHz must be easy with photochop :p:
10-17-2004, 11:40 PM
FDM

under 50sec now:
http://www.finaldrivemagazine.com/andy/4944.jpg
10-18-2004, 12:18 AM
SAE

Quote:

Originally Posted by FDM

under 50sec now:
http://www.finaldrivemagazine.com/andy/4944.jpg

That's what I call realistic and honest :thumbsup:

:D ;)
10-18-2004, 12:43 AM
FDM

Quote:

Originally Posted by SAE

That's what I call realistic and honest :thumbsup:

:D ;)

of course! All my scores are slow enough to be believed even if i was making them up :p:

Show 100 post(s) from this thread on one page