MrIcee breaks the 500 FSB mark!!

[OPPAINTER]

Looks like our friend MrIcee is back in action with his new DFI 939 SLI mobo and has maxxed it out.
Great work as usual Randi

See it HERE

OPP

[mrlobber]

Wow, nice round number all over these screenshots there ! Awesome work.

[Bennah]

Great work by MrIcee.... only thing I noticed wrong is that he spelt GSkill wrong in the second screenie

[krampak]

He is not the first hehehe

http://www.xtremesystems.org/forums/...hlight=dfi+500

overcrash86 did 520mhz.

[MrIcee]

Thanks Eric

Just an update, I've been Priming this puppy since 8:30pm last night at 500HTT/250Memory with OCZ PLatLE PC3500 BH-5 installed at 2-2-2-8 1T. At the 16 hour mark of Prime I ran a 16M SuperPi successfully. At the 16 hours 45 minute mark of Prime I ran 32M SuperPi successfully and Prime is still chugging along.

I'm taking screenshots along the way, I'm not sure what else to throw at it, it seems this board and Winchester love it at 500 Mhz HTT.

My only wish is that the Clock Generator could be pushed above 500Mhz as I'd love to find out how high this baby would go and sustain it.

Randi

EDIT
Just found the latest clockgen, let the fun begin. I was using the version that shipped with the board which only allows up to 500Mhz
Yep, and I did slip and misspell Gskill

[Evil_Spork]

that looks fun...... now is it actually the clockgen that cant go higher? or just a bios/program limit?

do you mean the physical clockgen on the mobo or the program? :p

[Bennah]

Thanks Eric

Just an update, I've been Priming this puppy since 8:30pm last night at 500HTT/250Memory with OCZ PLatLE PC3500 BH-5 installed at 2-2-2-8 1T. At the 16 hour mark of Prime I ran a 16M SuperPi successfully. At the 16 hours 45 minute mark of Prime I ran 32M SuperPi successfully and Prime is still chugging along.

I'm taking screenshots along the way, I'm not sure what else to throw at it, it seems this board and Winchester love it at 500 Mhz HTT.

My only wish is that the Clock Generator could be pushed above 500Mhz as I'd love to find out how high this baby would go and sustain it.

Randi

EDIT
Just found the latest clockgen, let the fun begin. I was using the version that shipped with the board which only allows up to 500Mhz
Yep, and I did slip and misspell Gskill

Gets even better Prime stable and super pi 32m stable... even more congrats to you

[bachus_anonym]

Nice numbers there, Randi

But can you honestly tell what is the point of all that "HTT chasing" ? Besides the fact that 500HTT looks nice, is there any other benefit to running it this high?

Aren't 416x6=2500Mhz DDR500 6-3-3-2.5 (divider120)
or
357x7=2500MHz DDR500 6-3-3-2.5 (divider150) same ??? I think they might be

But anyway, good luck on getting it above 500HTT :thumbsup:

[comment]

Yes, that is seriously and sweetly so insane

[neokenzo]

Thats some insane speed, congratulations dude

[mcnbns]

I am loving that screenage, even though it isn't the first 500MHz HTT I've seen. w00t!

On a personal note, I need to upgrade very soon.

[MrIcee]

Nice numbers there, Randi

But can you honestly tell what is the point of all that "HTT chasing" ? Besides the fact that 500HTT looks nice, is there any other benefit to running it this high?

Aren't 416x6=2500Mhz DDR500 6-3-3-2.5 (divider120)
or
357x7=2500MHz DDR500 6-3-3-2.5 (divider150) same ??? I think they might be

But anyway, good luck on getting it above 500HTT :thumbsup:

Let me tell ya a little story

Back in the day , before we created the first 3DMark Team and the 3D benching craze took off, I got hooked up with OPP and DDTUNG who were purveyers of trying to attain the highest FSB possible, prove it stable thru Prime95 and run Bandwidth benches, and developing bandwidth enhancing tweaks along the way.

This is way back at Icrontic. We were the first over 170, 180, and Eric, DDTUNG and I were the first to eclipse 200 on the AMD platform, and by that point we used that to enhance our 3D benching.

This is not purely about the bandwidth produced, because of course it can be ascertained that at the same CPU Mhz, a BH-5 270Mhz 2-2-2-5 1T will outperform 325Mhz 2.5-3-3-7. Not always tho, depends on the bench.And at 250Mhz the bandwidth is certainly not incredible, but used as an illustration of stability among many benches.

Although I was unaware of some of these guys going over 500HTT, I wanted to illustrate the stability of the achievement, which I still have much to post with Prime, SuperPi etc etc. But the fact that THIS is a plateau that few have achieved, and even fewer have achieved in stability tests.

FSB overclocking is just like total CPU Mhz overclocking, or trying to attain the highest bandwidth(at highest 1-1 FSB/Mem overclocks) such as you do and run stable 16M and 32M Super Pi's.

Its an achievement and goal for all who can accomplish it in whatever their niche is.

Certainly if others had achieved this from this forum, it would not have been asked what significance it has except "HTT chasing".

Thank you for the kind words otherwise my friend

Randi

[EclectiK]

Let me tell ya a little story

Back in the day , before we created the first 3DMark Team and the 3D benching craze took off, I got hooked up with OPP and DDTUNG who were purveyers of trying to attain the highest FSB possible, prove it stable thru Prime95 and run Bandwidth benches, and developing bandwidth enhancing tweaks along the way.

This is way back at Icrontic. We were the first over 170, 180, and Eric, DDTUNG and I were the first to eclipse 200 on the AMD platform, and by that point we used that to enhance our 3D benching.

This is not purely about the bandwidth produced, because of course it can be ascertained that at the same CPU Mhz, a BH-5 270Mhz 2-2-2-5 1T will outperform 325Mhz 2.5-3-3-7. Not always tho, depends on the bench.And at 250Mhz the bandwidth is certainly not incredible, but used as an illustration of stability among many benches.

Although I was unaware of some of these guys going over 500HTT, I wanted to illustrate the stability of the achievement, which I still have much to post with Prime, SuperPi etc etc. But the fact that THIS is a plateau that few have achieved, and even fewer have achieved in stability tests.

FSB overclocking is just like total CPU Mhz overclocking, or trying to attain the highest bandwidth(at highest 1-1 FSB/Mem overclocks) such as you do and run stable 16M and 32M Super Pi's.

Its an achievement and goal for all who can accomplish it in whatever their niche is.

Certainly if others had achieved this from this forum, it would not have been asked what significance it has except "HTT chasing".

Thank you for the kind words otherwise my friend

Randi

The new clockgen allow up to 600 mhz

I do 500 too, but not stable, I'm looking for settings to be so.
Actually my max benchable is 497.

And max stable is 485, maybe more : http://valid.x86-secret.com/show_oc.php?id=5566

ECK

[bachus_anonym]

@MrIcee,

I want you to understand that I really appreciate and can't stress that any more, how important part in "benchmaking history" you play...

Although I was unaware of some of these guys going over 500HTT, I wanted to illustrate the stability of the achievement, which I still have much to post with Prime, SuperPi etc etc. But the fact that THIS is a plateau that few have achieved, and even fewer have achieved in stability tests.

Don't you think that maybe not many people are interested in achieving super-high HTT because they might know that it just doesn't benefit the system as far as overall performance compared to lower HTT but same mem speed?
To be honest with you, I haven't tried pushing HTT above what my mems could do and don't know how hard it is to get to 500HTT... I've always been very sceptical about that as well as posting record braking Sandra Mem BW benchmarks. That just didn't make enough sense to me as 3DMark runs or SuperPi did. But maybe I should just try and see what is the hole hype about

Its an achievement and goal for all who can accomplish it in whatever their niche is.

I understand that, sir...

Certainly if others had achieved this from this forum, it would not have been asked what significance it has except "HTT chasing".

I'm not exaclt sure if I understand what you mean... But I've not singled you out for that question of significance. It's just if I had to rate certain benchmarking records I would put 3D as first, then SuperPi (although, I would love to see sth more than 1M), then probably HTT and suicide CPU speed shots. But that's just me... Not sure If anyone else shares that point of view.

[Paxi]

real nice !

[Major]

congrats Randi !!

[cupra]

nice but in the athlon 64 platform its not a big difference because its stay in the 2g fsb, so if its in the 500mhz tha htt multy set lower.

but still nice bm

[OPPAINTER]

bachus_anonym

Its all about the Thrill of Victory.

It's like when you see high CPU screenys that are just sreen shots and nothing more. A lot of viewers dig it. Although in MrIcees case he has made this benchable.

There's nothing at all wrong with showing off 500 HTT benchable.

OPP

[MrIcee]

Thanks OPP

I guess you hit it on the head, benchable. That has always been the keystone to an achievement. What the heck is a screenshot if its not benchable, especially a bios screenshot

Back in the day screenshots were laughed at if you could not back it up with stability tests or at least benches. Thats why today the hobby can sometimes be a farse, which is a shame. Its nothing like it used to be back in the Icrontic days(that time period), when enthuisiasts demanded results for an overclock or FSB achievement to be legitimate.

Same thing with 3DMark benching, it used to be if you could not run the ENTIRE bench you were not taken seriously, because the measuring stick was the ability of your system to remain stable thru the entire benchmark. That has also majorly gone by the wayside.

Randi

[bachus_anonym]

Thanks OPP

I guess you hit it on the head, benchable. That has always been the keystone to an achievement. What the heck is a screenshot if its not benchable, especially a bios screenshot

Back in the day screenshots were laughed at if you could not back it up with stability tests or at least benches. Thats why today the hobby can sometimes be a farse, which is a shame. Its nothing like it used to be back in the Icrontic days(that time period), when enthuisiasts demanded results for an overclock or FSB achievement to be legitimate.

Same thing with 3DMark benching, it used to be if you could not run the ENTIRE bench you were not taken seriously, because the measuring stick was the ability of your system to remain stable thru the entire benchmark. That has also majorly gone by the wayside.

Randi

That's exactly what I'm trying to say...
I'm literally tired of all those suicide screenshots or just Sandra screenies... But it looks that not that many people still care about that. I just keep hearing analogies to drag-races and stuff like that

Some time ago I tried to show a difference...
I posted full set of benchmarks with my Winnie / BH-5 @ 276x10 6-2-2-2.0-1T. I would soooo like to see more stuff like that, seriously...

Just making sure... Randi --- respect!

[computersmsa]

Nice, but the WR is 543 MHz by Jeffchen

[IronAge]

Well thats an american world record then. :banana4:

[flutie98]

[cabojoe]

Ahhhh...The Icrontic days....With OPP as our fearless leader, DD as his trusty caddy, Mr. Icee (was and still is) the voice of reason, Joe trying to keep up but
making too many blue arcs with power supplies, and mbrock chasing my tail. Those were the days**sniff sniff**. Congrats Randi!

[OPPAINTER]

Ahhhh...The Icrontic days....With OPP as our fearless leader, DD as his trusty caddy, Mr. Icee (was and still is) the voice of reason, Joe trying to keep up but
making too many blue arcs with power supplies, and mbrock chasing my tail. Those were the days**sniff sniff**. Congrats Randi!

Good to see you around Joe

Yep, those were some fine times back then.

OPP