just got SL79L Prescott 3.0e 1mb HT

Finally, i am in the hyperthreading world!!! Just got a Prescott 3.0e SL79L stepping cpu retail for $269! WOOT, can't wait to bench this baby.

using the 1016 Asus P4P800 bios!! it recognizes prescott cpu perfectly. although i had huge problems in the beginning, stupid windows xp installed drivers for hyperthreading and it conflicted with LAN controller, and my mouse froze. fixed that, then i had to load cpu bios to default values because it wouldn't boot into windows. finally, ...........here is a screenshot of my new baby!

wtf? Drivers for hyperthreading? Did I Miss something o_O?


Anyway have some fun :) And OC that baby please.

Different HAL

Quote:

---

Originally posted by EmineM
wtf? Drivers for hyperthreading? Did I Miss something o_O?


Anyway have some fun :) And OC that baby please.

---

windows wouldn't boot the first time when i didn't load default bios values! but finally got into windows , and windows installed several new drivers for ht and asked to restart!! no one had this problem?? my old cpu was a P4 2.8B no hyperthreading.

Hyperthreading!! 2 cpus in task manager..

http://imagehost.vendio.com/preview/.../Prescott2.JPG

overclocking?? now way jose! lol my idle cpu temp jumped 17c !

P4 2.8 B ; 34c idle

Prescott 3.0e; 47.1c idle

crap, my system isn't using PAT mode anymore!! Ram is running at only 159.6mhz dual channel WTF???? I have DDR333, shouldn't it run at 166mhz by default. Hmmm.

my super PI score 1mb was 47 sec with Prescott. i got 51 with the 2.8 Northwood chip.
4 sec improvement.
:toast:


oh and my Aquamark3 score:
http://arc.aquamark3.com/arc/arc_view.php?run=749916785

1k more

OMFG, look at my new PCmark04 scores!!!

new:
http://service.futuremark.com/compare?pcm04=2473467

old:
http://service.futuremark.com/compare?pcm04=2450326

isn't that good? wow, i beat alot Athlon64s running stock.

time to update your sig too

Quote:

---

Originally posted by freecableguy
time to update your sig too

---

done!:toast:

Might be able to tighten memory timing up more or not run so much in the background when running pi.

:toast:

SL7J5

lol, nice score Fugger! beat me with 200mhz less clockspeed. yeah, i'm too afraid to mess around with the mem timings atm. Wow, those LGA 775 has lower voltages, and new D0 stepping.

I'm just happy i got a SL7 stepping. although i might never utilize it to full potential lol.

Are you using DDRII with that Fugger?

Centon DDR2 samples with Micron chips, they run the tightest memory timing with ease and yet to find the top end. I have not played with loose timing much yet.

This will pull 3.3k sandra unbuffered @ 1:1 200FSB

:)

Sweet!:toast:

LGA 775 chipset looks very promising!:cool:

a lot of bandwidth.

here is my sandra 04 run. i ran it buffered i think. not sure...

i was at 3.2k-3.3k with P4 northwood 2.8B:)

The ORB does not let me publish any scores from this rig.

Some applications think this thing is a nacona xeon so dont be fooled when you see a screen shot with one missing L3 cache

I wish it was

Quote:

---

Originally posted by Kanavit
crap, my system isn't using PAT mode anymore!! Ram is running at only 159.6mhz dual channel WTF???? I have DDR333, shouldn't it run at 166mhz by default. Hmmm.

---

don't you just wish those dang intel platforms weren't so buggy? ;)

sry, just couldn't resist, you always preaching how intel is foolproof

looking good, I think that it must be just a mem divider issue

FUGGER stop making me want to buy new rams, mobo, cpu, and video all for the sake of intel (as crazy as it seems, it's still temping)

hmm, Fugger isn't the nocona cores have absent L3 cache? Or is cpu-z just not properly reading it right.

Snowwie, your probably right, my cpu/mem ration did change form 4:3 to 5:4.

They have L3 cache in various denominations. The datasheet claims it can have up to 8MB L3 cache.

The newest version of CPU ID works correctly, yours is an earlier version with better identification on socket 478.

Kanavit see if your RAM will run in 400 mode (1:1)...if not give it a bump in volts (no it won't hurt it).

That's what this site is all about. Would get you DDR400 and PAT for no extra cost (and yes I know money is tight).

thx Soul burner, i will try that. just tried PAT and turbo performance mode, no boot. my memory doesn't seem to want to go higher than 159mhz. :(

maybe your right, i should add some voltage and set to 1:1.

Quote:

---

Originally posted by Kanavit
crap, my system isn't using PAT mode anymore!! Ram is running at only 159.6mhz dual channel WTF???? I have DDR333, shouldn't it run at 166mhz by default. Hmmm.

---

No, as I tried to explain on that other thread, 160DDR or 320MHz effective is the correct setting for 333MHz/PC-2700 RAM with an 800MHz FSB Intel Processor. That means you're setup with a 5:4 Ratio. On i865 PAT is lost when not running 1:1, that's why many bought i875.

The link to the site that explains Ratios, in FAQ's sticky is off BTW. A 5:4 in this case simply means 80% as it relates to the memory controller, not the front side bus=P Yes it's four fifths but for the memory. So once you figure that out, you see that 160MHz DDR is 80% of the 200MHz FSB 5=200 so 4=160. If you overclock, the RAM is ran at 80% of whatever the FSB is when a 5:4. If the Ratio were 3:2 or two thirds, it would be 66.7% or 200:133 for PC-2100/266MHz effective.

This is an Intel tweaked setting or a Plan B made by Intel. Intel did this in case PC-3200 wasn't ready in time. The user could use PC-2700 and PC-2100 with 5:4 and 3:2 Ratios respectively. NO, it doesn't cause a hit to performance, except in cases where absolute memory bandwdith matters, rare. Just the opposite, more times than not, the performance increases. I can explain that later as well. Good luck!

Donnie

Quote:

---

Originally posted by Donnie27
No, as I tried to explain on that other thread, 160DDR or 320MHz effective is the correct setting for 333MHz/PC-2700 RAM with an 800MHz FSB Intel Processor. That means you're setup with a 5:4 Ratio. On i865 PAT is lost when not running 1:1, that's why many bought i875.

The link to the site that explains Ratios, in FAQ's sticky is off BTW. A 5:4 in this case simply means 80% as it relates to the memory controller, not the front side bus=P Yes it's four fifths but for the memory. So once you figure that out, you see that 160MHz DDR is 80% of the 200MHz FSB 5=200 so 4=160. If you overclock, the RAM is ran at 80% of whatever the FSB is when a 5:4. If the Ratio were 3:2 or two thirds, it would be 66.7% or 200:133 for PC-2100/266MHz effective.

This is an Intel tweaked setting or a Plan B made by Intel. Intel did this in case PC-3200 wasn't ready in time. The user could use PC-2700 and PC-2100 with 5:4 and 3:2 Ratios respectively. NO, it doesn't cause a hit to performance, except in cases where absolute memory bandwdith matters, rare. Just the opposite, more times than not, the performance increases. I can explain that later as well. Good luck!

Donnie

---

good read, thx Donnie27. very useful info there. I may be getting DDR400 pretty soon. I'll keep everyone posted.

Quote:

---

Originally posted by Donnie27
This is an Intel tweaked setting or a Plan B made by Intel. Intel did this in case PC-3200 wasn't ready in time. The user could use PC-2700 and PC-2100 with 5:4 and 3:2 Ratios respectively. NO, it doesn't cause a hit to performance, except in cases where absolute memory bandwdith matters, rare. Just the opposite, more times than not, the performance increases.

---

Dropping to 5:4 after being at 1:1 can cost you as much as a 250mhz clock speed drop in 3D apps (shown by macci...notice how he "ALWAYS" runs 1:1? There is a reason).