alpha's Allendale E4300

Just a very brief torture session to ascertain how well the new E4300 core clocks

Hardware is as per my sig

It looks like 3.60GHz will be beyond WC unless a TEC is used though my WC rig has a NB plus GPU in the loop (plus it needs a good clean :o ) with ambient @ 34C today
3.30 to 3.40Ghz should be OK for most decent high performance WC though I would not recommend running @ over 3.30GHz due to the extra .05Vcore needed for only another 100MHz

I am reasonably sure that CoreTemp is NOT reading temps correctly as they do not correspond with Speedfan that I have adjusted to match coretemp when using my e6600 (using +4C offset) and I am positive that the CPU rebooted when running @ 3.60GHz due to exceeding its operating temp

Heres the first run @ 3.60GHz, very unstable but managed to do an 1M SuperPi run
Vcore of 1.60V was needed (set in bios, reading 1.560V with 1.57V @ idle reported by speedfan) to boot into windoz though it could boot into bios @ 3.60GHz using only 1.50Vcore

3.60GHz
http://i16.photobucket.com/albums/b29/alpha0ne/E433.jpg

This is @ 3.40GHz, still not stable though with good cooling it could possibly be OK, I set 1.550Vcore in bios that read as 1.512 in bios and 1.52V idle in windoz

3.40GHz
http://i16.photobucket.com/albums/b2...0ne/E433-1.jpg

Heres 3.30GHz using 1.50Vcore set in bios that reads as 1.456V in bios and 1.47V idle in windoz
This seems to be under the critical cutoff temp and seemed stable with my very limited testing

3.30GHz
http://i16.photobucket.com/albums/b2...0ne/E433-2.jpg

how high can u go with stock volt ?

Why do these Allendales seem to be crappy overclockers?

Quote:

---

Originally Posted by SaFrOuT

how high can u go with stock volt ?

---

+1 :wave:

Run Intel TAT program and compare between the two programs.
I'm not even sure that Speedfan has Core Temp detection for Conroes

Quote:

---

Originally Posted by Omastar

Why do these Allendales seem to be crappy overclockers?

---

1.6ghz crappy?
Its low budget chips which exceed anything overklocked.

For a new budgetrig those will be truly nice to have.
:woot:

there is a reason these are the cheapest, slowest clocked C2Ds - ever heard of binning?

these things will not be the second coming, they will be a nice chip for a mild overclock on grandma's rig and extreme clockers who use them as disposable chips...

haven't seen any evidence anywhere that they do much over 3.4+/- without way more voltage than their 6xxx bretheren.

I see you posted in my thread about TAT / CoreTemp, but I really do believe TAT is way off this time. As Eva said, the tool was never intended to be used by the public and it was for mobile chips. And please explain to me this screenshot if TAT is correct....

http://www.fileden.com/files/2006/6/20/80195/temp3.JPG

TAT is obviously NOT correct because I am not getting any throttling even when TAT reads over 95C, whereas CoreTemp is below the 85C threshhold. The reason why your system shut down at 3.6GHz was not because of your temperatures, but because the chip wants more voltage. I can cause the same rebooting on my chip if I give it too little volts on any speed.

My e4300 seems to be a litltle better, but yes, I have noticed a similar trend in needing higher voltages. To get 3.4GHz stable, I need 1.4875v in BIOS, which comes out to about 1.45v real. I am currently running Orthos(5 hours and going) at 3.5GHz with 1.49v real. Coming from a Yonah system, I'm pretty happy with my results

You also have to remember that this is the very first week/batch of the E4300, which is the lowest binned chip. Everyone seems to expect 'B' chip like overclocks, which isn't too reasonable imo. Who knows, maybe in a few weeks a magic week/batch will pop up, just like the "B" chips did.

Quote:

---

Originally Posted by Omastar

Why do these Allendales seem to be crappy overclockers?

---

How is exceeding the x6800 a "crappy" overclock? Honestly what do you expect for less than $200? Most CPU's can't do anything near 100% increase.

Quote:

---

Originally Posted by falqon

How is exceeding the x6800 a "crappy" overclock? Honestly what do you expect for less than $200? Most CPU's can't do anything near 100% increase.

---

yeah... and how does the performance scale out? That still sucks for a 9x multiplier CPU. I'd expect around 3.4ghz, and I can get it from an E6300, plus, you get better performance than an E4300 at the same clocks. Plus, an E4300 may clock 200 or 400 mhz more, but that E6300 will still match the performance.

Quote:

---

Originally Posted by StealthyFish

yeah... and how does the performance scale out? That still sucks for a 9x multiplier CPU. I'd expect around 3.4ghz, and I can get it from an E6300, plus, you get better performance than an E4300 at the same clocks. Plus, an E4300 may clock 200 or 400 mhz more, but that E6300 will still match the performance.

---

Keep in mind:
- this is a cheaper cpu
- probably slower bin
- still A revision
- with it's lower stock fsb, doesn't need so much higher-performance ram and
mobos. Look at it: at 333 fsb, you've got 3GHz and you can still run ddrII-667
ram and some worse-performing mobos.

Can you lower the multiplier and test for max FSB please?

Quote:

---

Originally Posted by StealthyFish

Plus, an E4300 may clock 200 or 400 mhz more, but that E6300 will still match the performance.

---

And where is this proof? Please tell me you're not extrapolating from stock speeds.

Quote:

---

Originally Posted by falqon

How is exceeding the x6800 a "crappy" overclock? Honestly what do you expect for less than $200? Most CPU's can't do anything near 100% increase.

---

E6300's stock speeds are only 33MHz higher, yet 3.5GHz is normal for them. They're also cheaper than the E4300's.

It looks like CoreTemp readings seem to be accurate once the high Vcore used for 3.30GHz and above is lowered to around 1.450Vcore and below for 3.20GHz ????...........interesting

I have not benched to find the highest MHz at default Vcore (1.3250V) as yet but looking at the numbers I think 3.00GHz @ 1.3250Vcore would not be too far off the mark

I benched @ 3.20GHz this morning and found the lowest possible Vcore to be 1.4125V set in bios with 1.3760V read in bios, windoz = 1.3840V @ idle and 1.360V under ORTHOS large FFT load with Speedfan and CoreTemp readings about the same

The temp reported in bios for the 3.30 to 3.60GHz runs was between 50C and 56C :eek: , now @ 3.20GHz with the Vcore lowered the reading is ~ 36C

3.20GHz
http://i16.photobucket.com/albums/b2...0ne/E433-3.jpg

I forgot to give the chip details earlier so here they are:

E4300SL9TB
Q640A***
Pack Date 21/12/06

Arrr BUGGER, I just noticed I used the wrong screenshot, I actually tortured it for around 12min :slap:

Quote:

---

Originally Posted by Gautam

E6300's stock speeds are only 33MHz higher, yet 3.5GHz is normal for them. They're also cheaper than the E4300's.

---

i found teh 4300s just slightly cheaper up here, hard decision to make. The best thing abotu athe 4300s over the 6300s is that you can use almost any kind of budget pc6400 ram and not have to worry about it being the bottle neck at all. Could save you a few bucks there

Quote:

---

Originally Posted by falqon

How is exceeding the x6800 a "crappy" overclock? Honestly what do you expect for less than $200? Most CPU's can't do anything near 100% increase.

---

Firstly, X6800s have 4MB cache, so to truly 'exceed' it, you'd have to clock an E4300 to about 3.3/3.4GHz or so, and the chip reviewed here apparently was nowhere near stable at 3.6GHz, so I'm assuming 3.3GHz would be about its max stable with reasonable voltage, maybe less. Secondly, countless people have achieved 100% overclocks on E6300s, on air no less. I was just expecting more seeing as so many people claimed E4300s would all be massive overclockers for a great price, and they're obviously binned under E6xxx parts; the only plus is the 9x multi for boards that AREN'T P5B-Deluxe. It's a good budget chip, but as usual it was overhyped by the typical herd of internet reviewers.

Quote:

---

Originally Posted by Omastar

I was just expecting more seeing as so many people claimed E4300s would all be massive overclockers for a great price, and they're obviously binned under E6xxx parts;

---

They're going to be sold for $118 in a couple months, what did you honestly expect? "great price" is a relative term. As for the binned under 6xxx, well that's just completely wrong, you don't even understand this things core. As for the 9x multi, it has a lot more to do with not buying $250 RAM than the $150 mobo.

They're going to clock amazingly.

They're going to be cheap.

Do we see a pattern here? :p:

There's no such thing as a free lunch, and if something looks too good to be true, it usually is. ;)

Quote:

---

Originally Posted by Gautam

They're going to clock amazingly.

They're going to be cheap.

Do we see a pattern here? :p:

---

Nobody said they're "going" to clock amazingly? And yes they're "going" to be cheap in the 2nd quarter of 07, how do you not understand that simple concept?

No, I see absolutely no pattern, just a strange post with no point.

OK once we/you can get over all the usual hype* (*see bull s h i t e) and take a look at the market segment and usage this chip is intended for it really does not have or should I say will not have any competition once the price settles

For a budget gamer (or a gamer on a budget :D ) this chip is what they have been waiting for :fact:

Cheap CPU, a mid to low end board that only needs to get to 350FSB, and nearly every board on the market can do that (except some of the really cheeeeep boards ;) ) coupled with the cheapest DDR2 PC5200 should leave enough left in the kitty to buy an 8800 vga card for a machine that will be the equal in games to rigs costing double to three times as much :fact: :clap:

Just because these chips wont run stably at ~ 100 to 120% overclocks :rolleyes: (well this batch anyway, who knows what the future will bring) and are not priced competitively as yet does nothing to take away the significance of this launch :banana: :toast:

This chip at the moment seems a better deal than a E6300, but once the e6300 gets a true 4 MB cache and becomes the "e6320" I'm sure I would say the decision isn't easy anymore. (I'd probably take the e6320 but then again I have the RAM to run the Mhz)

So for the time being e4300 seems like the best buy for the money (until 4mb L2 caches becomes standard on all E6XXX's )

P.S. E4300 is supposed to be $20 cheaper than E6300. So except prices to fall.

why is it some people here think these "Allendales" are conroe core? i dont get it the info is right in front of yall it is a L2 revision,not B2. Different cores and some seem to forget once the B2 conroe cored 63/64's filter out they will be L2 Allendale cores.

Quote:

---

Originally Posted by Evilsizer

why is it some people here think these "Allendales" are conroe core? i dont get it the info is right in front of yall it is a L2 revision,not B2. Different cores and some seem to forget once the B2 conroe cored 63/64's filter out they will be L2 Allendale cores.

---

Are you saying thats a good or bad thing?

Quote:

---

Originally Posted by Super strokey

Are you saying thats a good or bad thing?

---

could be a bad thing...could be good...hard to say...since the e4300 seems to be clocking around the norm for e63/64's no not a bad thing. though this could be good when intel gets the L2 "allendales" kinks worked out and we might see better clocks. just to early to say atm... im planning on a e4300 to mess with though. esp. since in gaming the extra 2mb's of cache doesnt really help.

Nice overclock. May get one of these cpus soon.
Thanks

Quote:

---

Originally Posted by Evilsizer

why is it some people here think these "Allendales" are conroe core? i dont get it the info is right in front of yall it is a L2 revision,not B2. Different cores and some seem to forget once the B2 conroe cored 63/64's filter out they will be L2 Allendale cores.

---

Yes thats what a lot of ppl are not thinking of so those who may think the present e6300/e6400 is the better core for high overclocks better buy one soon before they all become true Allendales :stick: :D

Quote:

---

Originally Posted by alpha0ne

Yes thats what a lot of ppl are not thinking of so those who may think the present e6300/e6400 is the better core for high overclocks better buy one soon before they all become true Allendales :stick: :D

---

Well I would wait until those E6300s and E6400s get 4 MB L2 caches....you can look at it that way too.

Quote:

---

Originally Posted by alpha0ne

I am reasonably sure that CoreTemp is NOT reading temps correctly as they do not correspond with Speedfan that I have adjusted to match coretemp when ...

---

Sorry, but you have no idea. Read what CoreTemp actually does! http://www.overclockers.com/articles1378/ Especially page 2: the T-Junction.

Quote:

---

Originally Posted by Unbornchild

Sorry, but you have no idea. Read what CoreTemp actually does! http://www.overclockers.com/articles1378/ Especially page 2: the T-Junction.

---

No I reckon alph is onto something, there is something different happening there

Here's the temps from mine:-
http://i135.photobucket.com/albums/q...emp-odd-wp.png


With my E6600 and this same board temps between Coretemp and TAT were always withing a degree or two of one another... as you can see in the screeny everything is way off, I wouldn't trust any of them.

It seems like TAT for whatever reason thinks that the E4300 has a Tjunction of 100C...
If you look at the temps, there's a 15C difference between the readings, correlating to the 100C vs 85C in Tjunction.

funny though tat also thinks the cpu is a P-M when it is not.

something else thats strange I found is ever since I installed vista TAT reads higher than coretemp lol

before they both read the same, now since vista core temp gives me my same old idle and load temps it and TAT did before and TAT is up to 10C hotter.

Quote:

---

Originally Posted by Unbornchild

Sorry, but you have no idea. Read what CoreTemp actually does!

---

No need to be rude :slapass:

Since lowering Vcore CoreTemp has been performing as it should so maybe its just an anomaly ???

I have found the lowest possible Vcore set in bios @ 3.20GHz fully torture stable to be 1.4125V with my crappy Anus due to the usual Anus massive Vdroop™

With a board with better power reg the bios Vcore should be ~ 1.365V actual which is excellent for low heat tolerant rigs such as an HTPC :clap:

I will be installing this e43 into my ga 965P-DS3P soon for further testing :banana:

Heres a dual 32M @ 3.20GHz whilst browsing

http://i16.photobucket.com/albums/b2...e/32Mx2E43.jpg

I ran a few tests using the x8 multi this morning and found that I could run @ 3.20GHz torture stable using .0125 less Vcore than when using the x9 multi ?????? :confused:

Can anyone confirm ??/