Super PI Test (The 30 fastest PCs from HardLimit)

[Obione]

Overclocking memories have never been very important and in the Sandy they are very little noticeable. Don't worry, we are old enough to not bother with benchmarks :chuckles:

On the other hand, I think you have put a lot of voltage in it, I hope it is not more than to pass a SuperPI. If I were you, I would not go beyond 1.4 in normal mode.

From what I have read, there comes a point when it asks for a lot of voltage or it does not go beyond a certain frequency. I have tested up to 4.5GHz and the demand for voltage is linear. I leave you the OC/Vcore graph I made, along with a table that also has temperatures. As you can see, the temperature does shoot up from 4 and a half but well, we are talking about pretty cool micros, nothing to worry about with a quality/price cooler, not to mention price. What cooler did you put in it?

http://imageshack.us/f/14/12219012.jpg/

[Bm4n]

Watch out krampak, at those speeds you're going to have to play with the rest of the CPU voltages, and then if you disable a couple of cores it will also help.

[Obione]

Cierrtoo, good point. From 4.8, it's time to cheer up PLL but watch out for overvolt eh

To give it a go with memories, VTT can also help, I think it's on Gigabyte for VCCIO.

Editing: My board is very sad with the voltage adjustment. It has Vdroop, I can only touch Vcore/vRAM, I don't have dynamic voltage for micro and the PLL adjustment is an on/off :hangover:

One of these days I'll try the latest BIOS, a user of the MSI GD65 told me it improved the Vdroop.

[krampak]

So that's it, I've only touched the Vcore to get 4.9. I know it's high, I'm not going to leave it like that. Now I have it at 4.4Ghz factory voltage, the heat doesn't keep up, anyway at 4.7Ghz 1.42V was stable and didn't reach 70ºC with all 4 cores maxed out, this beast I've built does wonders :chuckles:

What PLL/VTT voltages etc did you reach for the 5.1 Bm4n?

Krampak the memories have never been very important and in Sandy they are noticed very little. Don't worry about it, we're old enough not to get worked up about benchmarks

True, I don't get worked up like I did 6 years ago XD I mentioned the memories because I'm not sure if they have to have specific chips or if I'm not playing well with the voltages, that's why I asked what voltage they used.
I also have a lot of Vdroop, 1.5V in BIOS stay at 1.42V at full load, I don't know if there's an option like 'load line calibration' like my P35 had that avoided that problem.

[Tecnho]

Krampak, if that photo is from your build, it's out of curiosity because you have the fans upside down, do you have any inverted case or some kind of invention?

[Xevipiu]

@Obione:

Krampak the memories have never been very important and in the Sandy they are noticed very little. Don't worry, we are already at an age where we can afford to be picky with benches :risitas:

**Well I would say the opposite, in the Sandy it is more determinant, especially because of the chipset, it costs more clock by clock to make aggressive timings, besides you should never go over 1.1v (VTT Gigabyte / VCCIO Asus / CPU I/O MSI), no matter that you have the memories at 1.72v, but do not go over 1.1 of VTT, it will have sudden death xDD

The 5Ghz as a rule at 1.41v real, try to see what it does with only two cores, to have a reference of how bad the micro is

By the way Marçal, I passed by your town the other day xD

Garmin route Telemetric:
105km
http://connect.garmin.com/activity/101185115**

[krampak]

The other way round because? It doesn't do the same function by letting out air as by blowing? I have 32ºC at rest with the room at 26ºC, I don't think it's badly assembled, do I? :S

Thanks for the tip, I won't go over 1.1V xD

[Tecnho]

Well, it depends on the box, but the logical thing in a conventional box is to push the air to the back by pushing from the front to make the air flow, but it depends on the box and how you have it ventilated, that's why I'm asking.

Regards

[krampak]

The case is a normal ATX, if I get bored I will try turning it to see if it improves anything.

[Nemo]

If you put the rear fan in front you push the fresh air inside the fins, with the central one you expel it outside the heatsink and the rear of the case expels it outside the case.
As it is, the central fan has to absorb the fresh air through the fins and it is not as effective.
It won't lower it by 10º but you will notice something.

Best regards

[Obione]

Raising the RAM voltage can mess up the memory controller and it's very rare to notice the RAM, if at all, in synthetic benches and even less so on other platforms. Watch out for Crysis and Metro 2033 (at the end):

http://www.anandtech.com/show/4503/sandy-bridge-memory-scaling-choosing-the-best-ddr3/6

Let's see who spends a fortune on memory to scrape half an FPS.

Krampak you have dynamic voltage. It's more complicated and you have to fight a bit because there are several parameters. Stabilizing the peaks, getting it not to skyrocket under load and having a stable voltage both at rest and under load... I'm glad I don't have these options in BIOS because you'll get dizzy jejeje.

LLC must be there but, I don't know what it's called on the Gigabyte. In the current BIOS with the revised board it doesn't work with manual voltage but in the B2 it worked even in manual and I went from having Vdroop to having more voltage under load. In fact I had the voltages that people get from dynamic voltage but even so with the B3 and a later BIOS I needed less voltage. By the way I was referring to the difference in idle-load, not with BIOS.

[krampak]

Well, after seeing the results from AnandTech, I'm glad I spent 150€ on 8GB xD. I was looking at the Cas 7 and they were skyrocketing to over 300 euros…

[Xevipiu]

@Obione:

Raising the RAM voltage can mess up the memory controller

LLC must be there but I don't know what it's called on Gigabyte. On the current BIOS with the revised board it doesn't work with manual voltage but on the B2 it worked even in manual and went from having Vdroop to having more voltage under load. In fact I had the voltages that people get from dynamic voltage but even so with the B3 and a later BIOS I needed less voltage. By the way I was referring to difference at idle-load, not with BIOS.

**With that comment I made earlier, I meant that you'll mess up the controller sooner by accessing VTT/VCCIO/iO than by overvolting vdimm, I'm not saying this to say it, but it's a fact.

In the new bios, you'll see that there are levels to calibrate the vdroop under load, try to choose the one that suits you best.

After being away from everything and Asus for 3 years, now when I finish my vacation, I will return as a pro-user of Gigabyte, who would have thought I would return after so much time xD**

[Obione]

@Obione:

My board is very sad with the voltage adjustment. It has Vdroop, I can only touch Vcore/vRAM, I don't have dynamic voltage for micro and the PLL adjustment is an on/off :hangover:

One of these days I will try the latest BIOS, a user of the MSI GD65 told me that it improved the Vdroop.

I forgot about the PCH voltage :rollani:

Have you seen the latest Xevipiu BIOS? The ones for the C45 are more packed. I don't know if you are referring to the dynamic voltage, that has been there since they came out but not on MSI. The ones that are working very well are the ASRock ones.

[Obione]

Prueba 2M:

24,358 (stock)
20,262 (CPU @ 4,4 - RAM @ 1600)

[krampak]

Put some screenshot man, otherwise I won't add you in the ranking

[Obione]

Now it has given me 19,684

[_Neptunno_]

Test 2M:

E5800@ 3.2Ghz + 8GB(4x2) DDR3@1333Mhz: 41,539s

And, with the board it has, you can't do OC unfortunately, so I don't know how much this micro would go up :cry:
That said, no matter how much OC you could do to it, it wouldn't even come close to touching the top of the ranking, lol.

Greetings!!

P.D: I wonder if I should pass it on to the i7 920.

[_Neptunno_]

Well, this morning I set out to OC the i7, "properly" and I'm a bit The truth is that with the DFI I hadn't gone over 3.6Ghz and with the Asus I hadn't even touched the OC yet.
After checking that it was easier to OC (I'm not an expert or anything) with the Asus, I started to go up, and up…and I've reached 4.4Ghz (that is, HTT off) and the worst part is that I still have the feeling that it can go higher :fumeta:
Maybe another day I'll try to see where the limit is, because maybe it's already close, I don't know.

Anyway, I'm putting up my screenshot, because I'm happy to be able to say that it's the day of my life that I've done the most OC to a micro, and forgive me for saying that I'm like a flan of emotion :risitas:

i7 920(DO)@ 4.4Ghz HT off : 20.421s

Out of the box it gave me about 32.338s.
That is, possibly more things can be polished in terms of memory and such, but you know that I'm not an expert like many of the forum members in these topics
That is, I didn't raise the vcore at all (it's on auto) and the cooling is air, which, putting my hand on the cold TT, is cold, cold, pardon the redundancy

Greetings!!

P.D: The truth is that seeing this what I posted last night about the E5800 is anecdotal, eh?

[Tecnho]

I reached 4.5 but it was too unstable and very hot, I know I could also reach 4.8 but I couldn't in the end, let's see if my little ones from LGA2011 arrive and when I have it with RL I'll give it a fire to see.