Ivy Bridge 3570K Testing, Opinions, Results, New Bios, 4.5Ghz At 1.236v

[Don Karnage]

lol Don, can you test it against the A8 with some real games/tests?

I was just playing around as i fixed the heatsinks on Goliath

 

[BallaTheFeared]

You have 1300MHz yet?

 

[Nemesis 1]

The irony was that I was laughing when prior to the release of bulldozer the pics of the AMD water cooling solution was leaked leading many to believe it would be required and thus it would be even more expensive to over clock. Now the shoe is on the other foot.

The Irony of your post is Intel offerred watercooling blocks befor AMD. So your right the shoe fitting was done and over with AMDs BD.
The other irony is these Benchmarks . Why compare what this cheap can do comparred to the K models. Intel Bins their chips . Tis topic is a fail. Until we see retail benchmarks its nothing but heresay.
When the BD romurs were all over the place the same people held those leaks as gods truth. We see were that went . Its he same people claiming IV is a failure even befor its been released or properly benchmarked by a quality review site. Same people differant song . But they still sing out of tune.

 

[richierich1212]

The Irony of your post is Intel offerred watercooling blocks befor AMD. So your right the shoe fitting was done and over with AMDs BD.
The other irony is these Benchmarks . Why compare what this cheap can do comparred to the K models. Intel Bins their chips . Tis topic is a fail. Until we see retail benchmarks its nothing but heresay .

Did you even bother to read the OP? It is retailed box . And it's a K-model.

 

[Grooveriding]

Was doing a burn in at stock settings. Now I'm doing 4.5GHz @ 1.304V and lowering the voltages till I find the lowest limit that is possible at 4.5GHz. Tried at 1.35V earlier and passed Prime, LinX and IBT.

That's pretty warm for closed loop cooling. Can you use less voltage for those clocks or is that after tweaking it ?

 

[LagunaX]

Just in case anyone was curious, A Chiphell user reported these numbers on a retail E1 3770k on a M5G and Thermalright Archon:

In summary
* TDP is at 95 W.
* Packaging is exactly the same as Sandy Bridge.
* Thermalright Archon is used as CPU cooling.
* At 3.9 GHz, full load temperature is 62 C (1.16 V).
* At 4.6 GHz, full load temperature is 78 C (1.25 V).
* At 4.8 GHz, full load temperature is 89 C (1.4 V).
* At 5.0 GHz, 1.47 V is needed to POST. BSOD as soon as Windows is booted.
* Stable with 2400 1T RAM set up. Can POST but BSOD at 2600. Corsair CMT16GX3M4X2133C9 2 x 4GB.
* Temperature drops really fast once in idle.
* System pulls 51.79 W at idle and 116.26 W at full load (AIDA64) with stock clock speed.
* System pulls 74.12 W at idle and 146.69 W at full load (AIDA64) with 4.6 GHz.
* System pulls 86.41 W at idle and 183.77 W at full load (AIDA64) with 4.8 GHz.

Looks like 4.6-4.7ghz might be the target speed to aim for.

 

[dxkj]

Hrm, built an aquarium pc for fun the other week. I wonder if it would be worth pulling out i5 SB and plopping in an IB to see what kind of temps and OC 15 gallons of mineral oil would dissipate.

 

[dma0991]

That's pretty warm for closed loop cooling. Can you use less voltage for those clocks or is that after tweaking it ?

I'm pretty sure I could as I'm currently doing about 1.29V and it seems very stable. I'm not too worried about the temps as it is within a range that I consider it to be safe. I'm not quite sure if I should drop the voltages further than 1.3V as LinX and IBT scores seems to be dropping as well, same clockspeed @ 4.5GHz.

 

[Grooveriding]

I'm pretty sure I could as I'm currently doing about 1.29V and it seems very stable. I'm not too worried about the temps as it is within a range that I consider it to be safe. I'm not quite sure if I should drop the voltages further than 1.3V as LinX and IBT scores seems to be dropping as well, same clockspeed @ 4.5GHz.

Safe for sure. Just to me looks a bit on the high side at 75C in a closed loop. I use about 1.34 for 4.7ghz and get mid to high 50s and 1.48 for 5ghz and get low to mid 60s.

Do you have temperature numbers for an OCed SB using your current setup as well ?

 

[dma0991]

Safe for sure. Just to me looks a bit on the high side at 75C in a closed loop. I use about 1.34 for 4.7ghz and get mid to high 50s and 1.48 for 5ghz and get low to mid 60s.

Do you have temperature numbers for an OCed SB using your current setup as well ?

Unfortunately I don't have a SB. The temps fluctuates a lot and it is hard to get a value that is constant. As far as WC goes, mine is only a step up above the XSPC Rasa RS240 kit, nothing extravagant like a 360 rad or higher end. Got to factor in my 30C-ish ambient temps as well.

 

[Absolute0]

loading at 70c at 1.3v on a 22nm IB with 30c ambient is actually quite good... 30C is frigging HOT ambient. This would be a 62C load at normal ambient.

My SB will load at nearly 70 at 1.3v and Im on water too. My chip is hotter than average.

 

[BallaTheFeared]

Just in case anyone was curious, A Chiphell user reported these numbers on a retail E1 3770k on a M5G and Thermalright Archon:

In summary
* TDP is at 95 W.
* Packaging is exactly the same as Sandy Bridge.
* Thermalright Archon is used as CPU cooling.
* At 3.9 GHz, full load temperature is 62 C (1.16 V).
* At 4.6 GHz, full load temperature is 78 C (1.25 V).
* At 4.8 GHz, full load temperature is 89 C (1.4 V).
* At 5.0 GHz, 1.47 V is needed to POST. BSOD as soon as Windows is booted.
* Stable with 2400 1T RAM set up. Can POST but BSOD at 2600. Corsair CMT16GX3M4X2133C9 2 x 4GB.
* Temperature drops really fast once in idle.
* System pulls 51.79 W at idle and 116.26 W at full load (AIDA64) with stock clock speed.
* System pulls 74.12 W at idle and 146.69 W at full load (AIDA64) with 4.6 GHz.
* System pulls 86.41 W at idle and 183.77 W at full load (AIDA64) with 4.8 GHz.

Looks like 4.6-4.7ghz might be the target speed to aim for.

So stock to 4.8GHz increases power draw by less than 70 watts but increases heat by over 40C?

Kinda lame actually.

 

[Don Karnage]

Just in case anyone was curious, A Chiphell user reported these numbers on a retail E1 3770k on a M5G and Thermalright Archon:

In summary
* TDP is at 95 W.
* Packaging is exactly the same as Sandy Bridge.
* Thermalright Archon is used as CPU cooling.
* At 3.9 GHz, full load temperature is 62 C (1.16 V).
* At 4.6 GHz, full load temperature is 78 C (1.25 V).
* At 4.8 GHz, full load temperature is 89 C (1.4 V).
* At 5.0 GHz, 1.47 V is needed to POST. BSOD as soon as Windows is booted.
* Stable with 2400 1T RAM set up. Can POST but BSOD at 2600. Corsair CMT16GX3M4X2133C9 2 x 4GB.
* Temperature drops really fast once in idle.
* System pulls 51.79 W at idle and 116.26 W at full load (AIDA64) with stock clock speed.
* System pulls 74.12 W at idle and 146.69 W at full load (AIDA64) with 4.6 GHz.
* System pulls 86.41 W at idle and 183.77 W at full load (AIDA64) with 4.8 GHz.

Looks like 4.6-4.7ghz might be the target speed to aim for.

Looks about right

 

[Olikan]

When the BD romurs were all over the place the same people held those leaks as gods truth. We see were that went . Its he same people claiming IV is a failure even befor its been released or properly benchmarked by a quality review site. Same people differant song . But they still sing out of tune.

BD rumors were all over the place, as you said...sometimes truly amazing, others very weak... ended up beeing the 2nd option...

IV, is like an opera, everyone saying the samething...what are the odds of everyone lied?

 

[notty22]

The wattage is more telling. Intel could have positioned the sensor in a certain area that reports temps higher than SB.
I'm sure we will see direct comparisons by review sites, with nice plotted curves of temp/TDP comparing IB to SB.

 

[T_Yamamoto]

The wattage is more telling. Intel could have positioned the sensor in a certain area that reports temps higher than SB.
I'm sure we will see direct comparisons by review sites, with nice plotted curves of temp/TDP comparing IB to SB.

Maybe so this is make temps more accurate to prevent degration?

 

[podspi]

Whoa. Hot, but that power usage is amazing Doesn't take into account the rest of the machine, but my machine hits over 400W (Oc'ed 1055t @ 3.3ghz + Turbo) at the wall under full load

 

[pelov]

That's on a Thermalright Archon, which is as good as a Corsair H80 as far as cooling goes.

Does anyone have anything on how it fairs on stock cooling? Just for comparative purposes.

 

[PlasmaBomb]

Just in case anyone was curious, A Chiphell user reported these numbers on a retail E1 3770k on a M5G and Thermalright Archon:

In summary
* TDP is at 95 W.
* Packaging is exactly the same as Sandy Bridge.
* Thermalright Archon is used as CPU cooling.
* At 3.9 GHz, full load temperature is 62 C (1.16 V).
* At 4.6 GHz, full load temperature is 78 C (1.25 V).
* At 4.8 GHz, full load temperature is 89 C (1.4 V).
* At 5.0 GHz, 1.47 V is needed to POST. BSOD as soon as Windows is booted.
* Stable with 2400 1T RAM set up. Can POST but BSOD at 2600. Corsair CMT16GX3M4X2133C9 2 x 4GB.
* Temperature drops really fast once in idle.
* System pulls 51.79 W at idle and 116.26 W at full load (AIDA64) with stock clock speed.
* System pulls 74.12 W at idle and 146.69 W at full load (AIDA64) with 4.6 GHz.
* System pulls 86.41 W at idle and 183.77 W at full load (AIDA64) with 4.8 GHz.

Looks like 4.6-4.7ghz might be the target speed to aim for.

So stock to 4.8GHz increases power draw by less than 70 watts but increases heat by over 40C?

Kinda lame actually.

Huh?

The CPU is only drawing like 70W at stock... and pretty much doubling the CPU consumption raises the load temp by 27 degrees celsius... (89 - 62= 27)

 

[Ferzerp]

So stock to 4.8GHz increases power draw by less than 70 watts but increases heat by over 40C?

Kinda lame actually.

No, it increases *temperature* by over 40C. It increases heat by 70 joules per second.

 

[BallaTheFeared]

Huh?

The CPU is only drawing like 70W at stock... and pretty much doubling the CPU consumption raises the load temp by 27 degrees celsius... (89 - 62= 27)

Yes, yes it was a last minute edit I was getting confused by the separation!

Boy two corrections in a row, feels like I'm back in school and still don't care!


What I was trying to say was that the wattage increase doesn't seem, at least to me, to justify the temp increase. But then again I've never once ran SB on anything but water so perhaps my perceptions of what air can do are off.

So about 70w at stock, and 132w overclocked... So um, why isn't the heat getting out of the chip? 132w should be a breeze for even low end coolers.

I know the die is smaller, but when was the last time that caused such a drastic increase in temps?

Are we positive the temp sensors reading programs are calibrated correctly?

Jesus I need sleep - But 40nm to 28nm didn't see a massive increase in temps, even with the same power draw, not sure I believe the story about how it's too small to dissipate heat. At this time I think it might just be a sensor error.

 

[Don Karnage]

I love microcenter. Check out the price

 

[LagunaX]

that is a steal!

 

[Borealis7]

can you please list max stable overclock with stock volts?

that's what i'm aiming for with my next build.

 

[T_Yamamoto]

that is a steal!

Too bad the only computer shop close to me is compusa