ASRock Z87M OC Formula Helps Break Memory OC Record - 4285.6 MHz DDR

[btarunr]

The one and only ace of overclocking Nick Shih continues to push DDR3 memory frequencies to new boundaries which no man has ever set foot on. He has reached a ludicrous DDR3 4285.6MHz with ASRock's Z87M OC Formula, the micro sized Z87 motherboard he designed specifically for overclocking, and it all happened on this year's overclocking show at COMPUTEX TAIPEI 2013.

ASRock's Z87M OC Formula meets Team Group Xtreem-LV-2666! Nick Shih tops the memory clock world record with an unbelievable DDR3 4285.6MHz score (validation). The masterpiece is done by the world's best overclocking team : Nick Shih, HKEPC's John Lam and Splave.



View at TechPowerUp Main Site

 

[HumanSmoke]

Not bad at all. That bandwidth is at the upper limit of the DDR4 specification.

 

[Jstn7477]

That CAS latency.

Still an incredible memory frequency considering many people still use DDR3 that is clocked at 1/3 of that speed.

 

[btarunr]

That CAS latency.

Given that frequency, it's comparable to DDR3-1600 at 4T CAS.

 

[dj-electric]

Oone does not simply compere CAS latancies to frequencies. They just work differant.

 

[NeoXF]

So is it Z87 OC Formula... or Z87M OC Formula!?

 

[George_o/c]

As DJ Electric said, timings do not matter when someone is aiming for the DDR3 Frequency WR. Not so long ago 3800 and even 3900MHz were considered unreal. And now suddenly, Haswell makes 4GHz look kinda easy with the right sticks. 4285MHz though is a whole other story, my god!

 

[seronx]

The memory speed for this overclock is actually 3213.98 MHz.

http://valid.canardpc.com/show_oc.php?id=2827664
146.09(bus) * 11(multiplier) = 1606.99(actual speed) * 2(DDR) = 3213.98(effective speed)

Also, don't actively trust anything with CPU-Z till 1.64.7/1.64.11 by the end of the week.

 

[George_o/c]

The memory speed for this overclock is actually 3213.98 MHz.

http://valid.canardpc.com/show_oc.php?id=2827664
146.09(bus) * 11(multiplier) = 1606.99(actual speed) * 2(DDR) = 3213.98(effective speed)

Also, don't actively trust anything with CPU-Z till 1.64.7/1.64.11 by the end of the week.

Actually this was done on the x29.33 (2933MHz) DRAM divider.
29.33 x 146.09 = 4285MHz. The score is totally legit

20 minutes ago hiwa posted 4289MHz, so we already have a new WR.

 

[Prima.Vera]

But how much faster is in real applications, ocnsidering the timings and all. Also, what voltage did they used?

 

[Deleted member 24505]

Nice board, this will surely bump up the sales of it.

 

[Geofrancis]

those timings lol thats like saying my car does 200mph on axle stands.

 

[George_o/c]

But that's what you're supposed to do when going for the memory freq WR. We're talking about raw MHz and nothing else

 

[treitano]

So is it Z87 OC Formula... or Z87M OC Formula!?

Z87M=MicroATX

 

[NeoXF]

Z87M=MicroATX

The original reference photo of the board/box was of the full-ATX one, not the micro.


Anyway, that's almost 70GB/s of bandwidth, right?

 

[tikblang]

Interested in 64 Gig

I know this will be impossible in ITX but can we have it (64Gig max) in Thin uATX. My plan is to get a VMware essentials installed on a 1U with commoditized parts.

 

[Exzulus]

^

So is it Z87 OC Formula... or Z87M OC Formula!?

Yeah, states micro.... so the question is, this is achievable on the M = both?.. vauge.

 

[EarthDog]

But how much faster is in real applications, ocnsidering the timings and all. Also, what voltage did they used?

LOL, its not. Memory bandiwidth, by far, is not a bottleneck.

But that's what you're supposed to do when going for the memory freq WR. We're talking about raw MHz and nothing else

People here just don't get it.. (and that's OK). You should see the posts when they cut back cores to get higher frequency on the CPU. People's mind must be blown when they look inside a NASCAR and notice there is only one seat, no radio... etc. Its for RACING ONLY... same with these records.