What's new
Carbonite

South Africa's Top Online Tech Classifieds!
Register a free account today to become a member! (No Under 18's)
Home of C.U.D.

Ballistix DRAM Crushes World DDR4 Overclocking Record at 5726MT/s

Sola Scriptura

VIP
VIP Supporter
Rating - 100%
109   0   0
Joined
Apr 22, 2010
Messages
10,171
Reaction score
1,238
Points
17,185
Location
Cape Town
Ballistix, Micron’s gaming memory brand, is now the official overclocking world record holder.

Overclockers used the Ballistix Elite 3600MT/s to set a new overclocking record for the fastest DDR4 memory frequency at a blistering 5726MT/s.
That’s 79 percent faster than the max JEDEC DDR4 speed of 3200MT/s and 115 percent faster than the 2666 MT/s considered mainstream today.

How's that for a speed run?
Yes, this record is a big, big, (big, big) deal to us here at Ballistix. But we care just as much about how we earned the top mark. We’re proud that we were able to use the same CAS latency – CL24 – used by most of the previous record holders. In addition, we set the record using the same production module of the Ballistix Elite 3600 available to gamers today.

That’s what we care about.
The record was set May 13 by Stavros Savvopoulos and Phil Strecker from Overclocked Gaming Systems, using an Intel i7-8086K CPU, an ASUS Maximus XI Apex motherboard, as well as, of course, a liquid nitrogen cooling system.
The veteran overclockers have tested other DRAM that was more finicky, requiring lower temperatures, Savvopoulos said, but the Elite 3600 broke the record without complaint.
“We were blown away by how surprisingly easy it was to overclock these Ballistix Elite DDR4 3600MT/s modules,” Savvopoulos said. “Other modules we’ve overclocked can be temperamental and need to train at temperatures lower than the one required for stability; but we didn’t experience that with Micron’s E-die, which scaled much better with both extreme voltages and temperatures. Overall, it was easy enough to call the whole experience plug and play!”
Ballistix gaming memory is engineered at the die level. The components used to create Ballistix memory are designed, built and tested in-house. Ballistix and Micron owe that degree of quality control to gamers, said Teresa Kelley, Vice President of Micron’s Consumer Products Group. That means that whether you’re using liquid nitrogen to overclock at 5000 MT/s or more, or if you’re a more cost-effective gamer using mainstream cooling at accelerated XMP profiles, the Ballistix 3600 will do the trick, straight out of the box.
“Breaking this world record reinforces our commitment to the enthusiast community,” Kelley said. “Our Ballistix product lineup provides the high-speed, low-latency and overclock headroom that gamers, content creators, and enthusiasts crave. We will continue to focus on offering an exceptional high-performance memory and storage portfolio, with Micron, Crucial, and Ballistix engineers striving to deliver leading-edge technology that redefines the performance boundary.”

 
Must say I have a couple Ballistix 2400Mhz (CL 14s I think) which I'm running at 3066Mhz 16-18-18-something and it's stable at 1.35V I think. I have no idea if this is good or not as I haven't done much OC before, but it did get me thinking, do I even need to buy anything faster now?
 
Got my 3200 CL 16 to 3400 CL 14 with a slight bump in voltage and some tighter timings.

Why did I never mess with RAM overclocking before? Because, damn, does it take time. Change one thing. Boot (if boot), run memtest (18 min). Rinse. Repeat.
 
Last edited:
Got my 3200 CL 16 to 3400 CL 14 with a slight bump in voltage and some tighter timings.

Why did I never mess with RAM overclocking before? Because, damn, does it take time. Change one thing. Boot (if boot), run memtest (10 min). Rinse. Repeat.

Also Blastix I assume?
 
Got my 3200 CL 16 to 3400 CL 14 with a slight bump in voltage and some tighter timings.

Why did I never mess with RAM overclocking before? Because, damn, does it take time. Change one thing. Boot (if boot), run memtest (10 min). Rinse. Repeat.

What is a good real use test case to see how much difference it makes? If I ran Cinebench for a living then yes quicker completion time is good, but how do I see how much real difference it's making? FPS for some I guess?
 
OK. Did a few tests @rinners .

3200 CL 16 XMP vs 3400 CL 14 OC (not straight CL 14, more like 14 17 16 16 32 49) .

UserBench

MCore Read/Write/Mixed 8,94% Increase
SCore Read/Write/Mixed 5.86% Increase
Latency 5.45 % decrease.

Timespy
Total 1.15% Increase (7385)
GPU 0.74% Increase
CPU 2.52% Increase

CSGO (1440p, FPS Benchmark, 2600x)
2.97% increase in FPS (from 303 AVG tot 312 AVG). Very repeatable.

So. Not sure what I have proven. Yes, FPS increases. By an insane amount? No yet. Maybe I will get better at RAM overclocking.

What I have read is that I should be measuring 1% and 0.1% lows, as with faster RAM I should be getting less lows.
 

Users who are viewing this thread

Latest posts

Back
Top Bottom