Crucial Ballistix PC3-16000 (DDR3-2000) 2GB Kit

Test Setup & Results

Test Setup
 
With DDR3 speeds ramping up on an almost weekly basis, It's extremely difficult to obtain two or more memory kits with similar specifications to perform any kind of worthwhile comparison. Therefore, the Crucial Ballistix PC3-16000 kit will be reviewed as a standalone product using the hardware specified below.
 
Test Setup Specs
 
To guarantee a broad range of results, the following benchmark utilities will be used:
 
Synthetic Memory Subsystem
• Lavalys Everest 4.0
• SuperPI Mod 1.4

3D Benckmarks & Games
• 3DMark05
• Unreal Tournament 3
 
 
Results
 
One of our biggest fears when receiving a memory kit clocked as high as the Ballistix PC3-16000 is that we'll be required to jump through hoops just to get it to run at its stock speed and timings. Initially things looked pretty bleak for the Ballistix kit, with our 790i based Asus Striker II motherboard refusing to boot whenever we attempted to take the memory past DDR3-1800. However, knowing that this was the first time we'd ever attempted to run DDR3 this fast on the Striker board, we enlisted the help of ASUS to see if there was anything wrong at their end.
 
Surely enough, after a single magical BIOS upgrade, the board booted first time at DDR3-2000 / 9-9-9-28. Proving once again that Crucial really do provide memory that works as advertised, right out of the box.
 
 Crucial Ballistix PC3-16000 Stock Crucial Baliisitix PC3-2000 Overclocked
 
Being Overclock3D we were obviously not content with the 2000mhz effective stock speed of the Ballistix kit. So by raising the FSB speed of the CPU and increasing the VDIMM voltage of the memory to 2.0v we managed a thoroughly respectable 72mhz (144mhz effective) overclock out of the kit, taking it to a total speed of DDR3-2144. The chart and graphs below show how this, and the other settings tested on the Ballistix kit, affected its performance:
 
Crucial Ballistix PC3-16000 Results
 
Everest Memory Read/Write/Copy
 
Everest Memory Latency
 
3DMark05
 
SuperPI Mod 1.4
 
Unreal Tournament 3
 
While you would expect the highest clocked DDR3-2144 results to be the champion across all of these benchmarks, this couldn't be much further from the truth. As the Everest Memory Latency and Bandwidth results above show, the Ballistix kit exhibited a sharp reduction in performance once passing DDR3-2000 with latency rising over 20ns and bandwidth dropping by around 2500 MB/s. We can only imagine that after passing DDR3-2000, some kind of internal divider comes into effect on the 790i chipset (or Striker II motherboard) resulting in crippled bandwidth figures.
 
However, good all-round improvements were seen by reducing the memory speed slightly down to DDR3-1902 and tightening up the memory timings to 7-7-7-28. This offerred a good balance of bandwidth and latency which was reflected in both 3DMar05, SuperPI and the Everest results.
«Prev 1 2 3 4 Next»

Most Recent Comments

22-05-2008, 08:39:59

JN
"With a stock speed of DDR3-2000 Crucial have one extremely fast DDR3 kit on their hands. But will it overclock any further? We find out."

http://www.overclock3d.net/gfx/artic...125750600s.jpg

Crucial Ballistix PC3-16000 Review

22-05-2008, 09:07:22

w3bbo
Impressive review m8. Glad to see they hit the magical 2000.

22-05-2008, 09:18:37

Bungral
Nice review Jimbo.

I get the whole argument about timings vs mhz but what is odd is that in the Everest Memory Latency tests with 9-9-9-28, the overclocked settings lost by quite a bit even with the same timings. That overclock seemed to bugger them up quite a bit.

Only lil bit of uniformity seemed to be when the 2144mhz at 9-9-9-28 and the 1902mhz at 7-7-7-28 won out in superpi and 3dmark... But even then it reversed to the other two settings in UT.

Man... The argument runs on.

22-05-2008, 09:49:50

Rastalovich
What bites me atm is 4x1g of this is looking at 600 with some possible playing around with volt/timings to accomodate a non set.

They do a 2x2g set for ~220 or so - now if ur buying components for a new system:

Q6600 + 4x1g (16000 as reviewed) = approx 730/740 ?

QX9650 + 2x2g (pfft who cares) = approx 770/780 ? plus u have a load of flexibility.

(figures are up in the air, but there`s a point to be had)

22-05-2008, 11:40:03

JN
Quote:
Originally Posted by name='Bungral'
Nice review Jimbo.

I get the whole argument about timings vs mhz but what is odd is that in the Everest Memory Latency tests with 9-9-9-28, the overclocked settings lost by quite a bit even with the same timings. That overclock seemed to bugger them up quite a bit.

Only lil bit of uniformity seemed to be when the 2144mhz at 9-9-9-28 and the 1902mhz at 7-7-7-28 won out in superpi and 3dmark... But even the it reversed to the other two settings in UT.

Man... The argument runs on.
Tell me about it :s For a lot of it you can put it down to certain benches liking the higher OC of the CPU or the lower memory timings, but the DDR3-2144 results are just wierd and something that ASUS are actually trying to reproduce at the moment.

22-05-2008, 16:50:58

Hassan
These are one nice looking beasts

Can I have 'em - pretty please
Reply
x

Register for the OC3D Newsletter

Subscribing to the OC3D newsletter will keep you up-to-date on the latest technology reviews, competitions and goings-on at Overclock3D. We won't share your email address with ANYONE, and we will only email you with updates on site news, reviews, and competitions and you can unsubscribe easily at any time.

Simply enter your name and email address into the box below and be sure to click on the links in the confirmation emails that will arrive in your e-mail shortly after to complete the registration.

If you run into any problems, just drop us a message on the forums.