OCZ Blade 3x2GB PC3-16000 CL7 Memory Review

Stability Testing & Overclocking

With every memory shootout, comes stability testing. Our methodology on stability will be put to the test with the high specifications of this OCZ Blade memory. Here is our stability testing breakdown:

Memory stability, what constitutes stable? What is not considered stable? These questions get hotly debated in enthusiast forums all over the internet like little brush fires on the fringe of an inferno that play havoc with forest fire crews. Everyone has their own opinion about stability, especially when it comes to memory stability. For some, stable means they can do whatever it is on their computer without it crashing, blue-screening, or restarting; whether that means gaming or just surfing the internet. To this user, stable means simply using the computer as they normally would.

Of course, that is not what we would call stable and we do a lot more thorough testing before labeling a memory frequency/timing stable. One of the toughest types of programs on a system has been found to be distributed computing projects such as Rosetta@Home, Folding@Home, World Community Grid, and more. Running 24/7 crunching for one of these great causes is a sure way to find holes in a system if there is truly some instability, unfortunately it takes a considerable amount of time to use them for stability testing so we use the list of programs below to all but guarantee the system to be 24/7 distributed computing stable:

• 8 threads of 32M SuperPi Mod 1.5 (ran at the same time using Hyper Pi 0.99b)
• 2 hours of 8 threads of Prime95 using Prime95 v25.7 Stress Testing Blend
• 2 hours of HCI MemTest Pro in Windows using all available memory
• 6 loops of 3DMark Vantage ran back to back

We don't take stability lightly and this list of stability testing shows it. There is also the simple fact that our results are going to be from a single kit of memory. You can't expect every kit purchased to have the same results. The Intel i7 platform relies heavily on the memory controller strength of the processor, especially at these memory clocks, and some Core i7 processors just aren't going to be able to handle DDR3-2000. This type of memory is not for the faint of heart, as the price indicates, and really needs to be mated to a strong setup in order to excel. More on this shortly as we go through our overclock testing section. First up is to test whether our setup can even run this memory at the specified frequency and timings.

Before we even do any "overclocking" we have to see if this memory will run its specifications. The reason overclocking is in quotes is because in order to run memory at DDR3-2000, we are already overclocking. Technically, running memory over DDR3-1333 is overclocking and if you call Intel about a processor that won't run DDR3-1600, even at the default base clock, they will tell you the chip isn't under warrantee for that. Needless to say, there are no memory ratio's to natively run DDR3-2000 without overclocking the base clock, so even without getting technical; we still have to overclock to get DDR3-2000 running. So, with the EVGA X58 3X SLI strapped in and ready to go, let's see how the OCZ Blade PC3-16000 handle their specified frequency and timings.

We decided to stay as close to the default base clock as we could with our stock testing so we opted for the 2:14 ratio and a base clock of 143. As it turns out, the 2:14 ratio isn't that good with this processor but it still handled the highly overclocked memory quite well. For voltages, we had to deal with the known drooping of vDIMM and vTT of the EVGA X58 3X SLI. At first we tried running with 1.65v for vDIMM in the BIOS which results in 1.602v under load but the memory just couldn't do it. Bumping voltage up to 1.70v selected in the BIOS resulting in 1.651v under load has us powering through the stability testing with ease. We also had to run vTT at 1.50v selected from the BIOS which results in 1.423v under load.

So far so good for the OCZ Blade. They run the stock settings no problem on the EVGA X58 3X SLI with our 975 processor. Unfortunately there is some knowledge and skill involved in getting the memory to run at the specified frequency and timings due to the lack of an XMP profile. The nature of the beast just doesn't allow an XMP profile to be very effective with the i7 platform and memory with these specifications so if you want to play with the Formula 1 car on the track, you had better learn how to drive it. We have included the voltage tab of E-LEET in the screenshot above for a reason. If you are using this combination, it allows you to see what voltages we have set in the BIOS for our setup. Obviously your setup may need some adjustments here and there, but it may give you some ideas when getting this combination running.

Now for the bad news, we were not able to repeat the stability testing with this kit and our i7 920 processor. This is the reality of DDR3-2000, some processors won't have a strong enough Internal Memory Controller to run memory at these frequencies completely stable. This is one variable many people don't understand, but obviously a very important one. Next up is our stable overclocking section.

Please do not attempt to try this at home. This overclocking has been performed by a trained professional on a closed circuit. Seriously though, remember, these results are from a single kit of memory. Not all kits are going to be able to overclock like this, and many kits will be better. The most important thing to keep in mind is that every setup is different. The memory controller on the CPU plays a massive role in memory overclocking and every single one is going to respond differently.

With that said, this section is going to be a little bit underwhelming for some people. To say we are expecting a solid overclock is a bit misleading, because we didn't expect the ability to overclock the memory past its specified ratings at all. We have had other more than willing memory matched up with our 965 processor and anything north of 1000MHz or DDR3-2000 has been extremely hit or miss, and certainly not stable enough for our full suite of stability testing. With the addition of the i7 975 processor, however, that changed a good bit as you will see with the 7-8-7 overclocking. What we do anticipate is some pretty impressive clocking at tighter timing sets with this memory. Here is what we have managed to squeeze out of this setup for daily use with reasonable voltages.

The 6-7-6 overclocking of this kit was great. Better than previous kits by a fair margin and is starting to get to a rather lofty frequency for CL6 24/7 stability. We have obviously gone over the recommend 1.65v for vDIMM in all of these results but we have kept the voltage within reason. Under load the 6-7-6 and 7-7-6 voltage is less than 1.73v as the EVGA X58 3X SLI droops vDIMM rather significantly under load.

As we can plainly see, we were able to reach 1000MHz or DDR3-2000 with the timings tightened up to 7-7-6-20 with these OCZ Blade sticks. Essentially, this is the specified frequency with tighter timings. We will be using this overclock in our comparison benchmarks in the next few sections. The performance of this memory at this frequency and these timings is absolutely lethal for memory bandwidth.

Our 7-8-7 overclocking wasn't limited to 1000MHz or DDR3-2000 thanks to the i7 975 processor. Our i7 965 pretty much runs out of gas with every kit of memory we have tested around DDR3-2000~DDR3-2020. With the i7 975 we were able to squeeze in a last minute overclock of 1023MHz or DDR3-2046. By last minute we mean, last minute before the review had to entering the editing process for publication. We are confident this kit has more in it with a little more time or a bit better processor. Here are the screen shots of the three overclocks we achieved here today. Again, the E-LEET voltage is displayed to provide the voltages that we have set in the BIOS for the various overclocks.

6-7-6 Stable Overclock click for full size... OCZ Blade @ 933 6-7-6-20 1T @ 1.727v under load

7-7-6 Stable Overclock click for full size... OCZ Blade @ 1006 7-7-6-20 1T @ 1.727v under load

7-8-7 Stable Overclock click for full size... OCZ Blade @ 1023 7-8-7-20 1T @ 1.751v vDIMM under load