ocaholic - Memory > DDR3 > Corsair Vengeance Pro Series 2x4GB DDR3-2933 CL12 1.65V Review

[pagebreak]

Closer Look

The top banana in the Vengeance Pro series is shipped in its own cardboard box enclosed with an additional two-fan cooling unit.

There are multiple colour schemes available for the Vengeance Pro, but Corsair have restricted the 2933C12 model to red only. In case this color doesn’t fit into a specific build, the red pieces at the top are removable, making the modules almost completely black.

As good-looking as the heatsinks might be, they are not what whe would call solid. The tops are very loose and the whole thing feels as if it was made out of plastic and not aluminium. The cooling efficiency, however, is not too bad: the modules were warm but not hot under full load.

The heatsinks are held on via a layer of thermal glue on the populated side of the modules and with a strip of thermal tape on the other blank one. Such attachment makes them relatively easy to remove, even without preliminary heating.
Having detached the heatsinks, we found our modules to be based on Hynix chips of 4Gbit density, commonly referred to as MFR. These chips are known for their ability to operate at high frequencies at loose timings, which made them a rather obvious choice for this spec.

The SPD module of the modules contains some basic information such as maker, part number and manufacturing date. It also features JEDEC profiles up to DDR3-1333, that are obligatory to make the memory work out-of-the-box, as well as an XMP profile up to its rated speeds at a change of just one BIOS option.
Detailed overview of the top JEDEC profile and the XMP can be seen on screenshots below.

Page 2 - Closer Look

[pagebreak]

Photo Gallery

Page 3 - Photo Gallery

[pagebreak]

Testing Method & Test Setup

To test the overclocking capabilities of the our kit, we are going to use an overclocked Intel Haswell platform due to its fast memory controller, which allows for all testing to be done within a reasonable period of time.

Motherboard	ASUS Maximus VI Impact (BIOS 0217)
CPU	Intel Core i7-4770K @ 4.5 GHz
Graphic card	ASUS GTX 580
Memory	Corsair Vengeance Pro CMY8GX3M2B2933C12R
SSD	Samsung PM840 Pro
PSU	Seasonic Platinum 660 Watts
OS	Windows 7, 64 bit SP1

Even though Haswell is very flexible on the memory frequency one can set, very few people actually do base clock (BCLK) overclocking on their daily setups. Therefore, we fix the BCLK frequency at default 100MHz, only adjust the memory multipliers and then minimise the voltage for each given setting.
As usual, our stability method of choice is HCI Memtest. Since we are dealing with an 8 GB kit, we use eight 750 MB instances and call things stable if we see all of them to go past 150% without showing a single error.
Not to get things too complicated, we only set the primary timings, command rate (1T) and the memory voltage by hand while the rest of the settings are left for the motherboard and SPD to agree on.

Results

Traditionally, our aim is to see how memory reacts to voltage and timing adjustments and in case with Hynix MFR we have an almost linear dependance between stable frequency and voltage using a constant CAS latency. What comes to second and third entries in the primary timings list, tRCD and tRP, the main factor for lowest stable values for those is frequency, with voltage playing a role only in borderline areas. Lastly, the tRAS value we use in each part of the testing is a sensible value induced by an educated guess.
A good indicator of quality on Hynix is voltage required for stability at DDR3-2800 at CL11. The yard stick in this case would be 1.70V with truly great sets being able to go under 1.65. In case with our Vengeance pro, we have a value of 1.72V, which isn’t exactly what we would call good.
Another bad sign is the voltage headroom that Corsair have left for specs at 1466MHz CL12. Given that some motherboards can undervolt the memory by more than 0.01V compared to what is set in the BIOS, we see a potential source of stability issues for those who only use XMP.
On a positive note, the memory is very flexible in terms of the timings it can run at lower frequencies and, given a good enough memory controller and enough voltage, can also be stable above DDR3-3000.

Page 4 - Results

[pagebreak]

Conclusion

As usual, we finalise the article by weighing what you get in comparison to what you pay. According to geizhals.at, the sets of CMY8GX3M2B2933C12R, just like ours, currently retail for more than 300 euros.
Which is a lot knowing that you only get 8 GB of memory in return. Surely, some applications do not require high volumes of memory, but if speed is what you’re after, then you won’t be getting 300 euros worth of it either. As these modules are single-sided, they will be slower at the same clocks and timings compared to dual-sided counterparts hence the extra clocks will not pay any dividends. And because Corsair have employed high-density (4Gbit) chips to build these modules, some of the secondary timings, such as tRFC, must be set looser in comparison to modules using chips of lower density (2Gbit) thus leaking more performance.
What you actually get for your money are the looks, the bragging rights and a fancy pirate logo. Not much, but then again, specific things matter to specific people.

Award

Overall, we rate the 2x4GB set of Vengeance Pro 2933 CL12 at 3 stars out of possible five.