In a test made by Tom’s Hardware in 11 April to see which memory can be more overclockable, they chose 5 manufacturers, GeIL included.

So, they tested 2 sets of GeIL memory. One bought from the shop, and one received from the manufacturer. While both memories were labeled as 333MHZ, the normal one reached a frequency of 421 MHZ during the tests, while the ‘review sample’ reached 471 MHZ. Meaning 12% more speed. Tom’s Hardware says that the difference between samples should be 2-3% as opposed to the 12% demonstrated in the benchmark.

Then was is going on?

We all know that in the hardware market, reviews are the one that decide if a product is good or not. But what can we do if the manufacturer sends ‘cherry picked’ product, meaning products that will perform better than a normal one?

How can we trust in a review, if the products that are in the stores will never have the same performance as the ones tested? How can we trust in a manufacturer if he tries to fraud its buyers with ‘arranged’ review?

There is one more aspect of this problem. Can you really give a verdict after testing only 2 sets of memories, while we all know that the frequency varies? Why did they not tested 10, or 20? And only after that they could say that the difference is this big and accuse GeIL of sending handy-picked products for testing.

After the reviewers saw the difference between the memories and blame GeIL for trying to get a favorable review, they asked GeIL for a statement. And here it is:

The ability of our standard memory modules to overclock or perform above its rated speed is somewhat different from batch to batch. Typically all of GeIL memory modules will run higher then its rated speed. Our testing standard allows approximately 5% headroom for standard Values series DDR2 modules. However, in some case some modules are able to run at higher speed then others depending on the batch and manufactured date. There are difference in performance in our Ultra and Values series as well, the Ultra series will overclock at higher speed even when rated at the same speed. – Victor Chu, Vice President of GeIL USA

GeIL representatives assure us that there is no guarantee on overclockability of a module; it is only guaranteed to run at the advertised specification.

Could they be right and not trying to save the situation? Maybe, maybe not.

It is for you to judge, the ones that actually by their products.

In any case, as a precaution: read as many reviews for a product, not only from the big review sites who can have hand-picked products for testing, but also from the smaller ones. Look at the notes or verdicts and only then make an opinion.