20nm SSDs Giving G.Skill and Other Companies Problems

By on April 5th, 2011 14:29 GMT

Generally, moving to a more advanced semiconductor manufacturing process is a good thing overall, because it means a lower cost of production, lower power draw and, usually, a better performance, but it seems that NAND doesn't exactly fit into this norm.

Some end-users may be aware of the fact that, not too long ago, G.Skill officially introduced the new Phoenix Evo SSDs.

That series is composed of drives that use NAND Flash chips based on the 2xnm manufacturing process technology.

As such, because of the lower manufacturing costs, the drives should be cheaper than their predecessors and rivals.

Unfortunately, there is a certain disadvantage to the 2xnm manufacturing processes, in the way that NAND chips based on it have fewer write cycles.

In other words, solid state drives constructed with them might not be as reliable as users would like.

What's more, this implies that more space than usual be dedicated to over-provisioning, since storage solutions of this sort need to carry out error correction algorithms.

In order to cope with the complications inherent in this situation, G.Skill, along with all other makers of SSDs, need controllers especially-designed for this flash memory with lower amounts of write cycles.

In fact, this is why the Phoenix Pro still use flash ICs that are still made on the 30nm process, for consistent capacity and rated performance.

All in all, the Phoenix Evo can be seen as more of an experiment, since truly reliable and capacious 2xnm-based SSDs will only be possible once controller makers like SandForce and Indilinx manage to develop chips fit for them.

Unfortunately, since said controllers will actually be more expensive, the price advantage of the NAND chips themselves might be neutralized. In the end, only time will tell if IT players manage to solve these uncertainties in a timely fashion, if at all.

Comments