SandForce Next Generation SSDs, SF-2000

[panda dim]

I have some information of SandForce Next Generation SSDs, SF-2000. Which are: SandForce has been significant me that the market is actually going to obtain exhilarating or exiting once its upcoming-generation controller is prepared. I did not actually consider it, merely because that is what each company informs me. But in this case, at least based on what SandForce showed me, I almost certainly should have. Drives will be sampling to enterprise clients in the upcoming weeks, but they most likely won’t see shipping hardware until Q1 2011 if the whole thing goes according to plan. This much information I have about this and require more information if possible.

[Talish]

I have some information about SandForce Next Generation SSDs, SF-2000. The overall structural design of the SF-2000 leftovers unaffected from what we have today with the SF-1200/SF-1500 controllers. SandForce’s controller obtains around the inherent troubles with writing to NAND by purely writing fewer. Using genuine time compression and data reduplication algorithms, the SF controllers stock up a representation of your information or data and not the definite data itself. The compact data stored on the drive is also encrypted and stored redundantly crosswise the NAND to assurance against data loss from page level or block level breakdowns. Both of these features are made likely by the reality that there is purely fewer data to handle.

[Talleen]

Side effect of SandForce’s write-less policy is there is no require for an external DRAM to grip great mapping tables. It decreases the total BOM cost of the SSD and permits SandForce to accuse a premium for its controllers. These are the essentials, they have not distorted. The fresh SF-2000 controller is quicker but the basic algorithms remain the similar. The three areas which have been enhanced however are the NAND boundaries or interface, the on-chip memories, and the encryption engine.

[TalinF]

I actually don't find the mania or observation whit raid specifically raid 0. It is the IOPs (Input/Output Operations per Second) which count up for how rapid your system boots or begins programs and whit 60k IOPs I think you are covert. Placing these drives in R0 could really for a few data patterns slow them down as information is alienated in excess of 2 drives they have to get there at the similar time or one of the drives have to stay or wait for the other to become equal. You will observe a vast boost in sequential reads/writs but whit tiny casual statistics or data the advantage would unhelpful, and the generally advantage would be approximately up to a 6% profit. And the down side would be the upper risk of data defeat if one of the drives smashed down.

[Tajdar]

Typical PC boot and application loading is linear in nature, it is just standards which attempt to do a number of things at the same time, very limited applications or servers which require IOPS considerably more than random read or write performance. You are too mistaken about slowing them down waiting because if not the drives DRAM cache, there is the system chief memory cache, and on a number of raid controllers there is even the 3rd level of controller cache on the card.

[archer]

SandForce is not publicizing real products based on the original controllers only up till now, but I told I can anticipate drives from numerous of their existing partners. With this line of SSD processors, SandForce will first be bringing in products into enterprise space, but new products will finally drop down into the consumer space too. I have got the complete press release from SandForce, along with a few additional slide-wares. Taking into reflection or consideration the features and performance of their present hardware, I am excited to get my hands on upcoming-generation hard state drives based on the new SF-2000 series controllers while they become accessible.