But since my impression of the EMC blogs that I’ve encountered is that they are in fact effectively (nominally) informal extensions of its marketing this does not surprise me overmuch – and this is why I don’t bother to follow them more closely (since I’m not interested in wading through the spin to dredge up whatever nuggets of actual technical worth might exist).

As long as you and your cohorts confine your FUD to your own venues I couldn’t care less how outrageous it may be. But if I happen upon it elsewhere I’ll likely call you on it: I have a fair amount of respect for some of EMC’s products, but very little for some of its corporate behavior.

- bill

And I am an employee of EMC – I’ve never hidden that fact.

As such, were I to contact Samsung directly, I couldn’t publicly disclose what I know about their chips without their approval.

Therefore I have suggested that Beth might want to go back to the Samsung PR contacts she already has spoken with to ask some more questions for added perspective to this story.

As it stands today, this article might well be considered Sun and/or Samsung FUD, implying that the only “good” Flash is this “better” flash, and that everything else is not yet “of age”.

A detailed comparison of the 100,000 vs the 500,000 cycle flash devices would help everyone have a more complete (and hopefully more accurate) understanding of what’s actually been announced here.

But why stop there? What if Samsung’s new flash emits lethal fumes, or occasionally explodes? (I don’t know – just making an example.) I don’t think Samsung said, but if it only returns valid data on Thursdays – buyer beware!

My advice to anyone (such as thestorageanarchist) who actually suspects that these (or other) issues *might* exist is not to ask other people (who apparently have no such concerns) to contact Samsung but to do so himself – after which he would actually have something substantive to contribute here. But perhaps that’s not ‘the EMC way’.

- bill

Although 500,000 writes is more than the rated 100,000 on todays SLC NAND, these chips may only be “better” for the application that Sun requires. I don’t think Samsung has said, but these new “better” chips could well cost more – or have lower performance than – their current 100,000 SLC NAND devices. Drive vendors may not need to pay extra for these “better” chips, nor might they want to suffer the performance penalty (I’m not saying either is the case, just making an example). This is especially true for enterprise-class flash drives like the STEC drives EMC is offering – their technology effectively ensures sufficient life of their drives (>5 years) and performance already.

Until we see the actual pricing and specifications for these “better” chips, they are “better” only according to Sun’s intended use – which they have stated is NOT as an external storage device, but as internal “L3 cache” next to their processors. On a processor board, it simply may not be possible to implement all of the availability and wear-levelling logic that STEC packs into the 3.5″ disk form factor, thus Sun might be willing to pay more so that they save space (and power) on their motherboards.

(I do’t know this – just hypothesizing).

Given this apples-vs-oranges use of flash chips between server and storage vendors, you might reconsider your last paragraph as perhaps unwarranted and unjustified, if not alarmist.

Perhaps you could give Samsung a call and ask about pricing, packaging and performance specs as compared to the enterprise SLX NAND they currently offer? Hopefully Samsung PR can answer without having to call to Korea :*)