What exactly is the "freshest stuff"?
20nm is a smaller process which inevitably means the raw NAND endurance should decline. IMFT introduced a high-K + metal gate design that keeps the P/E cycles to around 3k (same as 25nm) but how do you know that their claims are accurate? We don't.
Furthermore consumers are getting endurance anxiety and some would prefer 25nm over 20nm just because they believe there is a difference in endurance.
It appears there is some minor performance differences between 20nm and 25nm and if you're trying to sell "high end" SSD's where performance matters, why would you introduce 20nm which would just confuse the heck out of consumers.
I had the same reaction when I saw the news about V3.20 but after thinking about it and seeing OCZ's response to Anandtech I think it's a pretty fair strategy.
I think we all jumped to conclusions without knowing what their strategy is (yes, I'm sure they have a strategy). Kristian Vatto of Anandtech just wrote that he spoke with OCZ about the use of 20nm. He says that the lack of 20nm on the Vertex 4 and Vector is because there is a noticeable difference in performance between 20nm Micron MLC and 25nm Micron MLC. It's not a huge difference, but if you look at the new specs of the Vertex 3 there is a subtle difference. It should be clear though that endurance is unchanged because 20nm uses a high-K + metal gate design.
The interesting thing that he found out was that OCZ is about to EOL the Agility line (3 and 4) which would make the Vertex 3 their only mainstream offering (which appears to be their strategy). The product line would look like this.
Vector - 25nm
Vertex 4 - 25nm
Vertex 3 - 20nm
The use of 20nm in the Vertex 3 is actually kind of smart. First of all, it lowers the BOM of the Vertex 3 considerably (~15%) so it's almost like it negates added cost of using a 3rd party controller. Second, the lower speed actually helps to separate it from the Vertex 4. With these changes there appears to be 3 clear performance tiers and a fairly clean strategy.
It wouldn't surprise me to see OCZ unveil another premium SSD later in the year (using the new SF3?) and then introduce a Vertex 4.20 to replace Vertex 3.20 in the hierarchy.
Point I'm trying to make here is the lack of 20nm support for E2 and BF3 has nothing to do with OCZ's capabilities. Will this strategy pay off? Who knows, but it makes sense and is consistent with what they've been saying about the future of their business.
Why use Sandforce at all? If anything it should be Vertex 4.20 because of the $10 cost savings from using their own controller (let alone the added reliability of using a modified Marvell controller).
It seems they're going to delay the financials as long as possible so that they can show the "new OCZ" (I think we're on OCZ 3.0 now since 2011) whenever they do report. I just don't know how any significant customers can stick with them through all of this.
Plextor is basically the consumer brand of lite-on. Lite-on has a close relationship with Toshiba and just recently (past few months) have a working agreement to develop FW for Marvell.
I know these are rumors but these all seem reasonable.
Plextor talked about being the premium SSD manufacturer at CES and uses Marvell controllers. If they want to remain a player they're going to need to get their own controller. I'm not sure what sort of Enterprise presence Plextor has but I'm sure that's what they want to get into.
Sounds like there's room to decrease inventory a bit. As long as they're close to break even that should allow them to continue to pay off the WF line of credit.
Also mentioned they could sell off the PSU division, but that'd take a while to find a buyer.
Easy to agree with it now in hindsight. Hard to believe they knowingly deceived investors but it seems to be the case. If you reread the transcript from Q1 earnings it's painful. I guess we'll never know about that 'backlog' will we...
Yup. They were right. Ryan was an unbelievable story teller. Certainly got me and a whole lot of institutional investors to believe.
The OCZ/Merriman relationship is still odd. Russell Stanley is ex-Merriman and is now BD for OCZ.
Taipei, Jan. 30 (CNA) Flash memory products are likely to see a sharp price decline over the next 18 months, drawing more enterprises to adopt the super-fast storage technology, American technology firm IBM Corp. said Wednesday.
Kurt Bilderback, vice president of Flash Storage Global Solution Sales at IBM's Systems and Technology Group, said most companies are still evaluating whether they should use flash memory systems, because of the relatively high costs.
Based on the cycle of technology development, however, the prices of flash memory are forecast to have dropped by as much as 50 percent in around 18 months, making it easier for companies to adopt the system, Bilderback told CNA in an interview.
IBM announced in August last year that it was acquiring Texas Memory Systems Inc. (TMS), a high-end enterprise storage technology provider, to expand its storage offerings.
Following the acquisition, IBM plans to invest in and support the TMS product portfolio, and will look to integrate TMS technologies into a variety of solutions including storage, servers and software, the company said in a statement.
Houston-based TMS has more than 300 corporate clients in about 50 countries, dozens of which are Taiwanese firms, mostly in the financial sector.
Bilderback predicted that the quick response time and stability of flash memory will draw more investment from companies that believe "time is money."
For example, he said, people will have to wait for a shorter time during their online transactions on commercial websites that use flash memory systems.
"If you click the checkout button and it takes too long, then you may not check it out," Bilderback noted, adding that flash memory could also shorten the data access time for medical services, Internet-based certification and cloud computing services.
Moreover, flash memory has been used in real-time data computing for weather forecasting and earthquake
Art should have seen this coming and pulled in the reins on the business. Read the transcript and he talks about their capital management model which clearly was clearly wrong.