LONDON – PC and consumer goods firm Apple is in talks to buy startup Anobit Technologies Ltd. at a price of between $400 million and $500 million, according to a report in a Hebrew publication called Calcalist.
Anobit (Herzeliya, Israel) has developed NAND flash controller technology that can improve the apparent endurance – the number of reads and writes flash memory can perform. The raw endurance performance of flash memory reduces dramatically with reduction in the geometry of the manufacturing process technology, to the point where flash memory becomes unsuitable for many applications.
Anobit was founded in 2006 to make solid-state drives. It's MSP (Memory Signal Processing) technology is a combination of error correction and memory management schemes that compensates for errors and evens out wear thereby allowing higher apparent performance.
Low endurance can result in the need to install additional memory to provide endurance headroom. Anobit began production of its MSP2020 NAND flash memory controller in cooperation with Hynix Semiconductor Inc. earlier in 2011. Chips from Anobit are already incorporated in a number of Apple products, including the iPhone, the iPad and the MacBook air, the report said.
Anobit is examining a large financing round with a leading Asian flash memory manufacturer, the report said, and the talks with Apple may end up with both companies becoming strategic investors in Anobit, the report said.
Anobit is believed to have received more than $72 million from venture capital firms including Pitango, Battery Ventures, Intel Capital and Micron Technologies and has about 200 staff.
In August Anobit announced that it had shipped more than 20 million flash controllers and that its MSP2025 supported up to 256-Gbyte of NAND flash implemented on 20-nm and sub-20-nm process technology.
Prior to co-founding Anobit, Professor Ehud Weinstein was a co-founder, chairman
and CEO of Libit Signal Processing, which was acquired by Texas
Instruments in 1999. Ariel Maislos is a fellow co-founder of Anobit prior to which he was co-founder and president of
Passave, which was acquired by PMC Sierra in 2006. CTO Avraham Meir was previously vice president of corporate engineering at SanDisk and the CTO at M-Systems until it was acquired by SanDisk.
Anobit entered the Silicon 60, EE Times' list of emerging
startup companies at version 12.0 in 2011. The latest edition of the Silicon 60 is version
12.5, which is the topic of a detailed technology and employment digital
edition which can be accessed via http://e.ubmelectronics.com/Silicon60/index.html
Apple may also be looking to do heterogeneous integration of logic and other functions into the flash memory products via 3D/2.5D stacking. The reliability of memory components is paramount to the overall product reliability of a 3D/2.5D stacked product and perhaps this is one way Apple thinks it has an edge over others in higher yields.
Apple is actively participating in IC designing and slowly the may be interested in fab. It is good move as they have huge volume to support it. Alongwith acquisition, Apple should also make more efforts on innovation.
The only reason Apple blocks IP is to stand in the way of innovation, which is not a "Good move" at all. In this time of global uncertainty they challenge anyone and everyone which could cause lots of people to lose their jobs, that is just selfish. The Patents laws need to be overhauled and there needs to some limitation on exercise patents just to screw people over.
Since Apple is already using their tech in all its products, I guess they are just blocking the IP so that competition wont get any hands on it. Good move.
btw the valuation is too good, for a 5 year old semi startup. shows that still there are ways to do a semi startup without burning a lot of VC cash.
I don't know exactly what Anobit do to improve memory endurance. But given the backgrounds of the founders in digital signal processing, and that their technology is called memory signal processing (MSP) I suspect they have some way of reading data back from memory cells and can tell about the likelihood of a cell failure. With this information is should be possible to push endurance closer to the edge.
They may also have worked out superior ways to encode data so that when a cell failure does occur they can recover the lost bit and move on.