SanDisk’s Harari: NAND will prevail
“Does Boaz Eitan think there’s a problem with our technology? He’s wrong.”
Gitit Pincas 29 Jun 06 17:56
One of the questions most asked by people interested in the flash memory industry, as well as many Israeli investors is, “Who is right?” The question refers to the well-publicized argument between two of the most highly admired managers in Israel: M-Systems Flash Disk Pioneers Ltd. (Nasdaq: FLSH) founder, chairman and CEO Dov Moran, and Saifun Semiconductors Ltd. (Nasdaq:SFUN) CEO Dr. Boaz Eitan.
The story so far: Saifun announced NROM Quad technology, which enables the storage of 4 bits of information on each memory cell. Shortly afterwards, M-Systems announced its x4 technology, which enables the storage of 4 bits of information on each flash memory cell. Without getting bogged down in technicalities, we can say that the two things are not the same. M-Systems’s technology is based on 16-level flash, and its main innovation is at the system level. Saifun’s technology requires four-level flash, and its innovation is at the physical level.
Eitan told “Globes”, “I don’t care that people say that this is possible and could happen. I’m telling you - and remember this well - it’s impossible to create such a flash.”
Moran responded, “I don’t think that Boaz knows what happening at every fab in the world, and it will be a pity if in a few months he has to feel uncomfortable about what he said.”
Among other things, Eitan explains that the multi-level cell (MLC) processors that currently enable the storage of two bits per cell will soon face a problem. In his opinion, it will no longer be efficient with 50-nanometer technology, and there will then be a need for technologies like Saifun’s. MLC was launched by SanDisk Corp. (Nasdaq: SNDK), founded and run by chairman and CEO Dr. Eli Harari, an expatriate Israeli. SanDisk is a $10 billion company, and Harari entered the fray today in an exclusive interview with “Globes”.
“Globes”: Who is more correct today? Moran or Eitan?
Harari: “Boaz and Dov are both good friends of mine, and I think that this story is a storm in a teacup. Boaz is an excellent physicist, Dov is an excellent systems man, and they look at the same things from different directions. All options are open: both could be right, both could be wrong, or only one of them could be right. If you want me to bet which version will ultimately be right, I’d pick the M-Systems’s version. I believe in NAND flash, and I believe that it will continue to be the leading technology in the coming years, not Saifun’s NROM.”
Harari said that Eitan’s claim that SanDisk’s MLC won’t be good enough when the market reaches at least 50 nanometers was incorrect. “Boaz is a physicist and I am also a physicist. The difference between us is that I know more about MLC. I explicitly state that we can go down in the direction of 30 nanometers, and even 25 nanometers. It won’t be easy, there are a lot of obstacles on the way, but it will happen in the coming years. Boaz is an expert in NROM, and what he doesn’t know, he doesn’t know, and he could err. I prefer to speak about facts, and they are: we will reach 55 nanometers by the end of the year, and we’ll reach 45 nanometers by 2008. SanDisk is now investing billions of dollars in new plants and fabs intended for precisely these technologies, together with Toshiba. Do you think I’d make investments like these if I thought there was a problem? You don’t invest such amounts and then pray. You have to be sure.
“Moreover, in order for Saifun to succeed in its direction, SanDisk must fail. If, on the basis of my knowledge, I had concluded that NDOM Quad would succeed and our technology would fail, I’d have been the first to get on a plane to Israel and sign a contract with Saifun. But that’s not the case. I greatly admire Boaz, he worked with me for many years, and I wouldn’t dismiss what he says, but I see things on the ground. I know more than he does about how NAND MLC works. It won’t stop anytime soon.”
Harari adds that only in a few years will the facts become clear, and it will be possible to determine where the industry is going. “At the moment, everyone is talking about x4, but it should be remembered that Toshiba and we are the only companies making x2 properly. Samsung is trying, but not really succeeding. In any event, this isn't a life-and-death struggle, and it won’t change the world.”
Published by Globes [online], Israel business news - www.globes.co.il - on June 29, 2006