Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Rambus (RMBS) - Eagle or Penguin -- Ignore unavailable to you. Want to Upgrade?

To: Zeev Hed who wrote (58425)	10/21/2000 11:35:01 PM
From: Bilow	Read Replies (3) \| Respond to of 93625

Hi Zeev Hed; The patent infringement claims are based, I believe, on the following: SDRAM & DDR: Programmable latency DDR only: Differential clocks Internal DLL 2-Bit Prefetch dramreview.com The only strong claim is the programmable latency, and like Dan says, that can be avoided fairly easily. I differ with Dan, on the other hand, on the question of how much trouble the industry would be willing to go to, to design out that patent. My guess is that the courts will assign a rather nominal value to the patent, (1% is out of line by 10x), but that industry will pay it, if it holds up in court, at whatever rate the courts say. The next generation of memory would naturally avoid it, but that technology won't be dominant until at least 5 years from now. The Differential clock patent is highly unlikely to survive, the technique is ancient and obvious. Avoiding it would be possible, but not convenient. If the industry signs up for the Jazio I/O, differential clocks could be avoided. The Internal DLLs is unlikely to survive, but if it does, it is avoidable with some effort. 2-bit prefetch has no chance of surviving, the technique is not just obvious, it is also the only solution to getting higher bandwidth out of a memory, and it has been done many times in the past. If they did get a patent on this, it would be a travesty in the courts, and it would be very hard to avoid. The concept of prefetching is how bandwidths are matched between busses that have equivalent bandwidths, but differing pin widths. Internally, DRAM has extremely high bandwidth, relatively low frequency, but very high pin (i.e. signal) width. Externally, with DDR or RDRAM (or, likely, even very fast SDRAM), the frequency is higher, but the width is lower. Converting the data from one format into the other is what 2-bit prefetch is about. Every experienced digital designer on the planet has used this technique repeatedly. My conclusion is that the programmable latency is the significant patent, and I believe that this opinion is widely held in the industry. -- Carl