Rambus and Spansion Announce Cross-Licensing Agreement
January 15, 2007 6:55 PM
comment(s) - last by
Spansion to pay royalty fees to Rambus for certain memory technologies
Rambus announced last week that it has won another major license agreement deal with a company called Spansion. In the cross-licensing agreement which will last
for five years, Spansion is entitled to a number of Rambus patents that focuses on high-speed memory
interfaces and memory designs. The agreement will be royalty base -- like most other Rambus license grants -- and will give Spansion an edge in flash memory design.
Spansion focuses its products on the mobile, automotive and networking markets, where demand for high-performance memory is high. The company was jointly created by AMD and Fujitsu several years ago and today is one of the world's largest producers of flash memory -- a market that AMD very much prides itself in.
"This license agreement with Rambus enables Spansion to expand our memory solutions for cellular phone applications," said Robert Melendres, executive vice president and chief legal officer at Spansion. "Working with the Rambus team to secure access to their patent portfolio will provide us more design freedom as we develop next generation Flash memory technology and solutions."
Spansion is one of many companies that are currently part of licensing agreements with Rambus. AMD, Elpida, Fujitsu, Qimonda, Matsushita, NEC, Renesas and Toshiba are just some of the companies out there utilizing Rambus' technology.
Rambus holds an iron fist over its patents and products its intellectual property with legendary ferocity.
has reported on numerous ocassions about Rambus and its exercises in the court room. Rambus
recently won a case against Hynix
for patent infringement but Hynix isn't the only one. Micron was also recently entangled in court with Rambus over a
staggering 18 patent infringement claims
This article is over a month old, voting and posting comments is disabled
1/16/2007 5:34:09 PM
Perhaps, he doesn't... But I do.
Rambus was really only seriously useful in the very high end where the horribly high latency was less of an issue than a very high bandwidth stream of data. This is useful especially in database applications where a stored proceedure may be able to fit into the L2 or L3 cache of the CPU and latency hits from swapping into and out of the L2/L3 cache aren't performance killers. Instead in this niche market (which also includes scientific computing, and video rendering), you use small amounts of instruction and need to stream the data at the CPU as fast as possible. For this to work the CPU has to have a fair amount of cache memory, like most server processors have.
Their only real major innovation was that their interface was serialized in a similar fashion to PCIe. This means that increased bandwith is simply a matter of adding channels (like PCIe). The problem with this is that when you have 4 channels, you have to add 4 sticks at a time, or 2 dual channel sticks. In the server space this is no big deal. You pack a box with 16 dual channel slots and add 4 sticks at a time at a high cost, but massive throughput. The enterprise has machines physically big enough, and they have enough money to shoulder the massive expense. The problem this led to was that the more sticks per channel you install, the worse your latency gets.
So for the server space they had some interesting ideas. For many (perhaps most) generic server tasks (non-static web serving, multiple service machines, etc) the amount of cache is simply not enough to offset the latency issues with instruction swapping and data dependencies. For home use and gaming, Rambus was a poor match. ESPECIALLY couple with the pentium 4 which had such amazing penalties to branch mispredicts due to the length of the pipeline. Rambus made this weakness in the P4 so much worse.
If you would like, I can pull some numbers showing how terribly latent Rambus was. BTW, DDR2 is available with fairly low latencies, far lower than Rambus has ever offered. 4-4-4-12
oh, heres some Rambus... Latency is 40ns!!!!!!!!!!
1/17/2007 12:54:07 PM
I'd recommend reading the processor related articles in the Ars Technica paedia:
John Stokes also recently released a book "Inside the Machine"
"The whole principle [of censorship] is wrong. It's like demanding that grown men live on skim milk because the baby can't have steak." -- Robert Heinlein
Judge Overturns $307M USD Rambus Award
July 18, 2006, 5:43 AM
Rambus Drags Micron into Court on 18 Counts of Patent Infrigement
January 13, 2006, 7:39 PM
Report: AT&T Eyeing $40B DirecTV Purchase
May 1, 2014, 8:00 AM
WebOS Class Action Settlement Costs HP $57 Million
April 1, 2014, 10:22 AM
IBM Workers Strike Over Terms of Deal That Will Have Them Working for Lenovo
March 6, 2014, 9:29 AM
Google Picking Up Artificial Intelligence Company "DeepMind" for $400 Million
January 27, 2014, 9:25 AM
Quick Note: Qualcomm Grabs up Palm, IPAQ, and Bitfone Patent Portfolio from HP
January 24, 2014, 9:18 AM
Verizon Buys Intel Media OnCue Cloud TV assets
January 21, 2014, 10:26 AM
Most Popular Articles
Quick Note: Nokia Devices Now Receiving Windows Phone 8.1 Update
July 15, 2014, 10:42 AM
Microsoft Kills Entertainment Unit, May Shelve Flagship Lumia "McLaren"
July 18, 2014, 7:40 PM
Boeing 777 Malaysian Airlines Flight 17 Crashes in Ukraine
July 17, 2014, 1:00 PM
Tesla Confirms “Model III” EV with 200+ Mile Range, Blames Ford for Missed "SEX"
July 15, 2014, 9:12 PM
FBI Report Suggests That Self-Driving Cars Could Be Used as Rolling Bombs
July 16, 2014, 11:02 AM
Latest Blog Posts
Space Terrorism is a Looming Threat For the United States
Apr 23, 2014, 7:47 PM
Facebook Aims to Provide Internet to "Every Person in the World" with Drones, Satellites
Apr 1, 2014, 10:20 AM
Retail Mobile Sites Experience Outages in Light of Simplexity's Bankruptcy
Mar 14, 2014, 8:48 AM
Tesla vs. BMW: Who Has the Safer EV?
Feb 1, 2014, 2:56 PM
Justice Leaks Details of Next HTC One Two Flagship Phone
Dec 5, 2013, 4:04 PM
More Blog Posts
Copyright 2014 DailyTech LLC. -
Terms, Conditions & Privacy Information