backtop


Print 28 comment(s) - last by Calin.. on Feb 12 at 3:11 AM

The FTC releases its final statement and penalties for Rambus after more than six months of deliberations

Thursday the Federal Trade Commission settled a long standing anti-trust case against Rambus Inc. Originally filed in June 2002, the 4-year case has at least indirectly affected every PC memory manufacturer in the industry. 

The FTC first raised charges against Rambus in June 2002 on grounds the company was violating federal antitrust laws.  The FTC’s decision found Rambus guilty of monopolizing 4 computer memory technologies which were eventually used in DRAM chips.  Specifically, the FTC complaint states Rambus participated in assisting JEDEC's validation and development of DRAM technologies, many of which Rambus had already patented.  The Joint Electron Device Engineering Council, or JEDEC, is an international panel that sets standards for all memory technology manufacturers and integrators to use.

The charges were dismissed in February 2004 in an initial decision and order by Chief Administrative Law, Judge Stephen J. McGuire. McGuire’s decision, however, was overturned in August 2006 by the FTC.

The FTC's decision to re-evaluate its decision against Rambus stated "We find that Rambus’s course of conduct constituted deception under Section 5 of the FTC Act. Rambus’s conduct was calculated to mislead JEDEC members by fostering the belief that Rambus neither had, nor was seeking, relevant patents that would be enforced against JEDEC-compliant products ... Under the circumstances, JEDEC members acted reasonably when they relied on Rambus’s actions and omissions and adopted the SDRAM and DDR SDRAM standards."

The official FTC opinion released by FTC Chairman Deborah Platt Majoras, states "Rambus was able to conceal its patents and patent applications until after the standards were adopted and the market was locked in. Only then did Rambus reveal its patents – through patent infringement lawsuits against JEDEC members who practiced the standard."

Rambus currently charges DDR SDRAM manufacturers that use its technologies a 3.5% royalty. The FTC’s decision has told Rambus it can only charge 0.25% for SDRAM products and 0.5% royalty for DDR SDRAM products. The FTC went on to say that after three years Rambus will not be allowed to levy any royalties against companies for use of its SDRAM/DDR SDRAM technologies.

The FTC explicitly states these royalty changes would not apply to DDR2 SDRAM or other post-DDR JEDEC standards.  Considering Intel transitioned from DDR SDRAM in 2004 with AMD right behind in 2006, the FTC ruling may have very minimal impact on memory manufacturers.

The FTC’s decision is planned to take effect in 60 days. Rambus is "disappointed" with the decision and believes the FTC "ignored key facts." Currently it plans to appeal the decision.

"While we believe it appropriate that the Commission did not reach DDR2, GDDR2, or succeeding generations, we are nevertheless disappointed that the Commission’s remedy with respect to SDRAM and DDR SDRAM continues to ignore the extensive findings of fact made by its own Chief Administrative Law Judge McGuire," responded Tom Lavelle, senior vice president and general counsel for Rambus. "Because we strongly disagree with a number of the Commission’s determinations, we plan to appeal its decision. We believe that a fair review of the underlying facts will restore the perspective of the Chief ALJ who exonerated Rambus by dismissing the complaint."

Rambus's legal stigma does not end at this FTC ruling.  Last year the company was awarded a $307 million patent infringement case against Hynix, a memory manufacturer.  This ruling was then overturned, and two weeks later the FTC reopened its opinion on the anti-trust suit.

And yet, Rambus plays a key role as legal aid and witness testimony against big memory manufacturers in a pending multi-state investigation that claims memory manufacturers artificially set prices for DRAM from 1998 to 2002 -- the same period the FTC claims Rambus monopolized on SDRAM royalties.


Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Old News
By 457R4LDR34DKN07 on 2/6/2007 10:48:53 PM , Rating: 1
Read about this last night on my Wii. Serves Rambus right for the outragous prices and disasterous Intel 820 chipset. Karma




RE: Old News
By GhandiInstinct on 2/6/07, Rating: -1
RE: Old News
By StevoLincolnite on 2/7/2007 12:19:15 AM , Rating: 1
When?
On the Pentium 3, SD Ram out performed Rambus.
On the Pentium 4, SD Ram was like 1/4th of the cost.
Plus with the advent of Dual DDR performance was either equal or better on the Pentium 4.

http://www.anandtech.com/showdoc.aspx?i=1245&p=16
http://www.anandtech.com/showdoc.aspx?i=1541&p=10
http://pcworld.about.com/news/Sep102001id61244.htm

In every case it was either slower, And/or more expensive.


RE: Old News
By GhandiInstinct on 2/7/2007 12:34:09 AM , Rating: 2
RD1066 was faster, maybe a tiny bit more expensive, not by a lot. :)


RE: Old News
By StevoLincolnite on 2/7/2007 1:36:23 AM , Rating: 2
RDRAM PC1066 systems provides significantly more memory bandwidth than single-channel DDR systems. Nonetheless, the performance difference is relatively small (figure on average about 7%, more for some games, less for office apps).

What a dual-channel DDR configuration does is essentially bring DDR up to dual-channel RDRAM levels, so the performance gain is about the same as the difference between single DDR and RDRAM, fairly small.

http://www.pcstats.com/articleview.cfm?articleid=1...

Dual Channel DDR (266Mhz X2) Edges out RDRAM 1066.

http://www.dewassoc.com/performance/memory/ddr_v_r...

Dual Channel DDR edges out RDRAM in most benchmarks.

http://www.simmtester.com/page/news/showpubnews.as...

Average Performance advantage of dual channel DDR was about 1.4%

http://www.gen-x-pc.com/ddrversusrdram.htm

And Again Dual Channel DDR is in the lead.

Think about it this way, When SD Ram PC 100 Cam along, it was only a few % Faster despite almost 33% Faster memory bandwidth. (Compared to PC 66).

ECC:

512Mb of RDRAM PC-1066 goes for about 240 - 260 bucks
http://www.memorysuppliers.com/sam51pcecc53.html

1Gb of DDR PC-2100 goes for about 140 - 160 bucks.
http://www.memorysuppliers.com/in51pc26rege1.html

Non-ECC:

1GB RDRAM PC-1066 - About 450 - 520 bucks.
http://www.memorysuppliers.com/noname5.html

1GB DDR PC-2100 About 140 - 160 bucks.
http://www.memorysuppliers.com/ddrdeme.html

So a slower memory, PLUS a huge price difference.
RDRAM was always expensive, I remember back when it was first introduced it was like 600 bucks for a 64MB stick.


RE: Old News
By masher2 (blog) on 2/7/2007 8:29:16 AM , Rating: 2
Holy link-o-rama. Unfortunately, I've never heard of half those sites...and your link that you labelled as "DDR edges out RDRAM 1066", was actually testing against the much slower RDRAM 800.

The previous poster was correct. RDRAM 800 was, when introduced, the fastest performing memory solution. Later, higher-speed dual-channel DDR began to beat it...then QC RDRAM 1066 came along, and again recaptured the performance crown.

You are correct in that RDRAM was always a very pricey alternative, however. For the average user, the slight benefit wasn't worth the substantial cost increase.


RE: Old News
By bkiserx7 on 2/7/2007 10:35:46 AM , Rating: 2
In college and buddy and I had very comparable systems (p4 2.4, 256mb ram), but my system having rd1066 and his having ddr226. We ran 3dmark2001 with the same video card (ti4200 otes) and my scores were significantly better.


RE: Old News
By StevoLincolnite on 2/7/2007 10:52:37 AM , Rating: 2
http://www.anandtech.com/showdoc.aspx?i=1541&p=10

"In college and buddy and I had very comparable systems (p4 2.4, 256mb ram), but my system having rd1066 and his having ddr226. We ran 3dmark2001 with the same video card (ti4200 otes) and my scores were significantly better."

DDR226? Dont you mean 266?

If so, Anandtech's benchmarks show that DDR333 Edged out RD-800. And thats not just one benchmark, But several.
DDR-266 was only a few % Slower.
And that isnt in a dual configuration. I doubt if you were to pit DDR 400 against RD-1066 things would change all that much. But with DDR266 against RD-1066 RD-1066 wins, But at what cost? Is the small % in improvement worth the extra huge cost? Personally I would use the DDR and over clock it.



RE: Old News
By GhandiInstinct on 2/7/2007 1:53:10 PM , Rating: 2
Lincoln give it a rest, you compared 1066 and 00 to DDR? Even dual-channel? Are you out of your mind, the competition back then was SD and RD then later DDR, even so 1066 edged out DDR. Shhhhhhhhhhh


RE: Old News
By Calin on 2/12/2007 3:11:21 AM , Rating: 2
There was a time when the fast Pentium3 processors had come with 64MB of RDRAM, in order to help adoption. At that time, the retail value of those 64MB RDRAM exceeded the value of the processor (year 2000 or so)


RE: Old News
By Calin on 2/12/2007 3:08:22 AM , Rating: 2
The initial P4 (up to about 1700MHz) was as fast with single channel SDRAM as with RDRAM. The difference in speed started to show up at 2000+MHz.
As for single channel DDRAM, the P4 2400MHz was starting to be memory bandwidth challenged.


RE: Old News
By shimman on 2/8/2007 9:41:00 AM , Rating: 2
totally agree

rambus is like a patentstroll; they are shady, too when it comes to the standardization. rambus is arrogant & hype making just like sony; as a matter of fact, both sony & rambus get alone well.

the royalty fee was the cause of the fiasco of d-ram price fixing, and we have been hit all thanks to rambus.

only ones i know like rambus were investors. i wouldn't miss rambus even if they are out of bussiness.

btw, it was really good thing that rd-ram was killed otherwise we would be forced to use pc with netburst architecture (p4 rings a bell?) with smaller sized & power hungry rd-ram (it had a heatsink attached with a warning sign "hot" when it came out first)

the main problem of rd-ram was the high latency; it was designed for the streaming processing (cell & ps3 & p4); the memory module was difficult to increat the capacity while maintaining the bandwidth.

thanks to those who price fixed (slaped badly by government), we don't have to use rd-ram as intel ditched it for good.

anyway, hopefully this will bring down the price of ram little


They had the patents....
By Comdrpopnfresh on 2/6/2007 11:55:28 PM , Rating: 1
Just because they are trying to receive due royalties on their patents shouldn't mean they have a monopoly. Hell- parts of the human genome are patented by drug companies! If they can pull that, why can't more concrete, intellectual properties be protected?




RE: They had the patents....
By clemedia on 2/7/2007 12:02:55 AM , Rating: 5
I suppose words like deception mean nothing to you.


RE: They had the patents....
By jon1003 on 2/7/2007 5:13:21 AM , Rating: 5
Do you even know what's going on? Here's a simple version:

Rambus helps set the standard for ram by being a consultant to jedec. Meanwhile, they were hiding patents for everything they were recommending to jedec. They just recommended their own stuff so they could set up a monopoly for the IP and secure themselves an infinite revenue stream from royalties. Then when the ram standard is finalized, they whip out the patents and start collecting royalties and suing everyone. Jedec probably thought they were helping in good faith to further technology, while they were actually working for themselves to manipulate the industry.


Applause
By outsider on 2/6/2007 11:14:38 PM , Rating: 2
quote:
we are nevertheless disappointed that the Commission’s remedy with respect to SDRAM and DDR SDRAM...


What do Rambus and Federer have in common? - They are both humble even when they win.




RE: Applause
By jtesoro on 2/6/2007 11:45:27 PM , Rating: 2
Isn't this a loss though?


RE: Applause
By KristopherKubicki (blog) on 2/7/2007 2:00:45 AM , Rating: 2
Are you kidding?!

All of Rambus IP on DDR2 and GDDR memory is untouched -- not to mention there is no retroactive punitive damages. The only thing Rambus is going to get hurt on is legacy memory products, which have been on the decline for sometime now due to the channel direction.


RE: Applause
By jtesoro on 2/7/2007 6:25:53 AM , Rating: 2
True, but I was looking at it from the perspective that they already had it all before, when the initial charges against them were dismissed in 2004.

This new decision takes something away from them. Not much maybe, but still less than what they started out with.


RE: Applause
By CascadingDarkness on 2/7/2007 3:38:01 PM , Rating: 2
Too little, too late I'd say.


nice
By tfk11 on 2/7/2007 4:55:07 AM , Rating: 2
This is long overdue. Nice to see it finally happen. We'll all be better off without companies like rambus.




RE: nice
By Murst on 2/7/2007 9:46:07 AM , Rating: 1
Without rambus, memory technology would be probably 2 years behind what it is now. Be careful what you ask for.

There is a huge difference between the technology a company develops and their business practices. Just look at intel... they haven't had the best business practices, yet where would processors be today without intel's research?


RE: nice
By CascadingDarkness on 2/7/2007 3:40:07 PM , Rating: 2
Maybe farther than we are. If there wasn't poor business practices like this other companies would have a better chance at making real changes.


What's with the hair?
By mezman on 2/7/2007 5:30:54 PM , Rating: 2
Why does the chick in the picture have hair like Camron Diaz after using the "special mouse" in There's Something about Mary?




By KristopherKubicki (blog) on 2/8/2007 5:45:17 AM , Rating: 2
Lol - that's the FTC Chairman. I suppose she is allowed to wear her hair anyway she'd like.


Is this why?
By FITCamaro on 2/7/2007 6:51:39 AM , Rating: 1
Is this why memory prices have been so high lately? I mean this time last year you could easily get 2GB of DDR for $150, cheaper if you didn't want as high a quality. Even DDR2 was cheaper. Now lowend DDR2 in a 2GB kit is a minimum of around $180 with good quality stuff running $200 or more. Heck we've got $600 RAM now. And they can make the stuff cheaper now.

Memory now costs as much or more than your damn processor. If Rambus is the cause of the huge increase in memory prices, the FTC needs to include DDR2 in there.




RE: Is this why?
By Cincybeck on 2/7/2007 6:01:12 PM , Rating: 1
Supply and demand is at least one reason. The demand has increased for a couple of reasons.

1)Both Intel and AMD are using DDR2 now
2)The introduction of Core 2 has lowered prices across the board, giving incentive for people to upgrade now.

Whether they're buying new DDR2 based products or older DDR based products for the value, they need new shiny RAM to go with it.


sigh
By scrapsma54 on 2/8/2007 5:21:54 PM , Rating: 2
Rambus, oh, how you could have been successful were it not for your monopolizing tactics.




"Death Is Very Likely The Single Best Invention Of Life" -- Steve Jobs














botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki