Slashdot Mirror
DRAM Price Fixing Investigations
An anonymous reader writes "A few days ago after FTC antitrust charges against Rambus were thrown out, the U.S. Department of Justice and EU have both begun probes against the 4 largest memory makers in accusation of price fixing during 2001/2002. News.com.com has information regarding the pending EU investigation. Anandtech and Silcon.com both have primers on the U.S. investigation. If you thought you paid too much for RAM in 2002, chances are you may have been more right than you originally thought."
I don't like paying high prices any more than anyone else, however I have to wonder...
Let's say I have a monopoly on widgets, or myself and my compeptitors agree to keep the price of widgets artificially high.
At what point are we no longer allowed to sell our widgets at whatever price we see fit? When do we cross over into breaking the law for price fixing?
Karma: Chameleon (mostly due to the fact that you come and go).
...for ~$5.48, from some RAM manufacturer. In 2007.
Man, I just deposited my $13.86 RIAA check yesterday.
If the money keeps rolling in like this from Big Greedy/Evil Organizations, I may quit my day job.
Man, memory seems so cheap these days. If it was being fixed before, I can't imagine what it'd be like without price fixing.
May we never see th
Will this lead to cheaper RAM? I hope so, because I really need more... um, what's it called?
If we had MORE coordination like this from countries like Japan and China (who are the prime violaters of undercutting, sweat shops, etc), we might finally be able to level the playing field a bit more in the international trade arena.
Perhaps if the US and EU could work more closely together they could force our Pacific Rim and Asian counterparts to see that their unfair practices are not... appreciated.
So, they put together a cartel to do some racketeering. Sounds illegal to me.
Evolution or ID?
If you thought you paid too much for RAM in 2002
I paid 70 for a stick of 512mb in 2002. That may be expensive compared to now, but with my first 486 I paid the same price for a 4mb stick. Unless you own a server farm, what's 10 per half a gig or ram?
When anger rises, think of the consequences.
Confucius (551 BC - 479 BC)
Price Fixing? Are they sure? In 2002, they where practically giving them away.
Let's get a class action going. A voucher for a single chip from a DRAM bank would surely make up for this affront. That and millions for the lawyers.
Saving my mod points....
While I'm willing to give any company the benefit of the doubt, it does seem rather suspicious that Micron chose to sell off their PC arm and focus instead on, the implied, more lucrative memory manufacturing business line. Circumstantial yes, but it never made sense why Micron would sell of a business line that was the only good alternative to Dell.
That being said, it's really hard for the DoJ to prove a conspiracy existed to fix prices of memory between manufactures. IANAL, however from my understanding basically a "smoking gun" would be the only way a conviction could be had - some emails between companies discussing price or marketing strategies perhaps. Other than that, it's almost impossible to get a price-fixing case with a favorable outcome for the prosecution.
I'd like to see the price of toner and ink cartridges go down. Those things seem so simple, I wonder why they are so expensive. A memory chip seems slightly more expensive to produce than an ink cartridge. Yet the prices are very similar.
http://github.com/gbook/nidb
I've read cases where the same laws have been used to prosecute companies no matter what they do:
A law that you can't know you're breaking in advance is no law: it's a license for prosecutors to go after anybody.
Just FYI.
DRAM?!!!! You benighted savages are still using DRAM?!!!!
I bought a Dell in 2002, I may have to dig up the papers to see what brand of RAM is in there. Not that it'd make any difference, we wouldn't see our checks for 2 or 3 years, and it'd be just enough to buy half a stick. ;-)
-jls
Techno-pagan
When rambus hit the streets it was way too expensive, incompatible and a one man show. The price was only one factor, and allthough a major one, I think many regarded rambus like people regard intel itanium. An incompatible architecture that is way too expensive comparing to the competitors.
Also while processors have always been controlled by a mighty few, this has not been the case with memory. One company trying to push *all* the others away in one strike is bound to have problems.
Linux might push Windows away one day, but only because it has aproven architecture and an unbeatable price, otherwise it wouldn't stand a chance.
...when I saw this. I paid 200 Pounds for 1MB of RAM (two 512KB SIMMs) in 1990. I just paid 105 Pounds for 1GB of RAM (two 512MB DIMMs) a couple of weeks ago.
I'm not saying that price fixing shouldn't be punished but that comparison pretty much puts things in context. When it comes to putting together a PC, getting a decent amount of memory isn't as financially crippling as it once was.
"Accept that some days you are the pigeon, and some days you are the statue." - David Brent, Wernham Hogg
that America does some have some resemblance to dignity
but with Ken Lay and his friends and family still enjoying the billions of dollars he stole from you, i dont really hold too much faith, when the goverment is as corrupt as it is, along with buisness in USA i think nothing will change,
same reasons as Enron is still in buisness so is Worldcom
this is nothing more than adding "feel good" as in "we look as if we are trying to clean up buisness but in reality we dont do anything to those we catch anyway, so carry on ripping off our citizens they dont care"
Compared to say about 1994-1995 I believe. I clearly remember paying $308.00 for an 8 meg simm. That was 8 times more memory than my first pc had so it was worth it !
SCO is doing just about the same thing as Rambus, but with much less success. Participate in Linux/UNIX standards groups, but later claim to own those standards and begin suing everyone.
Who do you get to be an expert to tell you something's not obvious? The least insightful person you can find? -J Roberts
how much I paid for RAM, it must to have been enough.
So why is it that RDRAM still cost so uch more?
And while we are on the subject, if the RIAA sent out those checks for losing tha price fixing case, why do CDs still cost so much?
Or did they just have to knock a few pennies of the top?
Class action lawsuits are becoming my new favorite pasttime. Consider:
In the past month, apparently I've been involved in at least 3 class action lawsuits. Both my wife and I got checks for $13.86 from connecticut's part in suing the record labels over overpriced CDs. Both of us have gotten paperwork regarding whatever claims are against Microsoft and software purchased in the late 90s (couple window versions, offices, etc.). I just submitted something for a company who were apparently inflating their stock value (or something) while I owned a number of their shares. And I can't even recall doing anything to get involved in the lawsuit to begin with. That's the best part. Christmas in March. I love it.
So, when are the consumers going to sue and and how do I convince the authorities to go after Corsair, as that's the only memory I purchased in that timeframe?
Back in the day when ram was little doughnuts, and it cost its weight in gold...
Back in the day when punch cards were book marks...
Back in the day when minimum wage actually kept you alive...
Back in the day...
-Grump
Is it true that more people vote for the winner of American Idol, than vote for the president? -Ali G.
but I have good news... .. I just saved a bunch of money on my car insurance by switching to Geico!
Commodity DRAM memory has been around $0.10 / megabyte since 2002. I remember slashdot stories about $100 gigabytes at that time. Until Wintel breaks the 2GB/32-bit limit the core memory cost is not a major factor in personal computing. in contrast, flash memory has fallen from $1.00 to $0.25 that time period.
The price plateaus when a chip generation matures. The next 4x generation seems a bit delayed.
...this was back in 2001, and RAM is dirt cheap today. Win some, lose some - big deal.
Hell they wont do anything about the oil companys fixing prices so why mess with computer companys. You can tell me 4 Billion in profits is not over inflated prices.
It's hardly surprising to hear there's an investigation going on. I got RAM cheaper two years ago than it is today. I've wanted to buy 500Meg modules for several years now, but the prices never seem to come down.
I'm no longer in the standalone DRAM business, but I did spend about 20 years, there.
Much of the time, there's no money to be made. Much of the time, just about everyone runs at a loss. The industry is also cyclical, and sometimes there's good money to be made. It doesn't help that it takes serious time to build a chip, so build-to-order doesn't really work very well. There's also time involved in packaging chips into a usable form, especially because it may involve transportation to a remote site. This aspect may be key, later.
IF there is price-fixing involved, and I suspect that there really isn't, it's the general idea of flattening out the bottom of the price curve a little in the cycle. I suspect it's far more likely that memory makers have decided, "It's just not worth bringing memory to market for less than $xx.xx." Remember the thing about packaging? At some price point, it may be better to not even bother packaging chips. It may even be better to grind them back into sand. Each manufacturer has different costs, but they're all doing the same thing. I suspect that they all have different, but similar package/hold/destroy price points for their chips.
This might appear to be price-fixing, but isn't. It's simple economics.
Years back, I bought my Mom a pair of earrings made from defective 4Mbit chips I had worked on for 6 or 7 dollars. At the time, perfect chips were selling on the open market for $4.50.
the price fixing went on for a long time afterwards, long after the factories were rebuilt and at capacity. the ram was artificially stockpiled and the world was kept in a ram shortage = high prices.
be thankful they dont limit the supply again, debeers style.
I had an eprom programmer from the early 70's ... it was maxed out at three 4KB memory boards (about RAM 16 chips on each), each costing $3000. It took a while to find the microprocessor on that thing because it was a 18-pin 4-bit chip.
But, this being slashdot, I'm sure someone would top that.
HIV Crosses Species Barrier... into Muppets
Good ol' slashdot. I post it, and it's rejected. An AC posts, and it's accepted.
I'm starting to wonder why I bother...
I don't want to sell anything, buy anything, or process anything. I don't want to sell anything bought or processed...
..."DAMN Price Fixing Investigations"
I mean, I guess I hate it too when that stupid damn FTC investigates those harmless, lovable monopolies for price-fixing.
I know I paid too much for RAM in 2001-2002. But what can we do now?
Even if the case went the way of the consumers two years later, the damage has already been done.
I'd like to see the price of a good steak go down. Cows seem so simple, I wonder why meat costs so much. A cheap calculator is far more complicated to manufacture. Yet the prices are very similar.
[insert your nonsensical comparison here]
This issue is a bit more complicated than you think.
If you think maybe 200 dollars for a peice of actual hardware is bad, they might want to start looking into software like windows trying to sell $1,000 for microsoft windows xp. Atleast you can actually hold the ram in your hand and be like "yeah! I have ram in my hand."
Additionally, the FTC ruled that Rambus was not guilty of trying to illegally monopolize the memory market. The decision, along with a recent court victory against Infineon, potentially clears the way for Rambus to seek royalties from memory makers. The royalty claims could ultimately reach past $1 billion, according to various estimates.
So the result is that prices of memory will go up. In the end, the consumer loses.
According to allegations made during the antitrust suit, Rambus memory was only expensive because the DRAM manufacturers conspired to make it so. Apparently they wanted to drive Rambus out of business in order to get its IP cheap. Hence the price fixing investigation.
I knew it when I paid $20 for a half a gig! That DRAM couldn't be worth more than $17, tops.
B
"I'm payin' taxes, but what am I buyin'?" -- James Brown
Well, since you say if you price your product below cost you can be accused of predatory pricing, the memory makers price fixing would be predatory price fixing since they were all still losing $$$ hand over fist even with this supposed price fixing. It might be against the law, but seems to me companies should be allowed to say, "hey, why don't we stop giving our product away with a dollar bill wrapped around each one..."
Actually, price-fixing and collaboration happens fairly often in the gas industry... it's just not caught/deal-with as often. However, the concept of price-fixing Vs price-ways happens quite often at the pumps.
Locally, we had some of the cheapest gas in probably the entire province (Canada). Longtime gas-stations kept trying to raise the price, bumping to the 70-80c range. However, two of the newer stations in town - both attached to grocery outlets - consistently stayed 8-10c+ under the competetion.
Least to say, the pumps that went up weren't getting much business, and eventually had to drop. This went on for some time, until the newer stations (at Safeway and the "Superstore") finally went with the bump and jacked their prices up.
It makes me wonder though, how often this type of thing happens in the gas industry. I've heard from employees about how the gas-stations will try to set a mutually-profitable higher prices, so maybe it just too awhile for them to get the "new boys" into the loop.
I don't know about 2001 and 2002, but RAM pricing still sucks in 2004. I just paid $85 for a 512mb stick of PC-3200, and $75 for an 80GB, 7200RPM, 8mb cache HD. Really quite sad.
The judge determined that the allegations against Rambus were fraudulent. Rambus circulated its patent portfolio among the DRAM manufacturers and made them aware of its intent to enforce it. The DRAM manufacturers then incorporated Rambus technology knowingly and conspired to get it without paying Rambus. They duped us into believing Rambus were the bad guys, when in fact Rambus was playing fair and they were attempting to steal Rambus' IP. If you read the antitrust judgement, it is very inequivocal in saying this.
A creator is granted copyright on something the moment that it is created regardless of whether the creator is a company or an individual or whether it took lots of money/effort or almost none to create.
Copyright is granted to give the creator a chance to make money from a creation, but the lack of such "return on investment" does not necessarily stifle creation..."Art for art's sake" and all that...
A bit more to the topic at hand, it does sometimes seem quite wrong that a copyright/patent holder can simultaneously price gouge the customer and prevent others from sometimes even mimicking the product and sell it at a lower price...I can understand patents on products that are quite expensive to develop, but as the grand-parent post said, razor refills and ink cartridges?...c'mon.
Actually...I guess I am somewhat split on the topic. On the one hand, as a creator (photographs and sometimes music) I can understand the desire to control how a creation is used, but on the other hand, as a consumer I would really like a cheaper product. There must be some kind of balance in there, right? Or must everyone be a profiteering glutton?
"Empathise with stupidity, and you're halfway to thinking like an idiot." - Iain M. Banks
Results were a little varied as 4-way Opteron systems seemed to fare the best, although dual Xeon configurations almost always beat dual Opterons.
:)
well, I prefere 2 machines with 1 Opteron than 1 machine with 2 Xeons
/ss
Wrongo. They are given patents because they apply for them, regardless of what it costs to make said product. Even if there IS no product. Or if it is just an idea.
Copyrights are something inherent, you don't have to be assigned a copyright.
Tell me, what happens once that initial investment has been recouped? Nothing.
Damn, this slope is getting very slippery....
My beliefs do not require that you agree with them.
Bah, what are you, 12?! I remember shelling out $17,000 big ones for a 1 byte expansioon card! You should be ashamed of yourselves! Why, in my day, we didn't even think in the kilobyte! Kilo-what? See, told you!
Don't call me a cowboy, and don't tell me to slow down!
We all understand why patents are necessary... so you're 100% correect but there's a big problem with the law
If you don't believe me, then look at the profits HP makes on selling printer refills compared to _all_ of it's other business wings combined. (over half the total revenue). The fact that all printer manufactors engage in the same policy can be regarded as a new way around the price fixing problem.
Lets face it, the industry is deliberately vendor locking their customers and then charging ridiculous prices. Mum and Dad get sold on a wonderful printer that costs only $150, but then sigh in resignation when the salesman tells them that _all_ the ink cartridges are very expensive.
So they get around price fixing charges by all producing different (but functionally identical) components and over charging for them. Seems like the price fixing laws need to be fixed.
Like all pain, suffering is a signal that something isn't right
I buy a new printer whenever I need ink again, and donate the old one to a school or charity as a tax write off. Thus far, it has been a much cheaper per page cost, not counting the tax value. Remember, many organizations let you say how much the item is worth on donation.
You can have it fast, accurate, or pretty. Pick any 2.
If you thought you paid too much for RAM in 2002, chances are you may have been more right than you originally thought."
If you had followed the major players in the DRAM industry, you will know that in 2002, not a single member of those top 4 memory makers made a profit. Every one of those 4 lost billions in combined net losses.
Although they might/probably did price fixing, it was still below the actual production cost of the parts. It just wasn't as low as South Korea was hoping Hynix could drive it (to kill Micron, probably), using IMF-funded bailouts of Hynix to keep it going.
Think of it like this (using the airlines analogies floating throughout today):
Southwest drops prices for a given flight $10, all other major airlines follow suit, and everyone starts losing $3 per ticket (from $7 profit before the drop). Southwest drops another $50. Everyone is losing money already, but Southwest is getting "debt" written off left and right by the US government and can "afford" it, while the others cannot.
The other airlines all get together and say, we'll drop only $5 at most, and let Southwest go belly up.
These competitors are still losing money, just not as much as if they tried to follow Southwest's pricing.
~AC, since I work for one of the big 4 memory makers.
Is that why to refill the cart Canon charges $10 for, I can buy ink from a 3rd party for about a buck a refill, and that's in small quantities? If I buy by the gallon it's even better; $85 per gallon, 15ml per cart, that's 252 refills, or 33 cents per refill. If I wanted to buy by the drum it'd probably get a little better.
Similar economics exist for the HP and Epson printers I've owned.
And, the ink is just as good, if not better. I've done both color comparisons and long-term (2 years in sunlight) fade comparisons. In fact the 3rd party stuff is better in many respects.
I bought DDR333 RAM in Jan last year, when it was significantly higher than it was in 2002, and it was back down below half that a few months later. Did they do the same thing here, or was that just bad luck for me?
I haven't read the decision. I started to, then decided it would raise my blood pressure too much. I've had over 20 years in the DRAM business. (I have another comment on this same topic that began at anonymous and made it to +5) I have also worked as a Rambus licensee, so I've seen and worked with the Rambus IP, *as delivered by Rambus*. Heck, I've been to their headquarters as a guest during the design process. I've had materials subpoenaed during the subsequent mess. I've read several of the Rambus patents, including the seminal 1990 one. Obviously I have to remain anonymous.
I will not begrudge Rambus in any way the material in their original 1990 patent application. (btw, it was not granted, but languished (or submarined?) in the Office until being superceded by the 1992 extension filing.) It describes a DRAM talking over a packet-based bus with data on both clock edges, and techniques such as the DLL for pushing the frequency up.
I will begrudge the subsequent filings, each based off of the 1992 filing, each of which 'drifted' to look more and more like SDRAM and DDR. IMHO, those derivatives stretch the original filing beyond recognition.
For one, DDR: The true seminal art in DDR DRAM lies in IBM's 1987 patent for 'toggle mode', where they deliver data on both edges of a signal called 'Toggle'. Rambus decided to call this a 'clock' instead of 'toggle', and claim DDR.
For two, CAS latency: IMHO this one was just plain 'obvious'. I worked on a 16Mb SDRAM, clearly on the early side, before things really took off. We had a CAS latency register, it came to us from the JEDEC spec. But it's a no-brainer that something of the sort will be needed, once you start clocking the beast and accept that clocks will ramp upwards. Here comes another point: Rambus used the programmable CAS latency for a completely different purpose than SDRAM. Rambus programmed CAS latency to adapt their whole memory system to it's configured size. (more about this, later) SDRAM programs CAS latency to adapt the performance of the part to the desired clock frequency, independent of the bus size. I'll agree that clock frequency *may* be a function of bus size, but that's not what's happening, here.
Oh, I don't know much about earlier Rambus designs, but I worked on the vintage of RDRAM that Intel pushed. Here's a good one:
The fundamental operation of Intel's RDRAM is different from the 1990 patent application that's the source of the current mess. The 1990 application tries to compensate for the physical size of the bus by making the whole thing look like it's "at the middle" of the bus. The Intel RDRAM allows each part to adapt to it's position along the bus. IMHO, the latter method is MUCH better, but it's MUCH different from the 1990 filing. AFAIK they haven't done nearly as much to protect their GOOD art as they have to submarine SDRAM and DDR.
But as a designer, neither I nor any of my co-workers ever looked at Rambus art and said, "This is neat! What can we pick up?" As a matter of fact, in the early 90's I said, "Why the heck is it that companies won't let their engineers do good engineering, yet they'll pay money to Rambus for just that?" At the time, we could have done all of those things, but NOOOOOOOO, we weren't allowed to specify terminated nets, continuous clocking, controlled impedence, and all of those "Rambus things", no matter how much we pushed for them.
Stations have looked at it. Turns out the people who use the pay at the pump are the people who would only buy gas anyway (often with the same credit /debit card that skims ~3% off the sale price), while those who buy the stuff inside go inside and buy it anyway.
With pay at the pump the don't need a clerk ($) to ring up sales for those who are only buying gas. Clerks costs money, if you can get by on one less clerk because of pay at the pump you are saving 5 bucks and hour. That adds up fast.
lets face it man, i only come here for entertainment. one cannot take this place seriously for anything important.
take infosec for example. 50% of slashdot users cannot tell the difference between a worm and a virus. cannot tell the difference between an exploit and stupidity.
why bother indeed. go AC and just use this place for what it is worth. entertainment.
tell that fat cowboy everybody seems to know of to take a shower. tell low uid slashdot users they really do need a life. One cannot post 10 times a day and try to pretend they dont have a problem. in other words just abuse this place. it will soon die off into obscurity. matter of fact, the smart people i meet online hate slashdot with a passion. wonder why eh? i dont.
My take on this is twofold. One, they charged so much because demand was high. Two, prices were high because of short supply. One could also argue both points as being true. Price fixing is an awfully big charge to throw around lightly, let us hope that there is evidence to back up the claims made in this article by the EU and the DOJ.
I hate sigs.
Why not go after some big, international price-fixing cartels. Like OPEC?
== Jez ==
Do you miss Firefox? Try Pale Moon.
I bought a 256MB stick for $10 after rebate.
Just do what the canadian goverment did. When gas prices where WAY too high they brought in there own chain of gas stations "Petro Canada". There actually goverment ran and force the others to keep a low price.
Just my idea
Solosoft.org - Your Online Resource to Nothing
I remember getting my dad to buy two extra 16K (not 16M, but 16K!) boards for my Atari 800 back in 1979 or 1980 since it came with only 16K. They cost $139 each, and that was a pretty good deal compared to what a friend with an Apple II had his parents pay to upgrade from 16K to 48K a year before that!!