Hot on the heels of the US Federal Trade Commission's decision two weeks ago to drop a seven-year antitrust lawsuit against Rambus, the memory chip designer is pitching its bag of technology tricks, collectively known as XDR2, as the future of main memory and graphics memory for PCs and servers. If you are a cynic, you might …
any company using lawyers to do the selling is an immediate fail. these steps are obvious changes. Partitioning is in use in every part of computing at the moment so I fail to see how that one is going to be considered revolutionary. Let's hope Rambus meet SCO in patent/lawsuit hell!
Rambus leaping in the wrong direction
Two ways to make memory access faster:
1) Put 4 CPU's in one chip, put the chip in a socket, connect the socket to another socket via a PCB, put another PCB in that socket and solder memory chips to that PCB. Then come up with some patented process to reduce the latency and power requirements of all those gigantic sockets and PCB's.
2) Put 1 CPU in a chip and use the remaining space for embedded DRAM (http://en.wikipedia.org/wiki/1T-SRAM, http://en.wikipedia.org/wiki/EDRAM). All the connections are low latency low power on chip busses.
Guess which is the best choice for small cheap computers?
Hey RAMBUS! Why don't you join an industry consortium and help everyone by getting your designs included in the next industry wide standard for memory modules!
I can't quite put my finger on it, but something tells me that they shouldn't be trusted!
First off, any AM3 AMD desktop chip will run DDR3 if you have a motherboard to support it. That's not terribly exciting, since performance is more or less the exact same with DDR2 as DDR3.
Second, that's a fun chart there for memory speed. RDRAM is missing, probably because it was a horrible disaster. Second, the speeds are for a 64bit memory module, which would discount speed increases from going from dual to triple to quad channel ram. That Mangy-Cours AMD deal is supposed to run quad channel DDR3, while the i7 / Xeon 55xx uses triple channel. Both would be far above that curve, even if you re-draw it with a baseline of dual channel, which everything uses.
I wish the DoJ would have burried Rambus, since that is clearly what they deserve. How did the world come to a state where it is more profitable to sue based on IP than sell products based on that IP? So sad...
AMD does support DDR3
AMD has supported DDR3 on AM3 motherboards for several months.
Oh not you again
Look, Rambus, we've been through this.
I know you were offering bandwidth of 21GB/s back in 2005, and I'm only just getting that on DDR3 now while my PS3 has had it the whole time.
But that 21GB/s is FUCK ALL USE.
I don't need data in packets of 2KB. I need it in cacheline-sized chunks. Like xDDRn gives me.
Yes, I know you give me that too, but at HUGELY degraded bandwidth. 1/8 of bandwidth when I'm accessing 128B at a time is not acceptable. Not every algorithm is specifically tuned for YDRAM.
And I don't need huge fucking latencies either. Great, you can queue commands! Don't use that as an excuse to take all year to turn them around.
Yes, DDR3 will need replacing in a few years. Shit, guys, you don't think it'll be replaced by DDR4 do you?
Yeah I still have a RAMBUS Module...
...from an Intel motherboard from the early noughties. Also, I remember a gushing article in The Economist ("Rambo RAMBUS" was it?). Apart from that, RAMBUS only evokes massive FAIL for me. Do I care about that company or its products? This would be like caring about "Scooter" Libby. Would make me feel unclean.
Rambus Is an Inovator
Or at least they have been so far. RDRAM had high performance for it's time, the problem was the high cost which made it a poor seller. But the technical side was good. So a lot of memory manufacturers borrowed a little bit here and a little bit there and incorporated that into their products. And that's why Rambus did a little better in court than SCO, other companies were using things which Rambus had invented and patented. And I really doubt that Rambus planned things to go that way.
I hope that they have something good. The article says nothing about the cost and that's probably not a good sign.
In honor of this article
I just shutdown my rambus-based Dell Precisions... For the last time. They go to the scrapheap tommorrow. I transferred their workloads to an IBM RS/6000 and a Sun Ultra 1 - and the funny thing? Both those systems have more RAM than the Precisions did. Could it possibly be because buying RDRAM would have fucking bankrupted me? Even when compared to ridiculous crazy proprietary nutzo-RAM? Yeah, I think that's it.
Re: Rambus Is an Inovator
The problem E_Nig is that Rambus joined a standards setting committee and steered everybody else in the industry towards their IP without disclosing their patents. This was a big no no as well as borderline fraud and why all the negativity towards the firm and their ambulance chasing legal team. Its sad but I have seen this in my own industry recently as well. A certain big clueless wreck of a company was behind the eight ball releasing a software product that used a new fancy standardized protocol so to cover their own ineptness they instead took out patents and didn't tell the standards body until after they legally threatened a smaller competitor. Modern idea-only (no implementation) patents are evil and like most of the institutionalized evil in the world today only exist because scum sucking lawyers are the ones who write the laws (the ultimate conflict of interest).
"Hey RAMBUS! Why don't you join an industry consortium and help everyone by getting your designs included in the next industry wide standard for memory modules!"
I really hope your being Ironic (use joke icon), otherwise you need to read up on the RAMBUS vs FTC case
That's a load of rubbish, rambus joined a standards organisation and then changed their
secret patents to cover fundamental aspects of that standard.
The only country that allows them to get away with this blatant act of dishonesty is the US
everyone else told them to take a running jump!
Paris becuase she knows a bunch of shysters when she sees them!
Re: Oh not you again
"I don't need data in packets of 2KB. I need it in cacheline-sized chunks."
No, you don't. You used to - back in the days when the world was grey and men walked quickly. These days, however, it is far better to design your data structures to be cache friendly and copy large amounts of data into the L2 cache, work on it there (where you will then get cache-line sized transfer to and from the L1 cache) and then copy the results back out to system memory. And, unsurprisingly, this is exactly how the SPUs on your PS3's Cell processor are designed to work. (And in fact it's the only way they work.)
Assuming the architecture will solve everything for you and stalling while you wait for the next cache line to be fetched is soooo last century. I bet you think letting threads run on any available core is still a good idea too.
Paris, because she likes large packets too.
re: FTC v. RAMBUS - But the FTC dropped their case which could only mean that RAMBUS is a pretty cool guy...
I really hope you're being Ironic, otherwise you need to have your sarcasm detector looked at! I reckon that RAMBUS only gets to pull that stunt once.
Of course the SPUs work that way. Cell and the PS3 were designed to be used in a system that had XDR memory.
"Cell and the PS3 were designed to be used in a system that had XDR memory."
I doubt the Cell was designed with that in mind. Forcing processor elements to only be allowed to access their local caches directly greatly simplifies the design of the processor. It makes a lot of sense from both power and cost perspectives.
The fact that RDRAM has such high bandwidth does make it the logical choice for Cell based systems with such a high SPU to PPC ratio, which could have been one of the guiding factors in the design of the PS3.