Long-Delayed Rambus Machines May Show at Comdex

Sokie writes "IBM, H-P, Micron, and Dell all plan to unveil Rambus based machines at the upcoming Comdex in Las Vegas, NV according to C|Net News. Intel, however, won't confirm that the 820 chipset will be out by then. I can just imagine the heads that would roll if Intel postponed the release again, the OEMs are still probably a bit grumpy after last time."

Show 'em

[Microlith]

That would definitely be a good change of pace from the ground-headed course Intel has been on with the 820. I'm pretty sure that if they didn't bring them, many large OEMs would seriously be considering moving to Athlons, which support > 100mhz bus speeds, and haven't had any problems yet.

Crazy conspiracy theory...

[chazR]

==BEGIN CRAZY CONSPIRACY THEORY

Intel are deliberately delaying 820 in order to create an artificial demand for BX chipsets. Motherboard manufacturers then have to rely on Intel being nice and supplying them with lots of BX chipsets which weren't pre-ordered. Now, are Intel going to favour manufacturers who are also producing Athlon motherboards?

==END CRAZY CONSPIRACY THEORY


Of course, Intel wouldn't do anything so cynical.

Feed the hungry. Save the whales. Free the mallocs.

Marketing Wizards

[Accipiter]

Does anyone else find it incredible how Intel manages to push out Faster and Faster clocked processors (700 mHZ!), but when it comes to the important stuff (i.e.: Bus Speeds) they choke?

Ever since Intel stopped manufacturing processors and started manufacturing marketing gimmicks, they've been slipping behind giving other companies like AMD the space to edge into the market. If Intel spent half the money on technology that they do on marketing and/or developing faster processors, there would be far less setbacks, and we would have superior machines.

-- Give him Head? Be a Beacon?

Tied to Win2k?

[bgordon]

Maybe the delays are so that the RDRAM release will coincide with the Win2k release. That way Win2k won't seem so slow, since the first computers it runs on will have the faster RAM.

Rambus doesn't blow your mind

[Cplus]

According to a review I read in a hard copy of Maximumpc Mag, the Rambus does show performance improvements, but it won't really be taken advantage of until software is written to take full advantage. I'm really not that excited about it and can't understand what's taking so long. They released the i840 and they've released some to OEM's which have been passed along to the reviewers, so what's the hold-up baby!

More special instructions for less gen. purp. comp

[CBravo]

The idea of general purpose computing is falling down. When you want a game-computer, you get game-instructions. When you want multimedia, you get MultiMediaX10sions.

This means that real general purpose programming is _*relatively*_ becoming slower.

Since all computing is getting faster that is not a real problem. The second solution to that problem is connectivity. When you need the solution to a problem, why not beam it to a computer that can handle all this speciality for your specific problem (rent Cray-time, etc).

Re:Rambus, DDR-SDRAM and the Athlon....

[bofh23]

The big problem right now for AMD is the fact that Intel has their L2 cache integrated on the die, and will allow their processors to ramp up in speed without the degredation of L2 speed (which will occur with the athlon).

AMD needs to get on die L2 soon, the 1/2, 1/3 divider for processor memory is really bad news.

The Athlon may have off die L2 cache but it has a huge, 128KB on die L1 cache. Coppermine only has 256KB on die L2 cache.

My bad

[Cplus]

The review was on a Micron system and one other that I can't remember.

Re:Rambus doesn't blow your mind

[David Greene]

This has been true for a long, long time.

Writing a piece of code to run well on a Cray or Convex is not easy. You need to worry about all sorts of nasties like NUMA, messaging speed and so forth. Even non-parallel code must be concerned about I/O time. Databases have used their own raw partitions and caching schemes for years.

Even some PC software is written with the memory heirarchy in mind. You want your working set to fit in the cache, page table or whatever. RAMBUS just presents a new organization for which to optimize. Writing good code that takes advantage of the hardware makes a big difference. It's for this reason that programmers need to understand the hardware they're working with.

--

Re:Rambus doesn't blow your mind

[Nagash]

...but it won't really be taken advantage of until software is written to take full advantage

Does anyone else find this trend annoying? Having hardware that has all kinds of special instructions? OK, for CPUs, I suppose it is understandable, but for memory? Do you really mean software to take advantage of the i820/i840 chipsets?

Geoff Wozniak
[e-mail]

Rambus, DDR-SDRAM and the Athlon....

[Dextius Alphaeus]

DDR-SDRAM 's capable FSB speed = 266mhz

Athlon 's current FSB = 200mhz (to be raised to 266 once a chipset is created to support DDR)

RAMBUS 's capable FSB speed = 400mhz ? (there is a less powerfull version that will not allow 133mhz fsb speeds, which I don't understand)

Intel's current mobo FSB = 133mhz
Athlon's current mobo FSB = 200mhz (on the CPU bus, but can take full advantate of the 133mhz memory)

The big problem right now for AMD is the fact that Intel has their L2 cache integrated on the die, and will allow their processors to ramp up in speed without the degredation of L2 speed (which will occur with the athlon).

AMD needs to get on die L2 soon, the 1/2, 1/3 divider for processor memory is really bad news. The fact is, that without the "celeron" tested technology, the new coppermine processors wouldn't be near as fast as they AREN'T..

I own a p3-550 right now, I will not buy anything, until the hardware manufacturers get over this little war over memory (and those poor people in Taiwan get their country back in %100 order).

One last note... Asus not releasing Athlon mobo's.. what the heck more does the justice department need to nail intel with unfair business practices? This is ridiculous, AMD is being forced farther into the red because nobody is willing to step up to intel?!

What a sorry state of affairs we are in right now...

-Dextius Alphaeus

No the memory itself......

[Cplus]

Apparently the concept of running memory at this speed hasn't been accounted for in the past. Nothing suffers from it but it could be better apparently. I have no idea of the technical aspects of it but.............isn't this the way it always is.

New hardware requiring new software

[zilym]

I don't find this trend annoying at all, as long as the new hardware has enough technical documentation freely available for Linux developers to work with. I think the greatest gains are usually only possible if you don't have to maintain backwards compatibility with old software. As long as documentation is available, new hardware that increases performance and decreases cost can only help propel Linux (and other open source software) further ahead of the closed source competition.

Think about it. A lot of new and exciting things are now coming out due to the source code availability of Linux. Empeg [empeg.com] car MP3 players, Playstation 2, that Linux powered high speed ethernet switch, etc all come to mind. So what if new hardware requires tweaked software? It just makes it harder for Microsoft to keep up with us.

iXBT labs has an interesting article...

[leiz]

on RAMBUS, intel chipsets, ram manufactures, sdram, and all the other good stuff...

"i820 - the story of one failure" [ixbt-labs.com]


_______________________________________________
There is no statute of limitation on stupidity.

Forget your AMD fantasies - OEMs depend on Intel

[Anonymous Coward]

While its very pleasant to presume that OEMs will dump Intel and adopt AMD chips, you can basically forget it. No large scale manufacturer has made a significant commitment to AMD.

Unfortunately, Intel managed develop brand power with their incessant advetising, and this has created an impression among OEMs that consumers will respond to this and look for the "Intel Inside" sticker.

Whether they do or not, its the self-fulfilling prophecy of advertising. The OEMs are strongly dependent on Intel. Now more dependent than they were on Microsoft. They'll take whatever they're given with a smile.

Re:Rambus, DDR-SDRAM and the Athlon....

[Erik Corry]

DDR-SDRAM 's capable FSB speed = 266mhz

The FSB (Front Side Bus) is the bus the processor runs on. Since the RAM (incl. DDR-SDRAM) doesn't run on the the processor bus, it doesn't have an FSB.

DDR-SDRAM will have a 133MHz bus, at double data rate, hence 266Mhz.

RAMBUS 's capable FSB speed = 400mhz ? (there is a less powerfull version that will not allow 133mhz fsb speeds, which I don't understand)

Again, this is no FSB. Also it is a DDR interface, so it's 800MHz really. But it is only 16 bits wide. The other busses in this discussion have been 64 bits wide.

You can have multiple RAM busses making the RAM bus effectively wider. You can have only one GTL+ (P6/P-II/P-III) bus in a system whereas for EV6 (Athlon/Alpha 21264) you have one per processor. Rambus chips are at 300/600Mhz, 356/712Mhz or 400/800Mhz. Most benchmarks are for the 400/800Mhz version whereas most produced Rambus chips are for the lower speed grades. Rambus chipsets only support certain multipliers between the FSB and the Rambus bus, though I have forgotten the details.

Intel's current mobo FSB = 133mhz Yes. This is normal 133MHz (no DDR), though as far as I can see there are no SMP motherboards actually available at this speed. All the SMP motherboards actually sighted in the wild (including hardware test web sites) are 100MHz. Counterexamples welcome.

Athlon 's current FSB = 200mhz (to be raised to 266 once a chipset is created to support DDR)

Yes. This is a 100MHz or 133MHz bus, with DDR, making 200MHz or 266MHz.

If anyone can remember what the allowed relationships beween the FSB and the Rambus speed are for the i820 and i840 then I would be interested.

L2 Cache on AMD

[Atomix8]

Everyone knows that the Athlon is mad fast and is an excellent option for those on a budget. It is not as cheap as previous AMD processors because it is advanced and FAST. However, if AMD goes through the cost of developing a new chipset incorporating L2 right now, they will be so far into the red that they will sink like a guy with a lead shoes. AMD is staying narrowly afloat by getting by on a shoestring development budget which has worked so far because they have done nothing revolutionary. In the future though Intel may be able to pull ahead simply because it's market share allows them to charge so much more for a processor that is currently good, but it also supports a massive bloated development effort.

For the future of computer processors, something revolutionary must happen in the basic development. The little copper wires in chips can only be so small and close before they stop working. Intel and IBM are currently working to deveolop the future which is supported by the bloated profit margins of both companies. They will be the ones who develop tomorrows technology. When they do, AMD will make it slower but cheaper, and eventually develop the technology slightly further themselves.

As for memory prices they are infuriating, but what are you gonna do?? So your paying what you would have paid 2 years ago for the same memory? There is nothing to be done but wait, curse, and try to find some newbie with some extra memory, punch him in the balls and take IT!!!!!