Intel to drop support of RAMBUS in new products...

[ST4RCUTTER]

Link

Looks like Intel has had enough of RDRAM's acceptance, or lack there of, at least in the server space. Intel will not introduce new products utilizing RDRAM according to this EBN article. This will be a hard pill for RAMBUS to swallow.

 

[NFS4]

Well spank my ass and call me Charlie!:Q

 

[Athlon4all]

This sucks big time. Dual Channel DDR will be same speed as RDRAM if DDR266 is used, but will in all likely hood be more expensive than 850+PC1066, and definately will be more expensvie than the nixed Tulloch+32-bit RIMM platform:Q:disgust::|

 

[wfbberzerker]

ok charlie


i think its about damn time. rambus has been trouble for awhile now.

 

[AGodspeed]

And just when RDRAM was starting to get more interesting...

 

[ST4RCUTTER]

And just when RDRAM was starting to get more interesting...


I'm surprised by this as well. Makes me wonder if Intel sees more promise in alternative memory types...a.k.a. DDRII.

 

[ElFenix]

somehow this is really old news. something about a tom's roadmap a long time ago. maybe even in the p3 days.

 

[JellyBaby]

In my best Monty Burns, "Excellent!"

 

[NFS4]

What has me scratching my head is that Anand had samples of PC1200 displayed in his IDF Day 2 article:Q

 

[Novgrod]

Thank gawd; about time.

Still can't bring myself to like a company that tried to screw over the memory industry.

 

[ST4RCUTTER]

What has me scratching my head is that Anand had samples of PC1200 displayed in his IDF Day 2 article

As the article points out, this has more to do with server and workstation boxes than desktops. We know that RDRAM will scale up to 533 FSB. After that though dual channel DDR is likely to supplant it. I'm guessing that the price and availability of PC1200 will likely be insufficient for mass production. Dual channel DDR, even though it costs more for motherboard manufacturers, may end up being cheaper.

 

[Dre]

Wierd that the above website is the only one I can find that is reporting this. Maybe it's not true. If it were true, I'd think it be all over the news.

 

[Pabster]

Why are you all so shocked?

RAMBUSt has been a big BUST from day one. Besides the fact that they tried to claim patents they didn't own (or have any rights to) and extort millions of dollars in royalties from every Tom, Dick, and Harry even remotely affiliated with said ... The technology itself just isn't anything to scream about. It is proprietary, expensive, and I believe cost Intel a great deal of market share.

Just look at the screenshots of engineering samplings from PC1066 batches. They run so friggin' HOT that RAMBUSt actually recommends an 80mm cooling fan just to keep 2 RIMMs at feasible temps. Who the hell wants that?

DDR is the future, for better or for worse. Dropping a dead horse is actually a wise business decision for Intel.

 

[NFS4]

<< Just look at the screenshots of engineering samplings from PC1066 batches. They run so friggin' HOT that RAMBUSt actually recommends an 80mm cooling fan just to keep 2 RIMMs at feasible temps. Who the hell wants that? >>


Can't be much worse than the cooling needed for Athlon processors

 

[AGodspeed]

<<

<< Just look at the screenshots of engineering samplings from PC1066 batches. They run so friggin' HOT that RAMBUSt actually recommends an 80mm cooling fan just to keep 2 RIMMs at feasible temps. Who the hell wants that? >>


Can't be much worse than the cooling needed for Athlon processors >>

Have you seen a Pentium 4 heatsink? They're significantly larger. A heatspreader would do AMD good though...

 

[NFS4]

<<

<<

<< Just look at the screenshots of engineering samplings from PC1066 batches. They run so friggin' HOT that RAMBUSt actually recommends an 80mm cooling fan just to keep 2 RIMMs at feasible temps. Who the hell wants that? >>


Can't be much worse than the cooling needed for Athlon processors >>

Have you seen a Pentium 4 heatsink? They're significantly larger. A heatspreader would do AMD good though... >>


I wouldn't mind a large heatsink to get rid of my jet engine fan on my SK6. And yes, I was astonished when I first installed a tiny Socket 478 processor and then saw how big the heatsink was. But I quickly shut my mouth when I heard how queit it was.

 

[Pariah]

I don't think a P4 really needs a HS that large. I think the size of it has more to do with ease of installation and noise. I don't think there is any doubt that the clamp system on P4 HS is the smartest and easiest installation (and uninstallation if that was a word) method mass produced to date. AMD really should adopt a similar system as well as adding a heat spreader.

 

[WilsonTung]

Like it or not, RDRAM isn't going to dissapear.



<< DDR is the future, for better or for worse >>


Easy for you to say if you don't have to deal with increasingly difficult signaling issues. DDR is a dead end. My friends and profs in EE all say that crosstalk, interference, and other physical issues are hindering parallel busses. While CPU speeds increased from 500 - 1000 - 1500 MHz, SDRAMs went from 100 to 133 and 266 MHz. It's very hard to make current SDRAMs run as fast as they do.



<< The technology itself just isn't anything to scream about. It is proprietary, expensive... >>


Last time I checked PC800 RDRAM prices are on par with DDR prices.



<< They run so friggin' HOT that RAMBUSt actually recommends an 80mm cooling fan just to keep 2 RIMMs at feasible temps. Who the hell wants that >>


That's what Apple said about x86 processors. They were too hot compared to PPC and needed noisy high speed fans.

DDR-II is vaporware at this point. RDRAM is not. RDRAM costs roughly the same as DDR. PC1066 and 32-bit RDRAM are just around the corner (well, PC1066 is already here if you OC).

The bottom line: People are letting their dislike of Rambus' (admittedly despicable) actions cloud their judgement of the technology. Look at how many people claim RDRAM is so expensive, all the while forgetting that 5 posts ago they whined about DDR reaching 88 $US a stick.

 

[Insane3D]

"I don't think a P4 really needs a HS that large. I think the size of it has more to do with ease of installation and noise. I don't think there is any doubt that the clamp system on P4 HS is the smartest and easiest installation (and uninstallation if that was a word) method mass produced to date. AMD really should adopt a similar system as well as adding a heat spreader. "

I think the four screw mounting system, while harder to install, is superior. Sure the Intel HSF installation is a bit easier, but have you noticed how it bows out the motherboards PCB underneath the CPU? It doesn't seem to damage anything, but over time who knows. This has even led HSF makers like Taisol to release S478 cooling units with a plate to mount on the bottom of the board to prevent this. I believe some of the mobo manufacturers are doing the same... IIRC the Soyo P4 board has a reinforcing plate underneath the CPU socket too.


Edit:

Yep..here is a shot of the bottom of the Soyo P4S board.

Soyo P4S under CPU socket

 

[BD231]

I think the big issue is what Intel will use in place of RDRAM, DDR sucks IMO.

 

[Midnight Rambler]

<< I think the big issue is what Intel will use in place of RDRAM, DDR sucks IMO. >>

They have an in-house memory design cooking, the codename is IDREAM ...

 

[Darien]

Man this sucks.

AS someone mentioned above, DDR (at this point in time) is pretty much a dead end.

And, we've all seen the benches at aces right? with the pc1066...

Too bad about this.

Have a nice day

Darien

 

[sandorski]

Everything is a dead end, what's important is what will hold until something new comes out. RDdram's supposed technical superiority should have been suspect the moment P4 launched with the need for 2 sticks of RDdram.

 

[Dre]

Go to Tom's Hardware site and in the news section, they are reporting that the above atricle twisted the words around and Intl will continue with Rambus. Just as I thought.

 

[fkloster]

Pabster you are so full of it. DDR is the future? lol Serial memory is the future and rdram will always find its way in my box's...