Unstoppable: DDR400 vs. Rambus

[fkloster]

""Gratiutous" implys that it is unjustified. Take a look at what Fkloster has posted. THOSE are gratuitous insults which warrant a like response, in my opinion." -Zakule

"Ah, I guess 6th period just ended, huh!" -Zakule

*...insult by Zakule speculating that fkloster must be a young school kid*

"See you tomorrow, my overzealous, under-educated adversary!* -Zakule

*...insult by Zakule speculating about fkloster's lack of education*

Zakule, here are examples of just 'two' of many unprovoked, gratiutious insults typed by you....

From the very beginning of this thread, all I have asked of you is that you provide support for your statements, which you have denied the anandtech community. Instead, you have made multiple, MINOR attacks of my charachter, while insisting 'I' do the benchmarking work to prove you wrong....and that is what I intend to do.




<< Man, where does this sense of entitlement come from? I have nothing to prove to anyone on this thread, >>

-Zakule

Ok, I don't know about you, but where I come from, when a man makes a statement or provides a point of view, as YOU have done countless times in this thread..... YOU HAD BETTER BE ABLE TO BACK IT UP or you end up looking like a fool.



<< ...I'm offering information just as everyone else is. Whether or not it is received as credible or not doesn't matter to me and why should it? >>

-Zakule

Is it just me or can this statement be interpreted that Zakule doesn't care what people think of statements he presents and also that Zakule doesn't understand why anyone WOULD care? I care a great deal about position or statements I make and I do care what people think.






-----------------------------------------------------------------------------------------------------------------------------------------------------







Zakule in Las Vegas:

Zakule: "Put 500 on red dealer"

Dealer spins the wheel, which ends up resting on black...when the dealer reaches for Zakule's chips he finds none there, & Zakule screams...

Zakule: "Hey what are you doing? I don't have any chips and I shouldn't need them....now ...can I bet again?"

Dealer: "You lost...the marble landed on black, you betted on red..."

Zakule interupts: "WHAT!!!! you EXPECT ME to BACK UP WHAT I SAY WITH COLD HARD REALITY?"

Bouncers escort Zakule down to basement for some reality lessons...







-----------------------------------------------------------------------------------------------------------------------------------------








Moral of the story is, Zakule, when you make a statement, and someone calls you on it...you had better be able to back it up...

YOU MADE THE STATEMENT (..."any dx8 3d app will run faster on ddr than on rdram PERIOD" -Zakule circa 5/6/02)

now

YOU BACK IT UP

(...won't matter though...it has come to the point now where I personally will do the work you should have done following your statements to prove, or in your case, dis-prove your statement...)

 

[Athlon4all]

I cannot believe you 2. Why can't you 2 just cut the personal remarks out and simply discuss computer hardware?

<< FRAPS is not a synthetic benchmark. It takes reads on frames per second while in 3D apps and displays them in the upper left-hand corner. Not 100% accurate in all cases, but fairly reliable. Results here mean much more than 3DMark2001 since it allows you to test a variety of Directx8.0 engines, whereas 3DMark2001 is uses only ONE directx8 engine. It's a simple, but effective utility that even the greenest computer tech can use. >>

Maybe. My point was that if it and the game benchmarks itself say 2 different things, then I'll say that FRAPS is no better than a synthetic. Because, in case you haven't seen any real world gaming benchmarks, PC1066 RDRAM beats DDR pretty much always, I have yet to see PC1066 be beat.

 

[Chad]

Zakule,

For someone that is claiming he is being called names, I've noticed about 70% of each and every one of your posts is an insult to your debater. Seems like you're the one who needs to do "the growing up".

 

[smp]

<referee>

I think Fkloster started it!


lol Zakule...for your 5th post @ anands....

That was a personal attack, and the first one in the thread.
Followed by;

lol DDR2? what the hell? How long has DDR2 been out and can you provide a link...oh....oh ya, I just remembered, you don't have any links to your outlandish statements....how on earth could ANY ONE speculate intelligently that DDR2 (if and when DDR 2 is even developed...lol) would 'bury' 32bit PC-1200 in performance????? It seems very likely that you might now know what the hell you are talking about

It didn't have to be put like that, where I come from, talk like that is rude.

I too would like to see some proof, but as far as I can see, a lot of this thread is just speculation from both parties. I say that time will tell and I am satisfied leaving it at that.


</referee>

 

[fkloster]

<< ...lol Zakule...for your 5th post @ anands.... >>



Oh my Lord, If I were going to visit a very, VERY large hardware forum (like this one) and make a few posts (for the first tiime) that are very contraversal and not back them up, I would not be surprised if people scrutinized me... I WOULD EXPECT IT AND DESERVE IT Its not like I was ripping on him in the beginning, I just found it a rather shocking for someone with such outlandish statements to also be so new around here...



<< lol DDR2? what the hell? How long has DDR2 been out and can you provide a link... >>



Most of my frustrations stem from Zakules unwillingness to back up anything he says so I can give him credit. In addition, I feel I have maintained composure for the most part

TESTS WILL BEGIN THIS EVENING

 

[Zakule]

-sigh-

Try though I might, I just can't stay away from this thread. You've all intrigued me to no end, so I'm just going to ride this one out to its conclusion, whatever that may be.

I've got a quote, here, from somebody whom I consider to be reliable source about RDRAM as it relates to NT6. Before I tell you who it is (because some of you will undoubtedly know him) I'm curious to see which of you will contest the validity of this information.



<< NT6 is going to render RDRAM systems almost unusable. MS will do their best to make the performance acceptable, but its gonna be difficult. NT6 will be built around a database file system, which will offer incredible advancements in what the OS is capable of, but like all databases, it will require quick multiple accesses to several areas in the file system, 'streaming' large amounts of data into memory will be almost non-existant and this will put RDRAM at a serious disadvantage as its bandwidth advantages do not become apparant unless you send large, contiguous amounts of data to and from the memory. RDRAM has a lot of trouble with the 'random access' part of RAM. ;> It is also serial in nature, which limits how many different pieces of data can be fetched from it at once, a limitation that parallel memory does not have. >>



What say you now?

(Incidentally, I've enjoyed this thread immensely and I want to thank those of you who've been the most responsive to my highly controversial post. Especially you, Fkloster. I swear, there were a few times in this thread where I could feel heat from the steam rising off of your forehead! Trust me, I know how you feel. Some new punk ass comes along and sounds like a know-it-all and you don't like it, especially when his information contradicts your previous findings. Been there, done that.

For those of you who aren't familiar with me, my acerbic manner serves only to incite people into debate. You see, it is during these moments when people are most compelled to share what they know. I've found it to be an effective method for breaking the ice and motivating individuals to volunteer their knowledge, even though they're not particularly fond of me afterwards. Ah, well. I can live with it )



<< Let me just restate something, I am not questioning the fact that RDRAM does inheriantly have in some cases higher latency that DDR, but as the clock speed increases, this latency decreases that is what I am trying to get across, that and that PC1066 has similar latency to DDR. >>



Similar.......hmmmmmmm.........not sure I could even agree with that. In most real world environments RDRAM has higher latency, contrary to what some of the recent benchmarks tell you.



<< What are the odds of our 20,000 member forum having the exact system and games as me but using i845 ddr? >>



-sigh-

I swear, with how much some of you loath me one would think that you'd actually read my earlier posts. Intel P4 chipsets DDR chipsets will come in second to Intel P4 RDRAM chipsets because Intel intentionally cripples the performance of their DDR chipsets so that RDRAM will come out on top (and, no, I don't have a link that will back me up on this, but it is true). If you're going to do a comparision you need to use an SiS DDR chipset vs. Intel RDRAM chipset. Otherwise, there's no point.

 

[Chad]

Zakule,

If you want respect, you will have to earn it. I don't know how it's done where you come from, but around here at Anand's... in order to get respect, you must learn to give respect.

You were challenged because you were introducing claims without any evidence or support, and the manner in which you were making them was not condusive to an intellectual debate. Your posts are hostile. You will eventually learn the ways of AnandTech and can fit in.

These are just some of the groundrules.

 

[Zakule]

<< You were challenged because you were introducing claims without any evidence or support, and the manner in which you were making them was not condusive to an intellectual debate. Your posts are hostile. You will eventually learn the ways of AnandTech and can fit in. >>



Aside from your irrelevant commentary on my character, this has been an intellectual debate.



<< in order to get respect, you must learn to give respect. >>



Despite my tone, I have some measure of respect for everyone on this forum - even, you! I just don't take myself, or anyone else, as seriously as some of the others on this thread. Youthful arrogance on my part? Perhaps. Am I jerk? Yes, I would say so. Regardless of any of that, I have a mountain of technical knowledge that I'm more than happy to share with all here at Anandtech. Respect me, or don't respect me, but the information I'm posting is valid and accurate.

 

[Chad]

Or so you keep telling us.

 

[Zakule]

<< Or so you keep telling us. >>



My god you've figured me out! Damn! Guess it was only a matter of time before you put two and two together. I mean, it is certainly in my best interest to spend my valuable time spreading misinformation. Yep, that's what I do. Made a whole career out of it, I have. There's big bucks in misinformation, ya know.

 

[MadRat]

Why don't you do like you said and just bow out? You've done nothing but troll this entire thread.

 

[FelixDeCat]

Just check the sig....

 

[Zakule]

<< Why don't you do like you said and just bow out? You've done nothing but troll this entire thread. >>



Because I want to see what happens next!

Trolling? Please. I've mixed in plenty of relevant information with my insults!

I'm the underdog, here, not the undertroll.

Besides, if I left then you would have nobody to direct your self-righteousness towards. You don't really want that, now, do you?

Funny that nobody has commented on my quote regarding RDRAM and NT6........

Could it be that some of you are starting to realize that I'm telling the truth? Nah, that can't be it, because you'd then have to admit that you might be wrong. What's the matter, people? I've already told you all that I'm ready to eat my own words if time tells us that I've been mistaken, can the rest of you say the same? I wonder......

 

[BDSM]

hello again thread

Athlon4all.. Ofcourse latency is reduced when you up the working frequency. And this is ofcourse true for both RDRAM and DDR. Some argue that latency will somehow magically increase with the move to DDR 400. This is not true. It's just that the current DDR400 modules are unable to run with the same agressive timings as good DDR333 modules can today.
But as I've said more than a few times already, I've seen pics of DDR400 cas 2 modules, which is at least a step in the right direction.
And we know some of the really good DDR 333 modules out there can run with agressive timings at about 185-190 mhz.. An indication of what is to come.

Moving from 800 to 1066 mhz is a 33% increase. Moving from 133 to 166 mhz (pc 2100 to DDR 333) is only a 25% increase.. So latency wise RDRAm should come closer to DDR levels. Though I believe that DDR will have a lower latency until RDRAM runs at the same frequency.
That is real in real megahertz, not datarate. That would be in PC1333 RDRAM. I believe that is pretty far away though.

This is from Hardocp's p4 2.53 Ghz review:

<< In closing, let me say this. Remember that true PC1066 RDRAM is far and few between right now. There are a few companies that claim to have it, but no one seems to be showing their cards right now. That, of course, tells me that there are some production problems somewhere. This could be a very bad thing for Intel, as they tout their i850E chipset that you can't buy memory for. (Although the board does fully support the PC800 spec and RDRAM can be readily purchased for that level of usage.) Much of the PC800 RDRAM we have had here will operate at PC1066 speeds, but don't think you are guaranteed such. >>



He also says:

<< Dually DDR channels is what the Pentium 4 needs. Intel knows it, VIA knows it, and SiS knows it. And pretty soon, all of you will know it too.) >>



HardOCP review: link


I know it.. and you will too, soon ...Intel needs something to make the p4 able to compete with the Hammer. That's why they are going for dual channel DDR 333.

nite

 

[Zakule]

<< Most of my frustrations stem from Zakules unwillingness to back up anything he says so I can give him credit. In addition, I feel I have maintained composure for the most part >>



I can see how that might be frustrating, and I'd say you're doing fairly well in the composure department.


FYI: I've just confirmed through one of my sources that Intel still has no intention of officially supporting PC1066 on the i850e chipset. It is entirely up to mainboard manufacturers to implement and support PC1066 with the i850e.

Guess this is hardly news at this point, though. I just noticed that BSDM posted it earlier. Didn't know it was all over the tech sites already.

I think I'll give the ASUS P4T533-C a whirl with PC1066 to see how it works out. If anybody can make it work the way it should, it's ASUS.

Incidentally, BSDM, some of the mainboard manufacturers are supporting USB 2.0 along with PC1066 RDRAM, but it is entirely the responsibility of the manufacturer to implement USB 2.0, as it is with PC1066 RDRAM. My guess is that at one or two of 'em will do a good job on it. However, I'm betting that the SiS645DX will take the cake. I'd put down good money that the P4S533 with some solid DDR333 will run directx 8 apps (real world apps, not benchmarks like 3DMark2001 that love to show us what hardware T&L can do. You could probably count on one hand the number of current game titles that use hardware T&L, and it's going to get used less and less as the months roll by. Most game designers won't even touch it) with a more consistent framerate than P4T533-C systems.

Nite, all. See you tomorrow if I have the time.

The pariah has left the building! :Q

 

[MadRat]

<<FYI: I've just confirmed through one of my sources that Intel still has no intention of officially supporting PC1066 on the i850e chipset. It is entirely up to mainboard manufacturers to implement and support PC1066 with the i850e.>>

Your vapor sources again, tsk tsk. Then you conveniently claim to have known before it was all over the news sites...

 

[Chad]

Heh, indeed. He keeps telling us all about it and how he's right, yet I've never seen a link to any credible site, and haven't seen anything other realy than his incessant rambling about how knowledgable he is.

Please, teach us stupid AnandTech'ers... we don't know squat about "TECHNOLOGY" around these parts. Thankfully, you're kind enough to educate us all with your words.

Like the other guy said... around here at Anand's, it's all BS until you can prove it.

::Still waiting::

 

[Athlon4all]

<< FYI: I've just confirmed through one of my sources that Intel still has no intention of officially supporting PC1066 on the i850e chipset. It is entirely up to mainboard manufacturers to implement and support PC1066 with the i850e.
>>

We already go through that. Even though 850e may not "offically" support, the support will still be there on every board (even Intel's own D850eMV2:Q)

<< Incidentally, BSDM, some of the mainboard manufacturers are supporting USB 2.0 along with PC1066 RDRAM, but it is entirely the responsibility of the manufacturer to implement USB 2.0, as it is with PC1066 RDRAM. My guess is that at one or two of 'em will do a good job on it. However, I'm betting that the SiS645DX will take the cake. I'd put down good money that the P4S533 with some solid DDR333 will run directx 8 apps (real world apps, not benchmarks like 3DMark2001 that love to show us what hardware T&L can do. You could probably count on one hand the number of current game titles that use hardware T&L, and it's going to get used less and less as the months roll by. Most game designers won't even touch it) with a more consistent framerate than P4T533-C systems. >>

Don't take this the wrong way Zakule, You say 645DX, I have just one question, are you going to then discredit Oldfart's 845-D DDR400 numbers? Or should we find somebody else to do the DDR testing to compared to Fkloster's RDRAM numbers? I wouldn't bet anything just yet, but in any case, I look forward to seeing the numbers

<< This is from Hardocp's p4 2.53 Ghz review:

<< In closing, let me say this. Remember that true PC1066 RDRAM is far and few between right now. There are a few companies that claim to have it, but no one seems to be showing their cards right now. That, of course, tells me that there are some production problems somewhere. This could be a very bad thing for Intel, as they tout their i850E chipset that you can't buy memory for. (Although the board does fully support the PC800 spec and RDRAM can be readily purchased for that level of usage.) Much of the PC800 RDRAM we have had here will operate at PC1066 speeds, but don't think you are guaranteed such. >>



He also says:

<< Dually DDR channels is what the Pentium 4 needs. Intel knows it, VIA knows it, and SiS knows it. And pretty soon, all of you will know it too.) >> >>

Do you want me to go through it again? I personally feel that Springdale/Granite Bay will be more expensive than PC1066 RDRAM while providing similar performance if PC1200 is used. That's my opinion, you and Kyle have yours

<< << NT6 is going to render RDRAM systems almost unusable. MS will do their best to make the performance acceptable, but its gonna be difficult. NT6 will be built around a database file system, which will offer incredible advancements in what the OS is capable of, but like all databases, it will require quick multiple accesses to several areas in the file system, 'streaming' large amounts of data into memory will be almost non-existant and this will put RDRAM at a serious disadvantage as its bandwidth advantages do not become apparant unless you send large, contiguous amounts of data to and from the memory. RDRAM has a lot of trouble with the 'random access' part of RAM. ;> It is also serial in nature, which limits how many different pieces of data can be fetched from it at once, a limitation that parallel memory does not have. >> >>

Interesting. By NT6 you mean LongHorn? If this info is true, then you're right, it would hurt RDRAM significantley, but I am not so sure you could trust your source. AFAIK, LongHorn is barely in the development stage, there's no beta (or even Alpha) versions out yet. So, I guess only time will tell.

 

[trueblue]

Wow you guys are going at it hammer and tong lol

Just to throw something into this great debate There is an article from the May 2002 issue of Australin PC World page 18> Not sure how to put on as a link as can not find article on web but it is in the latest magazine.

Title is Intel is cools on Rambus. It goes on to explain how Intel threw it's support behind the rambus concept then states that the excutives of Intel declined to confirm whether future chipsets would support the high speed memory interface technology, saying Rambus is NOT the current first choice of consumers.

Intel CEO Craig BARRETT said the company will continue to manufacture existing chipsets for the PC and workstation markets that support RDRAM but stopped short of saying future chipsets in this sector will support the beleaguered memory technology.

BARRETT stated "We are obviuosly supporting RDRAM today with our existing chipsets and the network
processors continue to support RDRAM going forward and the rest of the products, you'll just have to wait until until they are announced" he said.

Further to this I saw a small Article in a PC Gaming Magazine for May 2002 claiming Intel was officially dropping support for RAMBUS. I can't recall which one it was but next trip to a newsagency I will post the guts of the article.

Have to say it is a bit strange that articles like these are starting to come up regarding Rambus. Maybe where there is smoke there is fire?????Text

 

[fkloster]

Hey madrat...you been around here a long time...for the life of me I can't remember his username, but do you remember the guy who was @ anands for a while who 'said' he ran an engineering firm in Sweden and owned sports cars and ripped on people ect. ect. ect.?

oh well...back on subject...

1) I removed 2 of my 256mb rimms last night and downloaded this frap ute Zakule talks about and ran it in a shell.

2) Dungeon Siege in 800x600x32 all candy on floats anywhere from 25fps to 40fps depending on type of fight and atmosphere

3) Max Payne was very difficult to determine average but floated around 100 to 150fps

4) Will do RtCW tonight

I have a feeling this is all for naught now that Zakule has 'condemned' Old Fart's Intelish DDR rig because it is 'crippled' so, in addition to an overwhelming feeling that I have been 'seduced' into this work in a 'trollish' way, I am out of here...better things to do than to provide entertainment for strangers...

 

[Athlon4all]

<< I have a feeling this is all for naught now that Zakule has 'condemned' Old Fart's Intelish DDR rig because it is 'crippled' so, in addition to an overwhelming feeling that I have been 'seduced' into this work in a 'trollish' way, I am out of here...better things to do than to provide entertainment for strangers... >>

Fkloster, Zakule never once said that he was condeming Oldfarts numbers, I AM THE ONE WHO ASKED IF HE WOULD STILL ACCEPT DDR400 ON 845-D EVEN THOUGH ITS NOT SIS 645DX, you are putting words in his mouth. Please stop

 

[fkloster]

"I swear, with how much some of you loath me one would think that you'd actually read my earlier posts. Intel P4 chipsets DDR chipsets will come in second to Intel P4 RDRAM chipsets because Intel intentionally cripples the performance of their DDR chipsets so that RDRAM will come out on top (and, no, I don't have a link that will back me up on this, but it is true). If you're going to do a comparision you need to use an SiS DDR chipset vs. Intel RDRAM chipset. Otherwise, there's no point. " -Zakule

1) Athlonforall, let me ask you something? Is it your impression, from the DIRECT quote above, that Zakule is saying that the Intel DDR chipsets are intentionally crippled by Intel?

2) Athlonforall, does Oldfart intend to test HIS Intel DDR chipset against my rdram system?

3) Athlonforall, does Oldfart own a DDR Intel chipset? If so, MAY ONE DEDUCE that Zakule would say that Intel has crippled his chipset and that my rambus platform will win no matter what?





---------------------------------------------------------------------------------------------------------------------------------------


Zakule shouts: "CAR MANUFACTURERS MAKE RED CARS SLOWER THAN BLUE CARS!!!!"

fkloster realizes OldFart owns a red car while fkloster drives a blue car

fkloster states: "Zakule must think OldFart's red car is slower than my blue car & a race between the two would result in the blue car winning..."

Athlonforall screams: "FKLOSTER DON'T PUT WORDS IN ZAKULES MOUTH...HE DIDN'T SAY THAT!!!!"

fkloster rolls on the floor laughing Excuse me for using common logic!

 

[Athlon4all]

<< Is it your impression, from the DIRECT quote above, that Zakule is saying that the Intel DDR chipsets are intentionally crippled by Intel? >>

Yes, and I agree that it isn't beyond Intel to do something like that.

<< does Oldfart intend to test HIS Intel DDR chipset against my rdram system? >>

It is my understand that yes, he is going to.

<< If so, MAY ONE DEDUCE that Zakule would say that Intel has crippled his chipset and that my rambus platform will win no matter what? >>

Maybe, maybe not. You don't know that. Neither do I. What bothers me is that you put words in his mouth. *sigh*. I'd better stop here because I feel like I am getting pulled into all this more than I already am

 

[smp]

Does anyone have the SIS chipset? Can we have a third party please?

I too think that Intel would be very capable of crippling their DDR chipset to make Rambus look better. Capitalism makes strange bedfellows

 

[andreasl]

Just wanted to jump into this thread and comment on this statement:



<< NT6 is going to render RDRAM systems almost unusable. MS will do their best to make the performance acceptable, but its gonna be difficult. NT6 will be built around a database file system, which will offer incredible advancements in what the OS is capable of, but like all databases, it will require quick multiple accesses to several areas in the file system, 'streaming' large amounts of data into memory will be almost non-existant and this will put RDRAM at a serious disadvantage as its bandwidth advantages do not become apparant unless you send large, contiguous amounts of data to and from the memory. RDRAM has a lot of trouble with the 'random access' part of RAM. ;> It is also serial in nature, which limits how many different pieces of data can be fetched from it at once, a limitation that parallel memory does not have. >>



I too have read about the database file system nature of the next file system coming from MS. However a statement like this shows a big lack of understanding about most things computer related... A disk file system is the way the data or organized and ordered on the harddrive. A harddrive has an access time several orders of magnitude longer than RAM. Consider a typical harddrive with an access time of 10ms (simplified, most have a little less these days). RAM has access times of about 100ns (also simplified, and I'm including chipset latencies as well). 10 milliseconds is 10.000.000 nanoseconds. So by the time the CPU waits for the harddrive access it can access RAM about 100.000 times! The difference in access time between different RAM types is very insignificant in this sense (about 10-20ns or so).

Streaming or non-streaming data from the harddrive will have NO impact on which type of RAM you use, but which harddrive you use. I think the original poster (not Zakule) who made that statement is confusing access times between RAM and harddrives. 2 completely different things alltogether. I would also like a link to the full article/post where it was stated but I doubt I'll get one...