Sign in with
Sign up | Sign in
Your question

DDR1 vs RDRAM 16 bit

Last response: in CPUs
Share
May 1, 2002 3:52:21 PM

Mobo new set of intel chipset all very fast I845 with DDR 266 have beat sis 645DX with DDR 400.For me it seen clear that intel have boost as much as he can is DDR chipset i cannot say as much for is RDRAM chipset.

still no test of Rimm4200 also i like to point out for Pc1200 you need 150 quad pump FSB to reach top performance.

Any comment

cheap, cheap. Think cheap, and you'll always be cheap.AMD version of semi conducteur industrie

More about : ddr1 rdram bit

a b à CPUs
May 1, 2002 4:31:45 PM

Dude, Frank Volkel is an idiot. You can seen PC1066 clearly wasting DDR400, yet he only concludes that DDR400 beets PC800. I got news for everyone, Samsung PC800 does PC1066 speed, maintaining it's supremacy for P4's operating at "533", including current processors overclocked to that speed.

What's the frequency, Kenneth?
May 1, 2002 6:50:55 PM

The hick with tomshardware is not the benchmark itself but the little conclusion.

90% of hardware communoty are Pro AMD so they have tender to make better conclusion for AMD that Intel.Any website that start make pro AMD benchmark.

cheap, cheap. Think cheap, and you'll always be cheap.AMD version of semi conducteur industrie
Related resources
Can't find your answer ? Ask !
May 1, 2002 8:23:54 PM

However these results are disappointing and less flexible as:
1) The DDR400 is limited by both CAS and chipset
2) The bus was not set to 200MHZ FSB on P4 therefore a loss
3) They did not try AMD's Athlon with 400MHZ FSB and DDR 400 which clearly is the centre of attention. Once they Dual Channel DDR400, THEN it is a worth competitor to PC1200. Also notice how 4.8GB/sec of bandwidth made the P4 scream even more than jumping from PC800 to PC1066, oddly enough! I found the PC1200 results awe-inspiring, impressive. But it might be just because of latency, I dunno.
But DDR333 CAS2 was quite a good competitor for PC1066 at 4.2GB/sec.


--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
a b à CPUs
May 1, 2002 8:48:04 PM

Eden, what is your problem anyway? You seriously thing that having the RAM at 150% the CPU bus will make a huge impact on AMD performance? I got news for you! nVidia make a dual channel DDR board and it has little affect. I have more news for you-the fact that they HAD to run at Cas 2.5 is due to the fact that this ram is already being pushed to it's limits. So forget AMD, since this test has no bearing whatsoever on AMD performance. The world does not revolve around AMD regardless of your opinion. There ARE NO RDRAM boards for AMD, NONE, And why would anyone want that anyway? This test compared performance of RDRAM to DDR SDRAM, period. The only processor this is now relavent to is the P4.

Why don't you go to an automotive forum and when they test a new set of Brodix heads ask "But the real question is, what would these heads do for a Ford".

What's the frequency, Kenneth?
May 1, 2002 8:53:56 PM

ROFL

Bravo!!

Beauty is in the eye of the beer holder.
May 1, 2002 9:01:02 PM

Oh god you mistook what I meant again. I meant P4 with DUAL!!
Not AMD!


--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
May 1, 2002 9:07:04 PM

Quote:
3) They did not try AMD's Athlon with 400MHZ FSB and DDR 400 which clearly is the centre of attention.


You're right, it very clearly says "P4", right after- hang on a minute...

<font color=blue>Hi mom!</font color=blue>
May 1, 2002 11:17:08 PM

Quote:
Once they Dual Channel DDR400, THEN it is a worth competitor to PC1200. Also notice how 4.8GB/sec of bandwidth made the P4 scream even more than jumping from PC800 to PC1066, oddly enough! I found the PC1200 results awe-inspiring, impressive. But it might be just because of latency, I dunno.



Eden clearly was not comparing ddr on the athlon, he should have hit enter between his two comments though.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
May 2, 2002 12:23:17 AM

Yeah, my apology to all for the misunderstanding. I would never ever even say anything about DC advantages for Athlons, there just isn't.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
May 2, 2002 12:28:08 AM

I meant that they should have included Athlon tests, regardless of the name.
Unless they are planning to, <A HREF="http://www.x86-secret.com/popups/articleswindow.php?id=..." target="_new">http://www.x86-secret.com/popups/articleswindow.php?id=...;/A> states how good it is for Athlons in competing. I am open to different bench tests, but it'd be nice to have both CPUs tested, not just one, especially since the memory applies for both except RDRAM. Although PC1200 is impressive at the very least.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
May 2, 2002 3:39:52 AM

I think the main point of this article is to show that DDR SDRAM is a better memory technology than RDRAM. The reason I say this is because 333 DDR with CL2 delivers 2.7 GB/s of bandwith, while 2 x PC800 RDRAM delivers 3.2 GB/s of bandwith, but the lower bandwith DDR 333 delivers much better performance than the PC800 RDRAM does. Even DDR 266 at CL2 with only 2.1 GB/s bandwith outperforms PC800 RDRAM in Quake 3. I'd like to see how the Pentium 4 at 533 FSB with a dal-channel DDR chipset would perform comparable to the PC1066 system.

"Trying is the first step towards failure."
May 2, 2002 5:28:24 AM

Do something before i kill it

cheap, cheap. Think cheap, and you'll always be cheap.AMD version of semi conducteur industrie
May 2, 2002 4:28:40 PM

Athlon tests merely would've skewed the benchmarks, the Athlon platform uses memory in quite a different way than the P4 platform.
If they could've tested RDRAM on Athlon as well, then I can see how it would help to add Athlon to the mix.

<font color=blue>Hi mom!</font color=blue>
May 2, 2002 11:58:26 PM

Exactly burger!


The ddr versus rdram debate is POINTLESS, it shoudl be the ddr versus rambus ON THE P4 debate.

Rdram has the dual channel advantage which is NOT part of rdram itself, the very concept of the debate is flawwed imo.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
May 3, 2002 12:55:05 AM

So in a way, even comparing RDRAM DC to DDR SC is flawed? Well I could agree, but each memory is designed differently, no? Like P4's designed for high speeds and Athlon for IPC.

But what I do want, is that there should be another article that puts the Athlons in the mix, with the top runners such as AthlonXP+DDR400+400MHZ FSB, vs P4 with RDRAM PC1200+150MHZ*4 FSB, vs P4+DDR400 CL2.5 minimum+200MHZ FSB(although farfetched, there were many that were able to, otherwise use a high FSB like 150MHZ). This would prove interesting, exactly the test I showed you from that website link that shows DDR400 pushing Athlons far, therefore competing even the best of P4s out there.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
May 3, 2002 5:00:29 AM

Putting Athlons in the mix would only be valid if you could do RDRAM on an Athlon. Remember this is an article on the memory specifically, not just how different platforms fare.

RDRAM DC (you mean dual-channel, right?) and SDRAM SC isn't really a fair comparison, you're right about that. To be strictly fair to the technology however, you would also need to do the same datapath.

But if you're comparing currently available memory platforms (which this article is), as opposed to strictly theoretical, then everything that exists can be thrown into the mix. Except for Athlon (IMO, at least), because it would simply skew the results not having RDRAM for Athlon.

<font color=blue>Hi mom!</font color=blue>
May 3, 2002 6:57:05 AM

I was also surprised with the conclusions. The fact is that at a given point one technology gives more bandwith and better results. It doesn't matter if PC800 < DDR400, because one is the past of RDRAM and the other is the near(?) future of DDR. Compare top memoey products and, if you want, check with street prices and availability. That is the key.

Also I was asking myshelf about the limits of both technologies. It seems like RDRAM have more room for improvement than DDR. Hey, I am on the side of DDR, but that's my opinion (no, no master degree in computer science or similar, so please tell if I am wrong).
May 3, 2002 7:01:57 AM

Sorry, but know I was thinking that there is one forum for memory issues, isn't it?
a b à CPUs
May 3, 2002 8:10:11 AM

NO, DDR333 offers SIMILAR performance to PC800 on the P4. Which is why RDRAM is required for overclockers wanting SERIOUS performance on a P4. Because RDRAM can be overclocked by 33%, while there is no DDR that will work right at DDR444 (333 x 133%). Which is why I've been pushing for dual channel DDR chipsets for the P4.

What's the frequency, Kenneth?
a b à CPUs
May 3, 2002 8:11:45 AM

DDR SDRAM has a large latency advantage over RDRAM, but it would have to be operated in Dual Channel to match the bandwith of RDRAM, which is already Dual Channel.

What's the frequency, Kenneth?
May 3, 2002 11:52:26 AM

Yeah I getcha, perhaps I was not specific. My real demand is that they do a test with the new mem setups cranked to the max with FSB and such, to see OVERALL PERFORMANCE. Current test was probably for the memories on the P4, but in the future I wanna see the viability of these new memory sticks, and also compare overall performance when DDR 400 for Athlon and P4, and the new RDRAM for P4. I guess DDR400 for P4 sucks at the moment compared to PC1200, but if Dual Channeled, I am guessing a clear advantage for DDR400, IF they use a higher FSB and CL2.5 or CL2 RAM.

So again, I would like an overall performance, like everyone here looks for in a system, test to see which is the better buy when the future comes with them chips. (changes will occur to the performance in the future as chipsets come FOR them)

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
May 3, 2002 4:29:13 PM

Quote:
Sorry, but know I was thinking that there is one forum for memory issues, isn't it?


Yes, but people generally post stuff in this forum anyway, no clue why.

<font color=blue>Hi mom!</font color=blue>
May 3, 2002 4:30:09 PM

I agree, but the article would have a different intent. They could also probably simplify it and just use the fastest RAM on each platform.

<font color=blue>Hi mom!</font color=blue>
May 3, 2002 7:06:44 PM

Quote:
But if you're comparing currently available memory platforms (which this article is), as opposed to strictly theoretical, then everything that exists can be thrown into the mix. Except for Athlon (IMO, at least), because it would simply skew the results not having RDRAM for Athlon.


Actually, it looked like it compared memory TECHNOLOGY to me, which is a whole different thing, they should have used a p3 on the rdram single channel boards to truely show the strengths and weaknesses of the dram itself IMO.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
May 3, 2002 7:27:32 PM

Yes you're right, and single channel RDRAM would work, but it would have to be the same datapath and a mature chipset to be able to isolate the RAM technology itself (I don't think anyone would argue that the P3's RDRAM chipset wasn't mature).

<font color=blue>Hi mom!</font color=blue>
May 3, 2002 7:36:04 PM

IMO it would be less skewed than using a dual channel board for the benchmarks.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
May 3, 2002 7:42:05 PM

Agreed, but I still wouldn't accept the results. Sounds like the kind of article THG would write. They could also compare DDR and RDRAM by using the i850E and the KT266.

<font color=blue>Hi mom!</font color=blue>
!