Slashdot Mirror
Intel To Drop Rambus Exclusivity, Support SDRAM
Overload128 writes "Over on ZDNet, there's an article detailing Intel's plan to let Rambus memory stand on its own. It seems that it will stop bundling RDRAM with Pentium 4s and stop giving rebates to OEMs that use P4s and RDRAM. They are also releasing a new chipset soon, called 845 that will support SDRAM (SDR) with DDR support for P4s not far down the line."
from the article:
"I think that Intel is slowly but surely backing away" from RDRAM, said Kevin Krewell, senior analyst with Instat/MDR (formerly MicroDesign Resources). "When they do finally introduce the DDR version of the 845, they'll get good performance--not as good as RDRAM...but everybody, overall, will be happier."
I'm laughing so hard that I'm crying here.
I work at CompUSA.
Let me tell you something, most customers come in and want a cheap computer that will work when they want it to.
It used to be that people only wanted intel, but AMD's prices are so competetive that people don't care anymore, and will accept any system.
HOWEVER! The management of our store ENCOURAGES us to tell our customers that AMD computers have incompatibilities. The store makes more money from intel systems (most people are willing to pay more for a brand name, so we jack up the markups), and we make more commission on intel systems as well. Most employees will just come right out and tell everyone "AMD systems have compatibility problems" and then the customers will look at intel, and tell their friends not to buy AMD either.
Its bullshit.
Whats also funny, is that almost everyone i know who knows more than 2 craps about computers has an athlon, but the people who bought a store bought computer got suckered into an intel.
AMD is hands down better. There are no compatibility problems. AMD now has 23% of the market share of desktop computers (up from 5% only 3 short years ago), and you don't gain that much market share that quickly if your product sucks.
Intel is running scared now that dual athlon systems are out.
Erm, the P4 bus runs at effectively 400MHz (100MHz QDR) and RDRAM at 800MHz (although the bus width is narrow).
The problem with memory is not the bandwidth, its the latency. For example, the peformance difference between SDR DRAM and DDR DRAM is 5-10% running benchmarks and general apps, despite the later having twice the bandwidth. 3D graphics cards use DDR because they dont have the latency issues and hence benefit from DDR.
Another point is that RDRAM has worse latency then SDRAM but has good bandwidth. So what did intel do? They increased the cache line size on the P4 to 128 bytes, a four times increase over the P3. This little modification is what makes RDRAM usable on the P4 as 128 bytes is long enough to mask out the latency problem. It'ill be interesting to see how DDR performs on the P4.
I;m curious (and htis is NOT a flame - I'm really curious) If you've had the $$ for a new box and you want a speedster system - what kept you from getting an Athlon based system?
Try this, tell your companies excutives that when you call your software vendor for support they won't help you because you are not running Intel Processors in your server but only "compatible" processors.
Our accounting package, MAS90/200, is ONLY supported on Intel processors. Until software vendors ok, AMD, we can't as a company use them. Similiarly with other software we use, and it's not like you can just migrate from an existing installation of your core business apps to another because it limits your pc selection. Really, until HP offers Vectras with AMD, and IBM offers servers (our standards for machines), our company won't be getting AMD but is stuck with Intel.
I tend to think this is another reason that IBM, Dell, and HP do not offer AMD chips in their server lines (yes HP has some durons in laptops and AMDs in consumer line pcs but not Vectras or Netservers).
The only thing I can think of is Intel hopes this will put pressure on Rambus to lower prices, that a lot of people will only pay attention to the CPU clockspeed and price in purchasing, so once that starts affecting RDRAM sales, Rambus will reduce its margins. But DDR RAM isn't that much more expensive than SDR, so skimping on the tech doesn't make that much sense. Maybe it did when they were designing the chipset, and wanted it out as fast as possible?
Ooh, a sarcasm detector. Oh, that's a real useful invention.
OK. I'll bite. I haven't bought a P4 yet, but I might.
Firstly, for the application I'm interested in the P4 really screams! How do I know this? Because (an old version of) the application IS one of the SPECCPU2000 benchamarks (254.gap) and the P4 does really, really well on that one benchmark.
Secondly, the CPU cost is just not a big part of total system cost. RAM, disks, case, etc all add up.
Thirdly, I have reports of stability problems for high-memory Athlon systems under Linux. They're a bit old now, and possibly obsolete, but I'd want to be sure.
The big issue against the P4 is memory costs and lack of a dual processor platform, although this last is due to change soon. I like at least 2GB of RAM, ideally 4, and that does tend to point at SDRAM as the only affordable technology. Even DDR SDRAM may be a problem if the mobos don't have enough channels/slots. So, I need to know what damage SDRAM does to the P4 in my favourite benchamrk.
Since you ask, computational pure maths. Everytime I can double my data sizes, I can solve the next problem. I admit that 1GB or so would be enough for other purposes (I also do a lot of compiling and a lot of RAM really helps there).
I don't know how to break it to you, but the price for *all* memory decreased lately, not just SDRAM. Yet Rambus is still over 4 times the price of SDRAM, just like it always was.
And Rambus has some technical strengths, too, when matched with a chip design like the P4 - the max bandwidth of RDRAM is higher than equivalent (PC133 DDR) SDRAM, though latency is higher. THe problem with Rambus was twofold:
Bullshit. The bandwidth of 266MHz DDR is 2128 MB/s. The bandwidth of 800MHz Rambus is 1600MB/s. P4 uses two channels of RDRAM, making the total bandwidth 3200MB/s. That's why in P4 you need to install RIMMs in pairs, one for each channel. So the bandwidth of Rambus is less than that of DDR, and on top of that Rambus has much higher latency and generates a lot more heat. RIMMs need heat sinks for crying out loud!
I, for one, have about 100 PCs to buy once there's a product worth buying.
Ahem, Athlon. Or P3 if you are the Intel type (it outperforms P4 you know...)
___
___
If you think big enough, you'll never have to do it.
Your assumption "dualchannel ddr-ram will not speed up nforce" is wrong.
nforce uses UMA for graphics and just needs VERY FAST Memory, even when it only performces fast at linear access.
I would agree if you say "nforce+dualchannel will not be faster than other ddr-chipsets".
"Life is short and in most cases it ends with death." Sir Sinclair
Actually, I think AMD systems rock - I recommend them to my friends, and I have two of them at home.
The problem with AMD boxes in a corporate envirnment is consistency. When I buy a box for work, I'm not looking for blazing speed and lowest cost. I want a PC that'll be available in exactly the same config for at least 9-15 months from its date of introduction. There aren't any AMD-based systems being sold by the major manufacturers that meet that criteria - the typical "corporate" system (like the Dell Optiplex, Gateway E-series, etc.) will use a known stable chipset with a long planned lifecycle in the roadmap, Intel processors, and typically an ATI video chipset. When I buy them (in bulk), it's easy for me to use a standard system image to build the PC. Then all we have to do is generate a new NT SID, and set up the user profile. It only takes about a half hour for us to build a new PC as a result, including the time to image the PC.
As for the current P4 systems, the reason we're avoiding them is RDRAM. I have no interest in stocking multiple RAM types, and the performance isn't sufficiently optimized to make it worthwhile. For my purposes, there's nothing wrong with the P3/i815 platform (or BX, for that matter), but BX is gone and i815 will be rapidly phased out once i845/Brookdale is available around the beginning of October. Hotrods don't really benefit us at all (our current main platform is still 440BX/P3-650), but stability does. Brookdale is slated to be around for a couple of years, with the DDR version coming out early next year.
One other advantage of buying the "corporate" class systems is that all the manufacturers who want the business will cheerfully share their product roadmaps with you - including model names/numbers, availability dates, length of product cycles, and pricing. However, it's NDA'd, so I don't get to share, unfortunately. However, I should have i845 sample systems in here during August from a couple of vendors to test.
- -Josh Turiel
-- Josh Turiel
"2. Do not eat iPod Shuffle."
Rambus RAM is nowadays well under $1/MB (the Chip Merchant, for instance, sells 256MB of PC800 RDRAM for $170) - the only thing that makes it look expensive now is the utter collapse of the DRAM market. About six months ago, that would have been a reasonable price for PC133 SDRAM.
And Rambus has some technical strengths, too, when matched with a chip design like the P4 - the max bandwidth of RDRAM is higher than equivalent (PC133 DDR) SDRAM, though latency is higher. THe problem with Rambus was twofold:
First, the horrible business practices of Rambus, the company.
Secondly, when Rambus finally became available to the PC market (with the i820 chipset), the platform was so underwhelming that Rambus was effectively squashed. That was when the company turned their attention to litigating rather than working to improve the product - and we all have seen examples of companies that lead through legal agtion. They die.
And now that Intel is going to ship Brookdale, they might start selling some P4 chips at last. I know that, at least at my shop, we've held off entirely on PC purchases this year (except for a couple of servers and laptops), in order to wait for a viable P4 platform. I'm sure I'm not the only one - expect Intel's sales numbers to start rising again and some of the top-tier PC vendors to show signs of breaking out of their slumps. I, for one, have about 100 PCs to buy once there's a product worth buying.
- -Josh Turiel
-- Josh Turiel
"2. Do not eat iPod Shuffle."
The rambus strategy was pretty dumb but going with sdram instead of ddr is even dumber. What intel should do is dual channel ddr like the nforce but it will be a major accomplishment just to see regular ddr for the P4. That's how pathetic this situation is. Now this sounds like pure intel bashing but I'm being honest. I have great respect for Intel technologies over the past decade or so and still do (xscale and that new intel camera we just bought) but they really screwed up since the Pentium 3 and still continue to make mistakes.
---
he was talking about a 1.7 ghz P4 on sdram. It will suck compared to a 1.2 ghz athlon on ddr. And nforce gives the athlon dual channel ddr-intel doesn't have anything like that
---
I don't think the technology was so stupid -- otherwise all the other memory and CPU chip manufacturers wouldn't have picked it up so fast back when the standards were being discussed.
The problem was that Rambus couldn't execute on their designs, and then decided to try to make money by blackmailing everyone else under the sun... :-/
The RAM is still a lot faster, and if we can get rid of all of Rambus's patent nonsense, the price will eventually come down.
--
Your Servant, B. Baggins
Superior? Look at the benchmarks, man! RDRAM consistently falls short of DDR (or in some cases, _SDR_) SDRAM!
Just look at the benchmark results...
BandwidthGeneral Benchmark 2
General Benchmark
-- Minds are like parachutes... they work best when open.
I have 512MB RAM in the one I have for work, and it crashes once every couple of days. I thought this was due to instability in my video driver (a ATI All-in-Wonder Radeon), but that could be a different reason for my problems.
I have a SGI flat panel monitor and still haven't been able to get maximum resolution on it because the driver doesn't support DVI out yet. I heard a while ago that this was fixed in Xfree 4.1 - anyone know about this?
D
----
I think Intel lost big-time because of the RAMBUS stuff. I know I lost a lot of respect for them after that, and explicitly asked for an Athelon system for my desk. I wanted 512MB RAM, and I couldn't stomach the RAMBUS price difference for it.
I'm upgrading my Power Macintosh G4/450 dual processor home system to 1.5GB RAM today. I paid less for 1GB RAM than the RAMBUS folks have to pay for 256mb. I'll bet I'll get better performance on this machine than I would have on the P4, not because Intel is slower than the G4, but since I have so much more RAM.
D
----
I could have bought an AMD package for a bit less money and it would have probably run most programs faster than the Intel package. Why did I buy the Intel package? I don't care about Quake or minor differences in performance. I wanted a reliable and well supported system that wasn't going to have compatibility problems with hardware and software. The large memory bandwidth was also a plus for the Intel package.
Mea navis aericumbens anguillis abundat
The first boards will be showing up in August.
The Rambus folks have single handedly run one of the most vile business models around. Regardless of the quality of their product (which thankfully is low) its important to make an active stand against this kind of behavior, because in the end it hurts us all. Imagine if a company like this had got some patents on underlying internet protocals. The internet would cost a ton and barely exist.
Nice to see you could come up with an original .sig. In case you were wondering, that is pretty bad form.
Do you even know anything about perl? -- AC Replying to Tom Christiansen post.
I for one am damned glad this happened. I want a p4, but there's no way I was ever going to get rambus. It's just too damned expensive. I thought it was eventually coming, but I'm definitely glad that it's official now, this should drop the price of a p4 system by a few hundred dollars here, which is frankly good for everybody...
Plus, we all know rambus inc is evil with it's frivoulus litigations, so this is a good thing for the rest of the industry as well as intel and us.
--Gfunk
Send lawyers, guns, and money!
Really, for the hour or two every year, you might want to consider building your own system. If you go with decent parts (ie, nothing from a company going bankrupt) you won't have any install issues. These days you can boot off of CD and have an install going in minutes.
But I dislike using an OS I didn't install myself because I've got strong preferences for what gets installed, etc. Well worth the 20m it takes to install Win2k, or 45m for Linux (Takes longer, cause I always install a bunch more stuff.)
If that hour or two is really that hard to come by, or if you never do it and it would take a lot longer, consider making a list of parts and getting a local consultant or small store to do it for you.
Saving shipping alone is likely to pay for an hour or consultant time (Expect $50/hour, it's not really high-end stuff) and you get a lot more control over what you get.
This is proven by the simple fact that you can get a Dual Athlon this way, where many major companies (Dell, etc) won't even sell powerful single-CPU Athlons.
I've had great luck buying my computer parts from a hardware consultant. He burns everything in, with the appropriate manufacturer's util when applicable, downloads and burns drivers when needed, and delivers them to my home after 6pm, when I'm back from work. Being that he doesn't have a storefront it usually ends up being the same price (+/- 5%) as the local stores, even the really cheap ones. (With retail products, he doesn't do the volume to get really cheap OEM stuff, but these days I like the warranties.)
If Dell decides to start using AMD cpu's, then my next purchase would most likely be a Dell with AMD inside.
I'm a leaf on the wind. Watch how I soar.
Such as the Tyan Tiger Athlon MP motherboard for $250. Without video, SCSI and NICs though, but it does have IDE RAID.
Better than the $500 for the fully features Tyan Thunder...
I just hope that AMD get a move on with their southbridge redesign to make it more relevant (SuperIO on board, NIC, USB2, etc), with decent NB-SB interconnect speeds, etc.
What's the word on SMP AMDs???
Like the man said, AMD rocks, and for some of us, price is #1 consideration. Now give me a dual Duron system with oodles of SDRAM!
Intel who??
Live today. Tomorrow will cost a lot more!
For example huge databases, or applications like Photoshop, 3D Studio MAX, Maya, Premiere...even math software can consume memory like nothing when you do heavy number crunching.
;-)
Or maybe he uses Emacs!
All they need now is a case that has room for more than one of those beasts..
Instead of jumping all around, could you just read carefully what I wrote : I said that the SDRAM solution was dreadfully slow on the P4. I never said a thing about RDRAM. Sure, RDRAM has move bandwidth and bigger latency.
"The answer to the Question of Life, the Universe and Everything is... 42"
It's people like you that give techies bad reputation. If you used your brains and read what I wrote in the first place, instead of feeling to be on a crusade of bullshiting everyone, you'd have understood, as a lot of other people did too, that I meant the i845 chipset with the SDRAM. So instead of jumping around and crying foulnames because you think you know stuff about hardware, go pick up a pair of glasses. Next time you'll be able to read what I wrote.
"The answer to the Question of Life, the Universe and Everything is... 42"
I've read a review about this a few weeks back. The performance of a 1.7Ghz P4 is pathetic to even a 1.2Ghz Athlon with DDR SDRAM. Intel execs must be on crack or something to release this chipset for the P4. It will seriously lower performance. The thing is P4 will be much faster with DDR memory than 'Rambust' and Intel is afraid to admit their mistake. Do yourself a favor and never buy a P4-Chipset845 combo, you'll be seriously sorry.
"The answer to the Question of Life, the Universe and Everything is... 42"
Unfortunately, you're absolutely right. All this is going to do is drive up the prices of the actual processors.
I personally hope that the removal of RDRAM will force them to drive up the quality of the pentium series (or create a new series for Pete's sake).
Anyway, I dont have numbers, but I'm pretty sure that AMD systems are more than comparable these days. Maybe they'll start looking at them as more serious competitors now that they have SMP solutions.
Attention deficit disorder is a complicated issue, spanning several major... HEY LET'S GO RIDE BIKES!
When nVidia releases nForce, DDR will suddenly have a Dual Channel configuration witha 128-bit wide memory interface. That means that a nForce board with two banks of PC2100 DDR will have about 4.2GB of bandwidth with very low latencies.
Relevant THG Page
Of course this is all still theoretical, but if it turns out to be true, DDR will have a higher bandwidth and drastically lower latencies than RDRAM.
-Sokie
------
Where are the slash-groupies? I distinctly remember being promised slash-groupies!
Rambus, fail? That is one spectacular failure. To date, Rambus has been used in all Nintendo 64s, all Playstation 2s, and several video cards, not to mention every P4 system to date, and many P3 systems as well. I'd hesitate to say it was dead, because I'm sure they must be plotting something...
A deep unwavering belief is a sure sign you're missing something...
The P4 is single processor only. The Intel Xeon (P4 based) supports SMP.
Since I am a big believer in SMP, and the Xeon line is a little out of my price range, I will be switching to Athalons next time I upgrade (not until lower cost Athalon MP boards are available).
Beowulf has nothing to do with Rambus, Intel, AMD, or any other specific hardware company. You can use any group of PC's (not all the same, even) to build a beowulf cluster. I've seen DEC Alpha clusters, Pentium clusters, and yes, Dual-PIII clusters.
"The best laid plans of mice and men gang oft agley..." - ROBERT BURNS
Intel has stated (at the motherboard expo some months back) that their SDR chipset for P4 cuts performance by 25% & DDR cuts performance from RDRAM by ~12%... Or so their graph I saw suggests... Now Dual DDR channels could be very interesting (& if you look at potential bandwidth it could lead you to a figure beating a Dual RDRAM interface by ~12%)...
Oh one other note, currently Intel's SDR/DDR chipsets only support PC100 & PC1600 due to bus limitation (bus is 100 Mhz Quad Data Rate). Changing to a non-synced bus to support PC133 or PC2100 could cause a major penalty in performance.
we are all invisible unless we choose otherwise
Only Xeon P4's are currently SMP capable, standard P4's are not...
Athlon's on the other hand are all capable of SMP... It's just the MP Athlons that are supported by AMD due to a number of reasons...
we are all invisible unless we choose otherwise
Strangely I noticed my room was cooler after I switched from my K6-2 450 sys to my Athlon 1.2 (which I oced to 1.33 within 2 days & could take higher if I wanted)... I don't think most heat stats can be compared to each other easily...
we are all invisible unless we choose otherwise
my old system did have a TNT2 Ultra which was a very hot graphics card while my new system uses a Radeon 64MB VIVO, but if that makes such a big difference then soemthings wrong as the card is generating more heat than the system is...
The PS should not be a factor as my new PS seems hotter than my old one... drives & such are the same, ACPI is off on both...
we are all invisible unless we choose otherwise
The opinion is widespread because people associate the MB chipset with the proc. And AMD 3rd party chipsets were (and still continue to be even today) notorious for shoddy hardware and drivers.
My brothers K5 was a flaky machine. My friend's K6-2 decided to just die. He always had sound card compatibility problems. Many IRC friends had sound issues with the early K7s. Video cards were always iffy ("Use the latest drivers! Upgrade every week!").
Notice that I left out any mention of the Motherboard chipset manufacturer. Most people, even techies, don't make the proper distinction.
A friend who works for AMD has a nice 1Ghz Athlon system. In Win2k it will occasionally hiccup for about a second or so. The problem? The VIA drivers suck. It's like building a great house but letting someone do a crappy job on your driveway and lawn.
Now, of course, some people haven't had any problems or researched the issues ultra-carefully before diving into the Kx fray but those people are few. I'm happy to say that a friend and I put together a nice Athlon rig for colocation and it's been doing just fine. However, this is an issue that has plagued AMD's public perception from the K5 days and that's why there is this reputation, deserved or not. I myself tread very carfeully when dealing with MB purchases.
Fsck cluebie moderators. I'll say what I want, offtopic or not. And fsck having to qualify every bloody statement just
The comment from the submitter makes it seem as if this is something new and important. This is actually really old news. Intel's 845 (Brookdale) chipset, which supports regular SDRAM and DDR has been in the works and well known for a while now and even benchmarked. Intel is probably initially only releasing it in the SDRAM flavor because of exclusive contracts with Rambus Inc. It is expected that in less than a year the DDR version will be out. Intel publically stated they are less than pleased with Rambus Inc. a long time ago.
As another poster mentioned the performance of the P4 with SDRAM is terrible. This is because the P4 was designed for memory with high bandwidth such as Rambus RAM and DDR as opposed to regular SD RAM. Tom's Hardware, perhaps the foremost Rambus hater, has an article on the 845 chipset and its very poor performance with SDRAM. Ace's Harware also has a summary.
All and all Intel's relationship with Rambus and use of Rambus RAM has been very foolish. The P-III was not designed to take advantage of the high bandwidth of Rambus so the improvements versus SDRAM were limited and the price of Rambus made VIA's competing SDRAM chipsets and AMD's solutions much more attractive. Now that Intel finally has a chip (P4) for which it makes sense to use Rambus RAM, Intel is slowly moving toward abandoning Rambus probably in favor of DDR. Although, given how hated Rambus is among RAM makers and the continued superior price-performance of DDR RAM, Intel's moving away from Rambus makes a lot of sense.
Hmm, in real life P4/RDR systems outperform K7/DDR systems by about 3x on memory bandwidth. Since P4/RDR offers only 50% more theoretical bandwidth than K7/DDR, it appears that it is DDR that is having more problems reaching its theoretical limit.
Besides, 60% of 3.2 GB/s is still faster than 75% of 2.4 GB/s. I don't really care who is faster relative to theoretical peak, I just care who is faster. And that's P4/RDR.
Nope, you are wrong. The bandwidth of dual channel PC800 RDRAM is 3.2 GB/s (800 MHz x 16 bit x 2 channels). The bandwidth of DDR SDRAM is only 2.1 GB/s or 2.4 GB/s. The corporate push behind DDR is attempting to confuse customers by giving a name of PCxxx where xxx is the bandwidth, not the frequency.
The problem is that while dual channel DDR will supply that much bandwidth, K7's FSB is quite sluggish and won't be able to take advantage of it. You will already note that currently shipping K7 systems show very minimal performance speedup with DDR over SDR (it's i820 all over again). There's no point in throwing even more bandwidth at the chip, which cannot use the bandwidth.
The most exciting DDR product is the ServerWorks GCHE chipset which has dual channel DDR supporting quad Foster processors, which can actually take advantage of the increased bandwidth.
Am I the only one who still sees "DDR RAM" and thinks "dance dance revolution random access memory?"
so do you think my butterfly upswing another trickstyle will really push my computer THAT much faster?
no
No thank you until DDR chipset/boards arive.
At first this might seem like a good move, until you remember the only current selling point of the P4 is memory bandwidth (until SSE2 is mainstream). So effectively, the took the best feature and killed it to save on costs to the consumer. Too bad most consumers will see "1.8ghz now with 128 megs of SDRAM!" and buy it because they heard SDRAM is cheaper than RDRAM, yet the 1.8gig CPU is over $500.
--------
--------
It's OK to be social, just don't tell anyone about it.
Well, WinXP is just around the corner.
Not everyone deserves a 320i
Its not joe smoe who will switch to linux, it is the people that are on the verge of switching. Something like this will make people like me want to go completely linux, and people like my dad start to consider it more and more.
This Wiki Feeds You TV and Anime - vidwiki.org
Am I the only one who still sees "DDR RAM" and thinks "dance dance revolution random access memory?"
Am I the only one who still sees "DDR RAM" and thinks "RAM made in what used to be East Germany[?]"?
Will I retire or break 10K?
Plus, they designed the P4 from the ground up to support RDRAM. It'd be just stupid to quit supporting them now, that it's becoming almost affordable to buy Rambus memory.
I don't recall which Intel executive (one of the chief architects of the Pentium IV architecture) it was that lead to this common misconception, but he later refuted that this was what he meant. He had said that the Pentium IV was designed to take advantage of large bandwidth chipsets, like Rambus. He then went on to remind everyone that like most microprocessors, the Pentium IV was designed to depend on the supporting chipset to interface with the memory technology. The Pentium IV was therefor not optimized for a specific memory technology. Instead the chipsets were designed to optimally work with the Pentium IV and whatever memory they supported.
Perhaps Intel didn't design the P4 for Rambus from the get-go, but it was designed as a bandwidth-hungry chip. While the P4 has been reviewed and analyzed to death, and indications of its merit seem to be on a perpetual see-saw, a few things seem to be objectively emerging.
First off, the P4 is strongest when streaming and when the new SSE2 stuff is being used.
Second, the P4 is clearly bandwidth-hungry, and knows how to use it. This goes hand-in-hand with the first point.
Now a little more controversial, there have been some reports that on non-streaming problems, in other words, away from its design peak, the P4 is bandwidth-wasteful. In other words, the P4 needs the bandwith to run well, whether it's really needed or not. Between the cache line size and prefetch pattern, the P4 just needs lots of bandwidth. It is well mated to Rambus. It may also suffer badly without Rambus, or at least that much bandwidth.
I'll be curious to see benchmarks of:
P4+Rambus
P4+Intel SDR SDRAM
P4+Intel DDR SDRAM
P4+non-Intel SDR SDRAM
P4+non-Intel DDR SDRAM
From what I've heard on various performance websites, the Intel/Rambus contract forbade them from fielding a non-Rambus solution with > 1GB/s bandwidth. One could look at PC133 and say it was a violation, but I suspect one could finesse that by saying that you can never in practice achieve the peak. I believe the rest of this has been reported as waiting for certain contract provisions to expire. I don't know the details, only the same rumors as everyone else.
The living have better things to do than to continue hating the dead.
And if you are too busy to put together a list of components you want in your computer so a build-to-spec computer builder can make you a system (for much less than Dell), how do you find time to post to Slashdot? It seems the former takes quite a bit less time, and is (no offense to the present audience) a lot more fun. --- "The 24-hour Dell Technical Support center is currently closed. Please call back at another time." (The recorded answer to my first ever Dell tech support call. Not a joke!)
Next time I'm at the mall I'll leave my girlfriend in some fitting room and go test our local CompUSA sales clerks to see if they too came to the conclusion that AMD systems have compatibility problems.
I don't recall which Intel executive (one of the chief architects of the Pentium IV architecture) it was that lead to this common misconception, but he later refuted that this was what he meant.
This isn't a common misconception... it is, in essence, true.
The P4, as it stands today, has very small on-chip caches, certainly by comparison with, say, the Athlon, and the primary reason why that is is because of memory latency. Smaller caches have lower access latencies, mitigating the effects of RDRAM's high access latency. RDRAM also has very high bandwidth, significantly higher than even DDR SDRAM, which means that keeping these small caches filled is less of a problem.
Consider also the very long, 20-stage pipelines found in the P4. These also require very large memory bandwidth to keep them filled, and what memory technology does that better than RDRAM? Conversely, a branch-prediction miss by the P4 takes a long time to work its way through this 20-stage pipeline before the P4 can restart its work. This gives RDRAM breathing space to refill the P4's small caches from elsewhere in memory that it would not have if the P4's pipeline was shorter.
Cynics might argue that perhaps the P4's branch-prediction isn't as good as it could be, in order to make up for some of the access latency that RDRAM suffers from...
Don't forget the benchmarks either: where does the P4 really shine, and convincingly beat the Athlon? That'll be benchmarks where the data set that is being processed is large and (mostly) sequential. Why? Well, it's obvious, isn't it? It's down to RDRAM's massive bandwidth. Other tests which do not hammer the memory bandwidth as hard, or where the data set is scattered around memory, show the Athlon coming out on top.
All of this points to the fact that the P4 was at least designed with RDRAM in mind. If not specifically for RDRAM, then at least for a memory technology with large bandwidth and high latency, and how many types of memory do we know that fit that description?
I suspect that future revs of the P4, designed to work better with DDR SDRAM, will have larger on-chip caches, and (possibly) better branch prediction. The larger L1 and/or L2 caches will be necessary in order for the P4 to have competitive performance when using SDRAM. However, this will come at a price - a lot more transistors on the die. More transistors == larger die size, higher power requirements, and, worst of all for Intel, means that they will not be able to scale the clock speed up as fast as they would like to, which leaves me wondering whether they will genuinely be able to keep up with the Athlon in terms of performance.
Yeah... as a general rule that does seem to be the case. Those who know about computers build, and usually build AMD. Those who don't buy and tend to wind up with Intel.
/Brian
I don't know about Gateway -- their hardware can get very weird sometimes. A lot of their systems ship without reset buttons and have what seem to be soft power switches. I've had to unplug Gateways to get them to reboot.
/Brian
See what rambus did to themselves... THis is the last blow rambus, thus beginning the final stages of its demise. It was a stupid technology anyways
Syndicate
This page left intentionally blank.
The boards have been delayed to September, but I too am drooling at getting ahold of one of these. Screw the onboard video, I want the dual channel DDR!
Wu-Tang Name: Half-Cut Skeleton Get your own Wu-Na
finally! i might actually think about getting a p4 now... its good to see that intel cant just bully their products onto the market like ms has been doing.
anyone know if they plan on making it SMP enabled anytime soon, or is that too much to ask?
This is exactly like Intel pumping up Mhz for marketing purposes rather than actual engineering. High latency, on servers, sucks ass. RDRAM? No thanks.
"Rambus Owners Club"
Stories on how Rambus was the most underappreciated technology of its time/ way ahead of its time.
P4-Rambus support groups and Rambus hacks?
All I see is the future retro-technology, and the enthusists who will religiously follow it.
I demand a million helicopters and a DOLLAR!
I'm more interested in getting the most out of an old BX board or the newer Abit VA-6, Getting an Athlon means getting a whole new board and processor. Upgrading and or OC'ing an Intel PIII (now most are under $90) is a minimal investment. Going with Athlon only makes sense if getting a whole new bundle (memory,CPU, board). I never even considered the P4 because of the price and incompatabilities of Rambus.
After tons of debugging, it was in the implementation of the sqrt() function on windows vs linux. I think i was passing it a zero, and on windows, the return value somehow evaluated to false, but on linux it evaluated to true (or something along those lines).
In any case, same software, different compilers, totally different results. The FDIV error on P1's was supposed to happen like once every 5000 years or something. Of course, maybe that's on average, and *all* of the bugs were happening to you :)
Where are we going and why am I in this handbasket?
Assuming this is correct, the FDIV bug was more related to rounding errors on very high numbers, so I'd say that problem was probably not from the pentium bug, but maybe if it were optimized for 486's and then the exe was copied to the Pentium there *could* be a problem. Then again, maybe POVRay had a bug (gasp!).
Where are we going and why am I in this handbasket?
I think among tech-types, this is a bigger embarassment to Intel than the Pentium bug back in the day. I mean, that bug was fairly harmless really, but it was blown up quite a bit in the media. Of course the actual seriousness of that bug is probably somewhere in between what Intel said and what the media said.
But for this Rambus fiasco, the tech community has scolded Intel pretty hard since day one with one benchmark after another, and Intel's refusal to accept that they may have selected an inferior product may have hurt them in the eyes of people who actually respect them for their technology.
Where are we going and why am I in this handbasket?
It was economic reality that this would eventually happen. The average person might be willing to pay for this equipment, since they don't know there is any option. However the high end power users know there are other options, and they know that the RDRAM Memory is over priced, over rated, and under powered.
Power Corrupts,Absolute Power Corrupts Absolutely, leaving one person(group)in charge is absolutely corrupt.
RDRAM has it's ups and downs. Yes, it does offer more bandwidth per pin then DDR does, however it requires more control, power and ground pins to go along with the data pins. End result? Both RDRAM RIMMs and DDR DIMMs are 184 pins, and one RDRAM RIMM offers up to 1.6GB/s while a single DDR DIMM offers up to 2.1GB/s of bandwidth. RDRAM in the P4 is dual-channel though, so that number doubles in this case.
However, on top of this whole bandwidth issue there are MANY downsides to RDRAM. First off, the whole idea of the low-pin-count was to reduce the cost of producing motherboards and chipsets, but since you need two seperate 400MHz DDR channels to get decent performance from an RDRAM system (as the i820 painfully illustrated), the end result has been chipsets and boards that are MORE expensive rather then less so. Case in point, MSI just recently released the first 4-layer dual channel RDRAM board of any company (as far as I know), all other RDRAM boards are 6-layer ones, which apperently adds ~$15-$20 to the cost of these boards. DDR boards, meanwhile, are mostly 4-layer boards, and SDRAM boards are almost exclusively 4-layer products.
In addition to that there's also the much publicized latency issue (though RDRAM doesn't really have much worse latency then DDR SDRAM, to the point that a good RDRAM chipset will have lower latency then a crappy DDR chipset any day). There's also some issues with RDRAM being essentially a serialized I/O port being used as a bus (hence the requirement for cRIMMs to fill up empty RIMM sockets), which causes some bad transmission line-type problems.
All in all, RDRAM is a great technology for some uses, but the PC isn't really one of those uses (at least not now). The Sony PlayStation2? Now that's a perfect example of where RDRAM makes perfect sense. But a PS2 is VERY different from a PC..
Interesting FWIW, the i845 (Brookdale) chipset that Intel is shipping for SDRAM support has full DDR support built in already, several motherboard manufacturers have already demoed i845 boards using DDR memory. However, Intel will be disabling DDR support until Jan. 2002. Official reason, from what I've heard, is that the DDR stuff isn't fully functional yet. Unofficial rumor is that Rambus has got Intel by the balls (albeit by a quickly loosening grip) due to some legal/contractual deals. ie Intel produces a DDR chipset now and Rambus sues them for patent infringment (though it's becoming increasing clear that intelligence will prevail and Rambus will get their *sses thrown out of court for the whole SDRAM/DDR litigation thing).
On the subject though, IMO Intel DESPRATELY needs a DDR chipset for the P4. Though sales of this chip have picked up significantly, they're still less than Intel's initial forcast, and I think that a lot of it has to do with RDRAM. SDRAM gets rid of that problem, but then saddles the P4 with another problem, ie a 5-25% performance hit (with the largest performance drop coming in the areas where the P4 had previously been the strongest). End result is that it makes the Athlon look like quite an attractive platform from the performance side of things. Where previously the top speed grades of Athlons and P4's were pretty much neck and neck, with PC133 SDRAM the P4 looses quite badly in all but quite rare cases.
Err, umm.. you might try re-reading your sources when quoting them. The heat generated by a processor is a 100% direct relation to the power consumption of this chip (seeing as your processor doesn't give off any light/sound/other form of energy.. or at least mine doesn't :> ).. In other words, if one chip draws 70W of power and another draws 88W of power, the only way for the second chip to run cooler is that it has a better heatsink+fan and/or better case cooling. Now,if we can read beyond Intel's somewhat confusing new "Thermal Design Power" to figure out how much the chips actually use, I think you'll see that the Athlon and the P4 draw damn near the same amount of power for roughly comperable performance levels (ie a 1.4GHz Athlon and a 1.8GHz P4). Typical power draw for the P4 is a tiny bit lower (roughly 60W vs. 66W for the Athlon), with a maximum real-world draw being about 72W for the Athlon and slightly higher for the P4, and the maximum theoretical power being about 74W for the Athlon and 89W for the P4. As you can see, the P4 has a somewhat wide range of temperatures then the Athlon, but really heat is a non-issue when comparing these chips.
Now, when compared to the Athlon or P4 to a PIII, well that's another story (hence the reason why it took AMD over almost two years to get laptops using Athlons and it'll take Intel a while to get a P4 for laptops out).
FWIW, anyone interested in knowing where these numbers are from, check out AMD's Athlon PGA datasheets (where they quite nicely and clearly provide the above numbers) or Intel's Thermal Design Guidelines for the P4 (where the info is presented in a somewhat convoluted manner which requires a bit of reading and guessing from graphs).
It doesn't make sense for Intel to be as worried about which type of RAM it's chip can support. It needs to start worrying about forcing mobo's to start increasing BUS speed. It's still the worst bottleneck in my PC.
I really hate Dan Patrick.
Nice that they put out a new platform (if you wnat to get a budget p4-system) but why didn't they try to get it right from the start? Using non ddr-memory with the p4 is lika asking for crappy perferformance. Toms hardware tested the paltform here and came to the same conlusion. Still, I think I'll stick to athlon.
Unfortunately, it is due to some sort of brand loyalty. I WISH that I could get a nice elegant Athlon system from my favorite PC vendor (DELL) but they just won't do it. I have nothing but positive experiences with Dell PC's and am very reluctant to switch, but if anyone knows of another MAJOR PC vendor that bundles their systems with AMD I would be happy to look into it. But as of late, they are not THAT easy to come by. And please, spare me the "build your own system and save" replies - my time is precious.
Since when is "public safety" the root password to the Constitution?
It's nice to see Intel get with the program and dump Rambus. This outfit has been screwing with the system too long. They have created quasi-standards then demanded licensing for them; built an empire on suing manufacturers, and have been investigated for all sorts of fraud. It's time for Intel to move on to support real standards in an open competitive enviroment. It's nice to see them taking steps in that direction
--CTH
--Got Lists? | Top 95 Star Wars Line
Brookdale was already in Intel's roadmaps for a long while. They're just waiting for their contract with Rambus to end, since it's preventing them from releasing DDR products.
Plus, they designed the P4 from the ground up to support RDRAM. It'd be just stupid to quit supporting them now, that it's becoming almost affordable to buy Rambus memory.
Personally, I loathe Rambus as much as the next guy, but if you think rationally, this decision should have been taken by Intel a long time ago -- now it doesn't make sense anymore.
Join the NFSNET. Our prime goal is making little numbers out of big ones. http://www.nfsnet.org/
Odd, I notice that my mid-tower is very warm to the touch, while the old K6-2 300 wasn't that bad at all. The K7 is very warm, with a hotter ambient air temp in the case and more airflow.
OTOH, I'm wondering if your P/S or other component was running hot. When I upgraded to my "new" system, the old system donated a lot of parts, and the CD-drive/Hard Drive, although replaced, were newer (although the 40 gig HDD in the new machine is also a significant source of heat, IMHO). Also, APCI will reduce the temperature, although I have APCI turned off in my own machine (I'm trying to track down a minor stability bug). In systems I sell, APCI is turned on, and yet I notice the heat, even with the smaller hard drives and generic (8 meg) AGP vid cards (5400 rpm, 20 gig IDE).
OTOH, monitors seem to be running at the same/simular temperatures over the years, at least for their relative sizes. My old amber monitor (9") runs just as hot as the newer 15" monitor I have.
1 lb heatsink jokes aside, the P4 runs cooler then the Athlon. Hell, with the default Athlon OEM heatsink, 1+ ghz CPUs have no problem hitting 50+C. The slot athlon is rated to 70C, the socketed athlon is rated to 95C, which is only 5C short of boiling. My home CPU (K7 1133), with OEM heatsink and only one case fan has hit temps of 71C. (Yes, I'm shopping for another heatsink, I just kept the OEM heatsink on because that's what we sell to our customers, and I'm curious about the temp.)
On the other hand, the P3 @ 1.13 Ghz (and the P4 @1.5 Ghz) have a max temperature rating of 72C, which probably means its normal operating temperature is a tad cooler.
Now the difference in temperature measurements probably means that the actual temp vs the measured temp of the pentiums are more accurate, rather then the underreporting I suspect the K7's temp measurement reports. The pentium has its temperature measured by an internal diode. The K7 goes by max case surface temperature of the cpu. I'm guessing that the K7 method of temperature reporting will lead to a lower then real temperature then the P4's method of temperature reporting.
So, if heat is an issue, or if there is a lot of computers in a room and AC might be an issue, then there might be a reason to go with the P4. Of course, the ignores the fact that RDRAM is supposed to run hot, and I'm too busy/lazy atm to look up the total heat output of a P4 + RDRAM solution vs K7 + DDR-SDRAM solution. Anyways, now since the P4 is switching to to DDR-SDRAM, that is no longer an issue. Intel traditionally has had a better chipset then AMD's 3rd party solution, at least in the opinion of many people in the technical community. This is another factor in favor of Intel. Also, the K7 doesn't beat the P4 in *all* benchmarks. There are some benchmarks where the P4 will beat the K7. So even though the K7 is a great chip for the money, I could see using the P4 for a computer devoted to a specialized app if the time saved was worth the extra price of the P4.
There is also brand loyalty, and service/supplier agreements that lock a company into dealing with intel-only products. If a company is getting great technical support from an intel-only supplier, I cannot see any reason to change, great technical support is worth the extra cash.
Of course, "intel" and "pentium" are household names, everyone keeps asking me who "AMD" is and what they do. And the P4 has a higher mhz rating which makes the idiots flock to it.
From the consumer side, intel is hurting. I work in a shop that has switched over to selling only AMD, based on AMD's price. Sure, we can and will custom order Pentium CPU's, but after we show the customer the price difference between the P4 and the K7, they have all switched over to the K7. Intel still has a stronger market in multi-processor solutions, but the release of the Athlon MP (as well as the fact that all K7's, Athlon or Duron, support a dual processor configuration) and the availability of a dual-processor motherboard will change that, especially when the other motherboard manufacturers release their motherboards and the prices lower to a more reasonable level. AMD has a different SMP solution then Intel, and I believe that AMD's version will convince shops to switch over to the Athlon MP, as long as AMD markets a dual-processor configuration successfully, and can keep up with the demand.
Okay, I'm done playing devil's advocate. Last Pentium CPU I've owned was a P100. I'm quite happy with my Athlon, even if it is a portable heater.
References: Processor Electrical Specifications - Gotta credit my sources for the temperature ratings.
I can't remember that last time I saw a widespread AMD only problem
Last widespread AMD only problem I knew of was with Lotus Domino Server. It required a special build to even run on an AMD system. I know that's why some buisnesses don't get the AMD systems, and stick with Intel.
Fare thee well, poor comment. For thou hast been cast out amongst wolves.
Every time I see "Pentium MMX" I think "Pentium Mega Man X".
Gateway & COmpaq have been selling Athlons for some time. Dell's reluctance can only be attributed to a sweetheart deal from Intel - I too love Dell systems (My laptop is an Inspiron and I managed a support group who handled Dell servers and desktops - it was great) and can't understand why they refuse to add AMD processors to their lineup. I've been surprised they have held out this long - but Dell has a HUGE edge over others in that their corporate contracts which are for HIGH margin machines allow them to earn plenty of money without trying to attract low end buyers with AMD or Celeron type processors. I fear that until (if ever) corporate execs are willing to buy Athlon machines because they feel they are cuttin gedge, Dell will have no need to bother with AMD which is sad.
Top Most Bizarre/Disturbing Error Messages
I;m curious (and htis is NOT a flame - I'm really curious) If you've had the $$ for a new box and you want a speedster system - what kept you from getting an Athlon based system? The benchmarks I've seen between teh fastest P4s and Athlons haev been mixed - P4s excel in some areas while Athlons at much slower clock speeds excel in others the P4 can't seem to master. I'd paint it a draw unless all you do is play Quake :)
So is it brand loyalty? The need to have that ultra high (and meaningless) MHz number? P4s are more expensive, though sometimes the mboos are cheaper than equivalent Athlons.
I'm not trying to start a flame fest of replies - I'm just curious if maybe theres experiences out there that the numerous HW sites haven't touched on that make waiting for the i845 worth doing.
Yes, I tend to lean towards Athlons due to teh price/performance ratio. I can't stomach the prices Intel wants for some of their CPUs when a 1GHz Athlon 266FSB is now < $100 But in teh technical community that is /., besides brand bias/loyalty are there really black and white compelling reasons to shun Athlons in favor of a P4 with SDRAM support? Remember, some of the benchmarks that P4s excelled on were due to memory bandwidth of RDRAM and with SDRAM, they may not be so high - only time will tell of course.
So if you reply, try to be insightful instead of saying AMD sucks - who knows - this thread might be worth reading for folks who don't check the HW sites daily :)
Top Most Bizarre/Disturbing Error Messages
...is the asus version of the nforce chipset (though I can't find out when the damned thing will come out; the only info I can find is from the same nvidia press release that says "Q3").
But I agree - screw intel - and REALLY screw intel if the nforce chipset is all that it's cracked up to be... it'll make better use of the processer, ram, and the north/southbridge interface (not to mention the rockin' onboard sound and graphics)
I have yet to see one and I don't know anyone who owns one. Im not convinced they even exist.
-And if the band you in starts playing different tunes......
Is the rambus monopoly/evil empire era really over? All without Anti-Trust laws? How can it be that capitalism actually works?! :-)
===> An eye for an eye makes everyone blind - MG
I love how so many companies that grew up around the PC (umm... hmm... Microsoft and Intel ring a bell???), that helped in it's creation, and made a shit load of money off of it's one general principle (It's Open) are now trying to tell the consumer what they can do with it. I'm happy Intel repented on their mistake, because forcing people to buy something is a universally stupid mistake in a industry based on competition. Now if we can only get Microsoft to get rid of that stupid hardware configuration, 1 license = 1 computer crap they built into Win XP...
can't sleep slashdot will eat me
my RAM won't cost as much as my damned CPU/Motherboard combo.
Screw 3...
How long will it be until intel supports DDR? I just purchaced a computer, and I decided to go with AMD. Granted I've heard a lot about the benchmarks between AMD and Intel, but the main two factors for me were cost vs performance (in which AMD seems to almost always been ahead), and the fact that AMD has been going with DDR RAM.
And I think most of us saw Rambus spiraling down for a while now, so this isn't so much of a surprise.
This story is months old... The p4 was designed to work with Rambus... it was a stupid move, but something even more dumb is going back to SDR. SDR makes a p4 absolutely useless, the p3 will destroy it... its designed to be able to use high bandwidth. Even DDR kills the P4, the long latency of the p4 chip, and rambus work together, the small latency of DDR won't help it much. DDR won't be coming for a while anyway, next year from what I've heard... they have a contract with rambus about DDR... but not SDR. If anyone is stupid enough to buy a SDR P4 system, they should be shot. The P4 is a failure, it was not designed well, you can't change it. From heat issues, to using ridiculous amounts of bandwidth on board to due simple functions it just sucks.
DDR2 is in the making 400mhz to 533mhz in late 2002
This will rock, DDR is the future.
Perhaps thats why PC133 rams are sooo sooo cheap, dumping of ALL inventory, to make way for HUGE DDR demand. Ahh well at least NO ONE has an excuse to have less than 512meg ram any more in a SDRAM system.
Did anyone not see this coming? People were saying that Rambus was going to fail ever since it came out.
Personally I'd never buy another Intel, they're to expensive... plus AMD's preform just as good and in some cases even better! This is just IMO, I dont want to start an Intel vs. AMD war.
Find Escorts, Strippers, Massage Parlours, Swingers
I saw this coming a mile away. Intel was being stubborn support RDRAM, but now that Rambus is a sinking ship, Intel is bailin' out! On a related note, PC2100 still will make a P4 pale to a T-bird
Yeh, you can beowulf cluster Amigas (god bless Lightwave)...it works for any technology at all
Tom's Hardware reviewed a reference i845 3 weeks ago. As we might expect, the SDR neutered the P4. Now DDR would be another story... http://www6.tomshardware.com/mainboard/01q3/010702
- Dessicator