Printable View
http://www.inqst.com/articles/p4band...dwidthmain.htm
InQuest did a study on the Pentium4 and they mentioned limitations that I hadn't heard yet.
------------------
Q: How many UGA students does it take to change a lightbulb?
A: Two, one to call a friend at GT and another to follow the directions.
I found it very interesting, and the follow up discussion on Slashdot. My own take on it is that the P4 was released a year earlier than planned, to offset the success of the Athlon, and the PIII problems over 1GHz.
The 0.13u P4 might actually be quite good, if they have solved that bus bandwidth wastage problem (ooh, we have a 400MHz bus, bus 75% of the data going over that bus will never get used).
It also proves that I was correct a long time ago on Sharky Forums when I said that the P4 uses up over 70W of power - indeed it does: 72.9W at 1.5GHz. I wonder what the figure will be of 1.7GHz?
The Athlon 1.33Ghz also uses up 73W, for comparison. That is one helluva lot of power in anyones books. Palomino, and then Thoroughbred cannot come soon enough.
I read somewhere that Palomino was 45m transistors, and 150mm^2 - anyone else read this?
This is a misnomer. The architecture came out exactly when planned, and dozens of roadmaps can confirm this. The engineers responsible worked very hard, and actually made the agressive deadline for shipment. Of course, as in any design, decisions to cut back are made, and sometimes these have little to do with actual performance, and other times, they can hurt. I tend to believe that the Pentium 4 came out the best as was economically feasable at the time. On a smaller process, sure less compromised could have been made, but the rumor that the product was rushed is totally unfounded, and is a slap in the face to the hard working engineers that made it possible. Just my $.02.Quote:
Originally posted by Hattig:
I found it very interesting, and the follow up discussion on Slashdot. My own take on it is that the P4 was released a year earlier than planned, to offset the success of the Athlon, and the PIII problems over 1GHz.
And does this perhaps have to do with some rediculous analysis be a certain analyst at Inquest Marketing? This is actually completely false information. If the Pentium 4 was indeed that poor with bandwidth management, it would never score so amazing high on memory bandwidth tests such as Linpack and Sandra, not to mention scientific benchmarks like Swim (of the SPEC2000fp suite) that are world renouned for measuring memory bandwidth efficiency.Quote:
Originally posted by Hattig:
The 0.13u P4 might actually be quite good, if they have solved that bus bandwidth wastage problem (ooh, we have a 400MHz bus, bus 75% of the data going over that bus will never get used).
The analyst's main mistake was assuming that 64 byte transfers on a system supporting 128 byte cachelines would yield 64 bytes of unused information. From a person who has specifically worked with the Willamette bus, I can assure you that this information is completely false. The Willamette bus has an in order queue of 12, which allows many pipelined transactions to maximize memory bandwidth. It is actually far more efficient than the P6 bus. I do not have the strength of voice to put a stop at this particular rumor, so I know it will catch some wind in the process, but I hope that readers of this post will know the truth, straight from the horse's mouth.
According to the following specs:Quote:
Originally posted by Hattig:
It also proves that I was correct a long time ago on Sharky Forums when I said that the P4 uses up over 70W of power - indeed it does: 72.9W at 1.5GHz. I wonder what the figure will be of 1.7GHz?
http://developer.intel.com/design/pe...s/24919802.pdf http://www.amd.com/products/cpg/athl.../pdf/23792.pdf
The thermal design power of the Pentium 4 under average conditions is 54.7W @ 1.5GHz. The Athlon is 63W @ 1.33GHz. See the difference? Total maximum power draw is a useless figure to throw around. You can essentially write a program that takes the most power hungry instruction and run it in an infinate loop, and then measure power dissipation, and sure, it will be higher. Manufacturer's should design around the max TDP, just in case some wise guy writes such a program, hoping to fry another person's CPU.
However, in terms of comparing TDP of processor families, please refer to the average TDP, which is the average case in 95+% of the cases. The Pentium 4 is still 13.5% lower power than the Athlon, even clocked 166MHz faster. Both chips stand to emit a lot of power. However, the Pentium 4 was designed from the beginning to allow for better hardware to mount on the processor and allow better cooling. The Athlon is using the same heatsink as when the Thunderbird first launched last year, which at the time, was still not the best solution.
Again, this is maximum TDP, and a useless number for comparison. Still, that particular Athlon dissipates 63W on average, which is still high up there. Palomino is said to aid in lowering power, but how much is certainly a consideration I'm sure we'd all like to see for ourselves.Quote:
Originally posted by Hattig:
The Athlon 1.33Ghz also uses up 73W, for comparison. That is one helluva lot of power in anyones books. Palomino, and then Thoroughbred cannot come soon enough.
I read that, and it's possible this is a measurment of the server version of Palomino, which may include 512KB of L2 cache. If it is meant to be lower volume, it won't be so costly for AMD to implement the added cache. I kind of doubt that Palomino has enough architectural enhansements to warrent the extra die size and transistors.Quote:
Originally posted by Hattig:
I read somewhere that Palomino was 45m transistors, and 150mm^2 - anyone else read this?
------------------
Grand Knight
OC Crusaders
[This message has been edited by Arcadian (edited April 13, 2001).]
Thanks for replying.
I do not hold much faith in Sandra - IIRC this is basically a Stream benchmark, and although my knowledge of what Stream is is limited, it is probably: time = time(); read_a_lot_of_memory_consecutively(); timed=time()-time;number=1000000000/timed;Quote:
Originally posted by Arcadian:
[BIf the Pentium 4 was indeed that poor with bandwidth management, it would never score so amazing high on memory bandwidth tests such as Linpack and Sandra
[/B]
The P4 gets around 1500 - 1600MB/s. I have seen an Athlon optimised Stream get 1400MB/s (source forgotten, probably AMDZone or the like, and it was AMD optimised, and maybe a P4 optimised version could do even better, etc).
I would like to see some MemTach results for the P4, Athlon and PIII - it appears to be a much more comprehensive memory benchmarking facility. Shame that no-one is using it yet in anger, and that there is nothing to compare it against, and the online results browser *still* isn't complete - did the guy die or something?
I would also like to see Palomino results - the Palomino has prefetching enabled (and better branch prediction) (source forgotten, AMDZone, OCworkbench, hardocp, somewhere, okay?).
In the end the P4 performs between a PIII 1GHz and an Athlon 1.333GHz with current software, whatever the benchmarks say. Yes, the P4 is an architecture for the future (as long as you don't own a current gen P4 with the out-of-date socket already).
I stand by my belief that the P4 was essentially released a year early (okay, maybe 6 months early). Sure, it was working well a year ago, but when the die shrink takes effect, and the new packaging and new socket, and probable minor improvements to how some things are done are implemented by the end of this year, then we will see a much more complete product.
Intel management pushed the chip out of the door as soon as it worked reliably with a working chipset - I never said that the Engineers were crap (although I bet a lot of them secretly want to work on XScale or McKinley https://www.sharkyforums.com/images/.../2005/06/5.gif ) just that as with any deadline, work is rushed, things aren't perfect. I suppose that is what steppings are for though...
Thanks for the reply Arcadian, I was hoping you would have something to say about that. I wonder what Intels responce is going to be.
------------------
Q: How many UGA students does it take to change a lightbulb?
A: Two, one to call a friend at GT and another to follow the directions.
There are a lot of flaws in that inQuest article. Way too many, to begin listing them. Before jumping to the author's conclusion, read the .pdf's yourself.
------------------
--Wingznut PEZ
* Am not speaking for Intel Corp. *
is the part about the P4's integrated power management true? limited to 54 watts? (with standard cooling and room temp 20°, better cooling pays off i presume) i am not exactly sure people won't notice that their 'highend' P4 1.7GHz is just as fast as the a 1.5...
Intel is going to have a problem: angry customers without an upgrade option who paid big bux for a comparably slow rig, Willamette is a rip-off and if they continue to abuse their brand name and peddle snake-oil, they'll be out of business by 2002.
[This message has been edited by Captain Iglo (edited April 15, 2001).]
The whole point is that this is NOT true. 54W is simply the average power dissipation of the Pentium 4. Thermal Throttling kicks in at far, far higher temperatures. Read the .pdf.Quote:
Originally posted by Captain Iglo:
is the part about the P4's integrated power management true? limited to 54 watts? (with standard cooling and room temp 20°, better cooling pays off i presume) i am not exactly sure people won't notice that their 'highend' P4 1.7GHz is just as fast as the a 1.5...
The only problem I see is when misinformed marketing researchers continue to spread FUD about a processor that causes the entire enthusiast community to believe it's true right off the bat.Quote:
Originally posted by Captain Iglo:
Intel is going to have a problem: angry customers without an upgrade option who paid big bux for a comparably slow rig, Willamette is a rip-off and if they continue to abuse their brand name and peddle snake-oil, they'll be out of business by 2002.
On Ace's, I found a pretty interesting piece of information. McCormak, who wrote the article you mention, is actually working for a man named Van Smith at Inquest Marketing, who happens to be a long time Intel hater. He used to work for Tom's Hardware, and most of the reputation that Tom's Hardware is biased comes from this man, Van Smith. Don't you think there's a connection here? I'm sure McCormak is easily able to play off of the ignorance of the rest of the community by playing off of certain words in the Pentium 4 tech specs that might allow him to make the Pentium 4 look very bad.
Inquest marketing used to be employed by Micron to print horrid FUD about Rambus, and now it looks like they are employed by AMD to spread FUD about Pentium 4. Which marketing do you find the most ethical? Having the funding to advertise Blue men playing with a giant '4' symbol (while your competetor cannot finance such an advertising campaign), or hiring Inquest marketing to spread false, but believable FUD about your competetor (while the competitor in question wouldn't sink to such depths)?
------------------
Grand Knight
OC Crusaders
Arcadian said:
>Inquest marketing used to be employed by >Micron to print horrid FUD about Rambus, >and now it looks like they are employed by >AMD to spread FUD about Pentium 4. Which >marketing do you find the most ethical? >Having the funding to advertise Blue men >playing with a giant '4' symbol (while your >competetor cannot finance such an >advertising campaign), or hiring Inquest >marketing to spread false, but believable >FUD about your competetor (while the >competitor in question wouldn't sink to >such depths)?
Do you have any proof to support your assertion that AMD is involved in any form?
By the way, I don't have any problem with Intel's advertising. I don't see any reason why they can't use their hugfe advertising budget to their advantage. AMD would do the same if it could.
argghhghg
I guess I'll have to break down and learn some UBB code! https://www.sharkyforums.com/images/.../2005/06/5.gif
Call it a gut feeling if you will, but you should know the people in charge of articles for InQuest marketing.Quote:
Originally posted by scottd:
Do you have any proof to support your assertion that AMD is involved in any form?
By the way, I don't have any problem with Intel's advertising. I don't see any reason why they can't use their hugfe advertising budget to their advantage. AMD would do the same if it could.
http://www.inqst.com/contact.htm
See the name Van Smith? He used to work for Tom's Hardware Guide, and was the biggest opponent for Intel and proponent for AMD on the site.
Here are some articles that give you a good idea of Van's motto.
A Titan Falls - AMD Plays David to Intel's Goliath http://www6.tomshardware.com/column/...209/index.html
Intel Roadmap Q1 2000. http://www6.tomshardware.com/cpu/00q1/000228/
Intel Developer Forum Day 0: Notes From the Field http://www6.tomshardware.com/business/00q3/000822/
A Quick Look Into the "intel play" QX3 Microscope http://www6.tomshardware.com/consumer/00q1/000330/
Intel Developer Forum Day 1 http://www6.tomshardware.com/business/00q3/000823/
Well, I can go on, but you get the general idea. Van Smith leaps at the chance to give a dig to Intel, while at the same time, puts positive spin on anything AMD puts out.
Bert McComas is no better. Check out the summary of the last few articles that InQuest has put out. Every single one bashes Intel to some degree.
http://www.inqst.com/articles/pcixvib/pcix.htm
http://www.inqst.com/articles/athlon...lonplusddr.htmQuote:
Originally posted by Bert McComas:
Is there a difference between PCI-X and InfiniBand? Yes.
There is an indisputable need in the market for PCI-X. It is required as soon as possible to satisfy (in a compatible manner) platform bandwidth requirements that are upon us in the present and immediate future.
By contrast, InfiniBand is one of several options for the future. It may supersede and obsolete all other present and future I/O technologies as Intel seems to suggest, but there is reason for doubt. As a replacement for PCI-X, the arguments for InfiniBand come up short. As a replacement for Ethernet and Fibre Channel, we also have trouble seeing an unobstructed entry point for InfiniBand.
Like an amphibious vehicle that can function on both land and water, but makes neither a good car nor a good boat, InfiniBand tries to portray itself as a jack-of-all-trades, yet finds itself lacking or unproven in many areas. We do not get the impression that Intel is unaware of this (they created it). Nor do we get the impression that Intel is going to let up simply because the market begins to call their bluff. The RDRAM fiasco has proven that Intel knows how to take its own medicine when its strategies come back to bite them. InfiniBand could turn out to be an instant smash hit. Or it could turn out to be another test of Intel’s intestinal fortitude.
http://www.inqst.com/articles/p3vp4/p4vp3.htmQuote:
Originally posted by Bert McComas:
For as long as the P3 remains a static target at 1GHz and below, Athlon has it completely cornered in terms of price and performance. Intel’s Tualatin chip could only breathe new life into P3 if Intel drives clock speeds to match Athlon in the second half of 2001. At this point we doubt that Intel will do this, preferring instead to fight Athlon with the P4.
We expect that AMD will beat Intel to a fully integrated 512K L2 cache, by shipping a new CPU core in early Q1’01. This will be almost 6 months ahead of Intel’s transition to a 512K cache. AMD will also debut its dual processor platform for Athlon. This will be very attractive to the workstation market. Dual Duron platforms could also become quite a curiosity among enthusiasts on a budget.
All of AMD’s activity to expand the high end of the market will be complemented by VIA, ALI, SIS and others with inexpensive, high volume chip sets for Athlon and Duron in the mainstream, adding features such as UMA. With certainty, these same companies will also successfully deliver DDR to the P3 platform ahead of Intel, to strengthen its position in the midrange and the low end.
There is no doubt, that with the fall of the P3, the end of an era is upon us. 2001 will prove to be an interesting year - to see who will control the next Dynasty, and for how long.
http://www.inqst.com/articles/p3ddr/p3ddrmain.htmQuote:
Originally posted by Bert McComas:
The range of applications we have evaluated here suggests that SMP capability could have a valuable performance impact for many Windows 2000 power users. Without doubt we will continue to find other applications of interest to professionals and consumers that respond favorably to the existence of a second CPU.
Under such SMP friendly circumstances, a fast single processor P4 cannot really stand up to a comparison with a Dual P3 or Dual Athlon platform. Merely raising the clock speed of the P4 a few more notches cannot compensate for the performance deltas we have seen in this project.
We expect that as hardware reviewers upgrade their benchmarking tools and methodologies, and as users begin to identify the potential advantages of SMP, that more platform solutions will become available for high end and even mid-range markets. Already numerous new platforms are on the way for P3 and Athlon that make use of PC133, VC133 and DDR SDRAM.
You can see he is subtle sometimes, and straightforward in others, but every time, he gets a dig in whenever possibly. And in one case, he brings up dual Athlon systems when the article has nothing to do with them, and incidently, they are still unavailable.Quote:
Originally posted by Bert McComas:
Main memory system production will begin very soon. Manufacturers seem intent on delivering DDR into every market segment over the next 8 to 12 months. Even with all of this activity, Intel appears to be on the sidelines, languishing under a contract it wishes it had not signed. Intel’s next generation P3 platform, Almador, is often rumored to have DDR capability, but Intel is not forthcoming. All indications are that Intel may yet withhold its own P3 DDR platform until the second half of 2001. This leaves a huge window of opportunity for VIA’s customers to open the market, take the performance lead, build momentum and cost reduce DDR platforms for the mainstream, perhaps before Intel even samples a DDR chip set.
This company sure does have a bias, but what are other people saying about it? Check out this conversation on Ace's Hardware Forums between a concerned reader, and Van Smith.
http://www.aceshardware.com/board/ge...ge_id=30014483
http://www.aceshardware.com/board/ge...ge_id=30014490
http://www.aceshardware.com/board/ge...ge_id=30014484
http://www.aceshardware.com/board/ge...ge_id=30014486
http://www.aceshardware.com/board/ge...ge_id=30014511
At this point, Van refuses to answer. Of course, these can be read several ways, and I'm not going to pass this off as proof of anything, but my opinions are certainly swayed from the available information.
I don't like bashing of Intel or AMD. I don't like it when hardware becomes political. What differentiates this debacle from a senatorial election? Try to get more dirt on the opposing party, while making yourself look good. Isn't that what InQuest is doing on behalf of AMD? Or DDR? It seems they have made it their quest to spread FUD on Intel, and I don't like that any more than someone spreading FUD on AMD.
Let's get realistic here, people, and read things for what they really are. Sorry for my rant, but that's all I have to say. Thanks for your time in reading this. https://www.sharkyforums.com/images/.../2005/06/5.gif
------------------
Grand Knight
OC Crusaders
[This message has been edited by Arcadian (edited April 15, 2001).]
One has to admit, it does look a bit fishy.Quote:
At this point, Van refuses to answer. Of course, these can be read several ways, and I'm not going to pass this off as proof of anything, but my opinions are certainly swayed from the available information.
The following update from www.hardocp.com may be of interest regarding the whole clock throttling issue:
Arcadian said:Quote:
Kyle:
Now that you guys have had a couple days to get excited about what Bert McComas had to say about the P4 clock throttling itself, I thought I would throw in my two cents. First off, here is the statement that is stirring the pot so well.
Intel’s Thermal Design Guide has revealed that the absolute maximum power dissipation of the 1.5GHz P4 is actually 72.9 watts. This is 33% higher than the published system design specification, and essentially identical to the 1.33 GHz Athlon. In order to prevent the CPU from exceeding 54.7 watt, thermal throttling is used. If performance critical applications drive CPU power above its artificially low 54.7 watt limit, the CPU is halted with a 50% duty cycle (alternating 2 microseconds on; 2 microseconds off) until it cools down. This effectively turns your 1.5GHz processor into a 750MHz processor – just at the moment you demand peak performance. On the other hand, you will probably still be able to check your email at 1.5GHz.
While I don't know Bert, I have had the pleasure of meeting him and you have seen his links here on the [H] many times. On this occasion I think Bert has been sucking the crack pipe a bit too hard or either must have been in a terrible auto accident and had his cranium lodged in his rectal cavity and did not notice before he wrote the above statement.
We have been running an over-volted overclocked Pentium4 with the factory heatsink installed now for some time. It has been running here beside my desk folding proteins for Stanford University now for a solid month and has stayed at 100% CPU utilization. I track its performance and I can assure you that it has not ever slipped into the throttling that Bert speaks about above. If Bert's apps are running at 50%, it is because he does not have the sense to put a heatsink on the CPU or either he is operating his P4 system in Hell. Bert is taking an Intel safety device and demonizing the P4 with it. Here is what Intel PR George Alfs had to say about Bert's statements.
Hi Kyle,
You can run benchmarks all day on a Pentium(R) 4 processor with the benchmarks unaffected by the thermal protection circuitry. The key is to have a robust heat sink and thermal solution. With the heat sink setup we designed for Pentium 4 processor systems, I have yet to see thermal protection kick in.
George.
I have to fully agree with George's statements and have a few things to add. Also, I think that "robust" need not be in his statement.
What Bert may not know is that some mainboards have an adjustment in the BIOS that you can set yourself with the temperature that you want throttle to. (On our systems we have left it at default and never messed with the settings on the particular that is Folding.) Yes, YOU can turn this on and off and fully control it on some mainboards. If you want a shield in between you and a burned up processor, set it low; if you want to forego the safety feature, set the temp high. I know that MANY of you wished that AMD had the courtesy to include a feature such as this instead of leaving their Athlon and Duron CPUs totally unprotected.
We have never seen nor heard of the CPU throttling being active on any person's CPU and certainly have not experienced it ourselves (unless we FORCED it to happen) under conditions more strenuous than 99.9% of the P4s in the field will ever encounter. I do not suggest that DIYers or hobbyist go the P4 route if they want to buy a system for themselves, but bashing it on this front is simply bad journalism and transparent to many people.
We here at the [H] have a lot of respect for Bert McComas' work and think he should step back up to the plate and possibly rephrase the statements in regards to this issue. Bert, we love you man, but you were just totally out to show Intel in a bad light this time, or were simply not thinking through the issue properly because you are being misleading and it looks to us as if you were trying to do it purposefully.
UPDATE: I have gotten a few emails that have told me that I don't understand "clock throttling". I am fully aware of what it is, how it works, and how to identify if it is being used. Just for the record. https://www.sharkyforums.com/images/.../2005/06/5.gif My point here is not that what Burt is saying is false, but he is making it sound as if anyone that ever uses their P4 to its fullest potential will immediately be clock throttled and that is far from the case.
I couldn't agree more. Some of the flamewars in the CPU forum a few weeks ago were ridiculous. It got to the point where AMD and Intel could not be mentioned without howls of indignation. For whatever reason, processors are VERY important to some people, to the point where they act in a very illogical and uncivilized manner while defending their particular brand. I realize you are also talking about Inquest, etc...but it certainly applies to the forums as well.Quote:
I don't like bashing of Intel or AMD. I don't like it when hardware becomes political. What differentiates this debacle from a senatorial election? Try to get more dirt on the opposing party, while making yourself look good. Isn't that what InQuest is doing on behalf of AMD? Or DDR? It seems they have made it their quest to spread FUD on Intel, and I don't like that any more than someone spreading FUD on AMD.
It is virtually impossible to make someone change allegiance from one processor to another by simply telling them how great the alternative is. Another thing that seems to create a great deal of difficulty is speculation regarding future technologies. Palamino is a good example. Though some details seem to have leaked out, there is no way to guage its performance until it is in the hands of reviewers and the public.
Though I've only made a few posts, I've been a regular reader of the forums since I joined in October and you, Arcadian, have been outstanding in your efforts to initiate discussion on various topics and in staying away from pointless flames. Keep up the good work! https://www.sharkyforums.com/images/.../2005/08/2.gif
Scott
Thanks for the complement and the resonse, Scott. And thanks for the update from our friends at HardOCP. It's nice to see this issue is getting the right attention. Whatever has gone on, I'm sure we'll all get to the bottom of it. https://www.sharkyforums.com/images/.../2005/06/5.gifQuote:
Originally posted by scottd:
I couldn't agree more. Some of the flamewars in the CPU forum a few weeks ago were ridiculous. It got to the point where AMD and Intel could not be mentioned without howls of indignation. For whatever reason, processors are VERY important to some people, to the point where they act in a very illogical and uncivilized manner while defending their particular brand. I realize you are also talking about Inquest, etc...but it certainly applies to the forums as well.
It is virtually impossible to make someone change allegiance from one processor to another by simply telling them how great the alternative is. Another thing that seems to create a great deal of difficulty is speculation regarding future technologies. Palamino is a good example. Though some details seem to have leaked out, there is no way to guage its performance until it is in the hands of reviewers and the public.
Though I've only made a few posts, I've been a regular reader of the forums since I joined in October and you, Arcadian, have been outstanding in your efforts to initiate discussion on various topics and in staying away from pointless flames. Keep up the good work! https://www.sharkyforums.com/images/.../2005/08/2.gif
Scott
------------------
Grand Knight
OC Crusaders
ok, good to know, a power management feature is a good thing, because it's prevents the user from burning a CPU within 10 seconds in case the HSF is not placed correctly, which is a major problem for AMD CPUs, aside from fragility.Quote:
Originally posted by Arcadian:
The whole point is that this is NOT true.
but i have to critisise a few points:
1)the P4 has future potential, especially in bandwith consuming applications, also, a recompliation can do wonders in terms of performance (see Flask benchmark recompilation), but today's applications do not take advantage of P4's architecture, and i don't think we'll see recompiled editions of every single piece of software, which means that most of the P4's advantages won't be used for quite some time. this is NOT Intel's fault, of course.
2)Willamette's socket is a dead end, which will prevent current P4 users from upgrading.
3)Price/performance is ridiculous, 2 Athlon DDR systems for the price of one P4 rig. in many cases, the Vcard limits fps to 100 or below, professional users have multi-CPU systems.
i don't want to see Intel go bust, but today's circumstances indicate that it's just a question of time.
Incidently, I went to a lecture at the Oregon Center for Advaced Technology Engineering. It was about the Netburst architecture and was by an Intel engineer. Somebody asked about compiler optimizations, and he responded that Intel is currently working with Microsoft, and that a lot of Pentium 4 optimizations will find their way in the next version of Visual Studio, 7.0. Since the compilers in Studio are used the industry over, I'm sure there will be plenty of software support for Willamette once this release becomes available. https://www.sharkyforums.com/images/.../2005/06/5.gifQuote:
Originally posted by Captain Iglo:
1)the P4 has future potential, especially in bandwith consuming applications, also, a recompliation can do wonders in terms of performance (see Flask benchmark recompilation), but today's applications do not take advantage of P4's architecture, and i don't think we'll see recompiled editions of every single piece of software, which means that most of the P4's advantages won't be used for quite some time. this is NOT Intel's fault, of course.
------------------
Grand Knight
OC Crusaders
Quote:
Originally posted by Captain Iglo:
is the part about the P4's integrated power management true? limited to 54 watts? (with standard cooling and room temp 20°, better cooling pays off i presume) i am not exactly sure people won't notice that their 'highend' P4 1.7GHz is just as fast as the a 1.5...
Intel is going to have a problem: angry customers without an upgrade option who paid big bux for a comparably slow rig, Willamette is a rip-off and if they continue to abuse their brand name and peddle snake-oil, they'll be out of business by 2002.
[This message has been edited by Captain Iglo (edited April 15, 2001).]
Peddling snake oil? Out of business by 2002? I'm sorry for your ignorance, but a company that posted 6.7 BILLION dollars in revenue for FIRST QUARTER 2001, and is spending 4.2 BILLION on R&D in 2001 does not have such a grim outlook that you propose. They have the cash, and the smarts to put out a good product, it is people like you who need to pull the blinds up and look at the real world. lol
Also- I've been watching the forum for a while, just recently started posting. I like the way Arcadian avoids the flame! It's refreshing. I could care less each way what someone roots for, but senseless, unsubstatiated comments like the one I just replied to are irritating to say the least.
[This message has been edited by jen4950 (edited May 06, 2001).]
Really interesting string here... I have to admit that i flame the other cpu's all the time, but i'm constantly switching between intel and AMD. Whoever produces the lesser cpu gets flamed.
IMHO the athlon is the better cpu. either way you look at it. Price or performance.
I was reading that the latency with sdram can skip up to 4 cycles when it switches to read and write or vis versa, but on average the rdram has much less latency then the sdram. Is that not true? tell me if its not.
As for this whole cpu throttling. Dont most athlon mainboards have some kind of cpu throttling, regardless if the cpu has any kind of throttling built in or not. I know my motherboard does (a7m266). The P4 has it built in, so it throttles it down to 50%, that would still fry the cpu if you didnt have the heat sink properly put on. I've personally NEVER fried an athlon. I've built so many, that i've lost count.
I know a lot of people have had problems with cracking or burning up an athlon cpu. but most of these cases are people to used to posting their systems without fans on the cpu. and as for the cracking... thats more the user's fault. Its like saying a brand of car sucks because the clutch went out in the first week. stop putting it in reverse on the freeway?
Not saying the p4 is a bad cpu, just an inferior to the current alternative. When it comes to this whole bandwidth use argument. Todays applications dont use/dont need that much bandwidth. they really dont even need as much as the DDR athlons provide.. yet. someone said a while back they dont use 75% of the bandwidth. the figure isnt that high i'm sure, but applications and hardware dont need that much bandwidth. Its not really the bottle neck with current configs. maybe when we start seeing higher then 2 ghz cpu's....
And for the statement that the Tbirds didnt have an acceptable cooling attachment to the cpu (still using the socket 7 style) Every athlon motherboard i've used (mostly asus) has 4 holes around the cpu. There's quite a few cooling options that use these holes with screws. This solution at least doesnt require you to get a new case form factor for your new system.
just my 2 bits... dont take this as a flame. Just my opinion based on the facts I know.
[This message has been edited by Mr.Squish (edited May 09, 2001).]
The p4 is too expensive, and in some ways, intel shot themselves in the foot with RDRAM, but as more software supports the new instructions in the future, it will definately be worth it. The Athlon T-Bird is faster in business apps, but the p4 is better in games. I read in Maximum PC that intel is going to change the socket design and pin count for the p4 soon. Anyone have comments on this?
------------------
P III 1000
512 megs SDRAM @139mhz
Asus CUSL2
2 36bg scsi160 in Raid 0
GeForce2 Ultra
SB Live Platinum 5.1
TDK VeloCD 16x
Sony GDM F-500
Win2k / mandrake
The Athlon is definitely better at legacy code, however the P4 is a much more advanced design and should outperform the Athlon on most "new" software. This is already being seen in new games and 3D benchmarks (which are usually the first to take advantage of new techniques) and the trend will almost certainly continue.Quote:
Originally posted by Mr.Squish:
Really interesting string here... I have to admit that i flame the other cpu's all the time, but i'm constantly switching between intel and AMD. Whoever produces the lesser cpu gets flamed.
IMHO the athlon is the better cpu. either way you look at it. Price or performance.
I was reading that the latency with sdram can skip up to 4 cycles when it switches to read and write or vis versa, but on average the rdram has much less latency then the sdram. Is that not true? tell me if its not.
As for this whole cpu throttling. Dont most athlon mainboards have some kind of cpu throttling, regardless if the cpu has any kind of throttling built in or not. I know my motherboard does (a7m266). The P4 has it built in, so it throttles it down to 50%, that would still fry the cpu if you didnt have the heat sink properly put on. I've personally NEVER fried an athlon. I've built so many, that i've lost count.
I know a lot of people have had problems with cracking or burning up an athlon cpu. but most of these cases are people to used to posting their systems without fans on the cpu. and as for the cracking... thats more the user's fault. Its like saying a brand of car sucks because the clutch went out in the first week. stop putting it in reverse on the freeway?
Not saying the p4 is a bad cpu, just an inferior to the current alternative. When it comes to this whole bandwidth use argument. Todays applications dont use/dont need that much bandwidth. they really dont even need as much as the DDR athlons provide.. yet. someone said a while back they dont use 75% of the bandwidth. the figure isnt that high i'm sure, but applications and hardware dont need that much bandwidth. Its not really the bottle neck with current configs. maybe when we start seeing higher then 2 ghz cpu's....
And for the statement that the Tbirds didnt have an acceptable cooling attachment to the cpu (still using the socket 7 style) Every athlon motherboard i've used (mostly asus) has 4 holes around the cpu. There's quite a few cooling options that use these holes with screws. This solution at least doesnt require you to get a new case form factor for your new system.
just my 2 bits... dont take this as a flame. Just my opinion based on the facts I know.
[This message has been edited by Mr.Squish (edited May 09, 2001).]
Three years from now you can expect a 1.7 GHz P4 to be faster then a 1.5 GHz Athlon for almost all performance sensitive applications. If you don't plan to keep your processor more then a year or so then the Athlon really is faster. If you plan on keeping it longer the P4 will likely be faster in the long run.
RDRAM has a lot of advantages, but there are some tasks that will almost always run better on SDRAM. The problem is that these tend to be things like MS office that aren’t really all that sensitive to performance to begin with. I predict that when we get DDR for the P4 it will surpass the Athlon in sysmark (a benchmark the Athlon currently dominates) but in most cases gaming performance will suffer.
------------------
Life is complex. It has real and imaginary parts.
hehe, true true. I'll admit to that last post. But anyone that reads/posts in the "highly technical pc forum" prolly isnt gonna be keeping the same cpu for 1 year, let a lone 3 years.
Regarding the advantages of RDRAM over DDR, could you share your knowledge? Thanks.
Quote:
Originally posted by Moridin:
The Athlon is definitely better at legacy code, however the P4 is a much more advanced design and should outperform the Athlon on most "new" software. This is already being seen in new games and 3D benchmarks (which are usually the first to take advantage of new techniques) and the trend will almost certainly continue.
Three years from now you can expect a 1.7 GHz P4 to be faster then a 1.5 GHz Athlon for almost all performance sensitive applications. If you don't plan to keep your processor more then a year or so then the Athlon really is faster. If you plan on keeping it longer the P4 will likely be faster in the long run.
RDRAM has a lot of advantages, but there are some tasks that will almost always run better on SDRAM. The problem is that these tend to be things like MS office that aren’t really all that sensitive to performance to begin with. I predict that when we get DDR for the P4 it will surpass the Athlon in sysmark (a benchmark the Athlon currently dominates) but in most cases gaming performance will suffer.
I surmise most folks here keep their CPU for more than a year. That isn't to say that they don't purchase another CPU every year. My stable of CPUs date from from 1978 through 1999. If a box still does its intended job, why throw it away? Sure, I'll buy a couple of Foster's as soon as they are released - supposedly this month.Quote:
Originally posted by Mr.Squish:
hehe, true true. I'll admit to that last post. But anyone that reads/posts in the "highly technical pc forum" prolly isnt gonna be keeping the same cpu for 1 year, let a lone 3 years.
------------------
Live life as if money was of no concern.
Life is good. :cool:
I meant that in fairly generic terms. It has higher per pin bandwidth, more open pages (potentially), It makes more efficient use of bandwidth, It has additional power saving features (most of which aren’t used in PC memory). The type of thing you are probably already familiar with.Quote:
Originally posted by Conrad Song:
Regarding the advantages of RDRAM over DDR, could you share your knowledge? Thanks.
In the long run it's the bandwidth that really counts. Latency will still be higher then DDR or SDR in many cases, but the same was true of EDO Vs SDRAM. Here, higher latency, higher bandwidth SDRAM won out even though it had slightly higher latency.
Ultimately, I do not think it is practical to deal with main memory latency as a hardware problem. As CPU's become faster, the apparent latency in clock cycles seen by the CPU increases. There doesn’t seem to be any way memory latency can keep up with processor clock speeds.
On the other hand if we place the responsibility of dealing with latency on the application, then the hardware designer is left with the much simpler task of increasing bandwidth.
As the disparity between memory speed and processor speed grows this is really the only way to take advantage of these faster processors. In this context the generally higher latency of RDRAM isn't a big problem, while its greater bandwidth is a big benefit.
------------------
Life is complex. It has real and imaginary parts.