Both chips are good and both have issues. The P4 has the potential to be a great chip. Give it time and it will come into its own.for a 400mhz differance this is pathetic!
Well, I guess you guys don't know about the "power management" feature of P4's. Basically, if the processor heats up too much, it cuts the clock speed by half.
It runs for 2 ns, and rests for another 2 ns until the processor cools down to set temperature(automatic mode) If you look at P4's spec sheet: http://developer.intel.com/design/penti ... 919803.pdf , that's what it says - look under 7.3 "Thermal monitor". It also supports 12.5% on/ 87.5% off to 87.5% on/12.5% off in increments of 12.5%. I'd guess you'd have to change those settings in BIOS if it supports it.
That's one of many reasons why I don't like P4 - rambus latency, too deep pipeline, etc, already mentioned. They implemented this feature to keep the power disspation to something like 57Watts. If it were turned off, it'd run as hot or hotter than T-birds.
I'm not sure how often that happens, tho. I remember reading some site that said the maximum power disspation is actually something like 75Watts w/o the power management feature. So this may happen fairly often in gaming, I'd guess.
It runs for 2 ns, and rests for another 2 ns until the processor cools down to set temperature(automatic mode) If you look at P4's spec sheet: http://developer.intel.com/design/penti ... 919803.pdf , that's what it says - look under 7.3 "Thermal monitor". It also supports 12.5% on/ 87.5% off to 87.5% on/12.5% off in increments of 12.5%. I'd guess you'd have to change those settings in BIOS if it supports it.
That's one of many reasons why I don't like P4 - rambus latency, too deep pipeline, etc, already mentioned. They implemented this feature to keep the power disspation to something like 57Watts. If it were turned off, it'd run as hot or hotter than T-birds.
I'm not sure how often that happens, tho. I remember reading some site that said the maximum power disspation is actually something like 75Watts w/o the power management feature. So this may happen fairly often in gaming, I'd guess.
b1st, good find..but it really means nothing to us.
READ in bold print in the middle of that page
"Automatic mode is required for the processor to operate within specifications and must first be enabled via BIOS. Once automatic mode is enabled the TCC will activate only when the interal die temperature is very near the temperature limits of the processor."
Meaning it ISN'T automatically enabled, because it has to be enabled first. This is great for companies like Dell, Compaq and such, so people dont burn their cpu's up. Plus all the benchmarks ran by all these websites out here say/show nothing about this, so apparently it's nothing to worry about
Also it says "regardless of the automatic or On-Demand modes, in the event of a catastophic cooling failure, the processor will automatically shut down when the silicon has reached a temperature of approximately 135C.
This sounds like a great thing to me??? I dunno about you, but i'm pretty damn sure i'll never hit the silicon heat limit of 135C!!!!!!! And if it ever did, i'd be glad the cpu shut off than burning up like some KCF extra crispy chicken.
READ in bold print in the middle of that page
"Automatic mode is required for the processor to operate within specifications and must first be enabled via BIOS. Once automatic mode is enabled the TCC will activate only when the interal die temperature is very near the temperature limits of the processor."
Meaning it ISN'T automatically enabled, because it has to be enabled first. This is great for companies like Dell, Compaq and such, so people dont burn their cpu's up. Plus all the benchmarks ran by all these websites out here say/show nothing about this, so apparently it's nothing to worry about
Also it says "regardless of the automatic or On-Demand modes, in the event of a catastophic cooling failure, the processor will automatically shut down when the silicon has reached a temperature of approximately 135C.
This sounds like a great thing to me??? I dunno about you, but i'm pretty damn sure i'll never hit the silicon heat limit of 135C!!!!!!! And if it ever did, i'd be glad the cpu shut off than burning up like some KCF extra crispy chicken.
Actually, if you the sentence one right above that, it says "Thermal monitor uses two modes to activate the TCC. Automatic Mode and On-Demand Mode"
Which means, if you want to use automatic mode rather than on-demand mode, it has to be enabled in the bios. Otherwise, you're in on-demand mode.
However, TCC has to be enabled via the BIOS or in software first - this, I didn't know when I was posting the first reply. My assumption is that the BIOS's used in P4 boards turns it on because if it wasn't abled, the heatsink/fan that comes with the retail P4's can't possibly disspate all that heat (seventy some watts); will result the 135C sensor to kick in fairly often.
This introduces a possibility of a little utility that can disable TCC. As more people become aware of this, there will be one. I don't know x86 assembly enough to write such untility nor would I care to if I were able to. But then that'd be pointless since all you have to do is keep the processor very cool, then TCC modulation won't kick in at all.
Not many people know about the thermal management feature so I don't think the BIOS manufacturers even include the option to disable TCC. It's kinda like, back in the days, no BIOS/jumpers enabled the change of CPU core voltages because there was no demand for it.
Those two mode names are a bit misleading cuz both are automatic. It's matter of 50/50 or 12.5/87.5 in 12.5 increments.
As far as no one mentioning this feature, here's an outstanding article by InQuest Market Research: http://www.inqst.com/articles/p4bandwid ... thmain.htm , go down to where it says "Pentium 4 Power Management – A Performance Limiter" Again, I don't think most of the reviewers know about this.
As far as shutting down at 135C, or 130 or whatever, it's a whole different unit that does this. P3 uses this exact same unit, it's nothing new.
It'd be nice is someone with a P4 could test out my theory. Try using the retail heatsink/fan and run some tests; then, try using bigger/better combo and running the same test while keeping all other variables constant. It'd be nice if you've got some decent CPU temp monitor so you can see if the different heatsink/fan makes any diff in temperature first. I think the modulation kicks in at 35 or 40 C, I can't remember.
Which means, if you want to use automatic mode rather than on-demand mode, it has to be enabled in the bios. Otherwise, you're in on-demand mode.
However, TCC has to be enabled via the BIOS or in software first - this, I didn't know when I was posting the first reply. My assumption is that the BIOS's used in P4 boards turns it on because if it wasn't abled, the heatsink/fan that comes with the retail P4's can't possibly disspate all that heat (seventy some watts); will result the 135C sensor to kick in fairly often.
This introduces a possibility of a little utility that can disable TCC. As more people become aware of this, there will be one. I don't know x86 assembly enough to write such untility nor would I care to if I were able to. But then that'd be pointless since all you have to do is keep the processor very cool, then TCC modulation won't kick in at all.
Not many people know about the thermal management feature so I don't think the BIOS manufacturers even include the option to disable TCC. It's kinda like, back in the days, no BIOS/jumpers enabled the change of CPU core voltages because there was no demand for it.
Those two mode names are a bit misleading cuz both are automatic. It's matter of 50/50 or 12.5/87.5 in 12.5 increments.
As far as no one mentioning this feature, here's an outstanding article by InQuest Market Research: http://www.inqst.com/articles/p4bandwid ... thmain.htm , go down to where it says "Pentium 4 Power Management – A Performance Limiter" Again, I don't think most of the reviewers know about this.
As far as shutting down at 135C, or 130 or whatever, it's a whole different unit that does this. P3 uses this exact same unit, it's nothing new.
It'd be nice is someone with a P4 could test out my theory. Try using the retail heatsink/fan and run some tests; then, try using bigger/better combo and running the same test while keeping all other variables constant. It'd be nice if you've got some decent CPU temp monitor so you can see if the different heatsink/fan makes any diff in temperature first. I think the modulation kicks in at 35 or 40 C, I can't remember.
-
Hipnotic_Tranz
- Almighty Member
- Posts: 3750
- Joined: Wed Nov 22, 2000 6:35 am
- Location: Indpls, IN
- Contact:
Oh boy, I really shouldn't do this but it's been a while since I've been in an argument. Although, these are just my thoughts and don't mean jack to any of the rest of you
The point still remains, no matter what the hell the architecture of the P4, it clock for clock didn't win but a few benchmarks which is pathetic. I don't care if its new technology and you should let it break in or not; as of right now, the damn thing costs way too much for such little performance increase.
This is where I get upset; why do people support a processor that is not only slower (clock for clock in todays applications) but also requires specific RAM (which costs a lot by the way) and on top of that, is the most expensive desktop CPU out? It just doesn't make any sense. Not when there are alternatives (AMD) which is cheaper, faster (clock for clock again), and can use either SDRAM or DDR-RAM. At the point I am at right now, there is no reason I would go with a P4 over the T-bird. I can see people supporting the P3, atleast in the benchmarks it seemed comparable to the Athlon, but this? How can people be in such denile about how this processor (at this current time) just sucks?
The point still remains, no matter what the hell the architecture of the P4, it clock for clock didn't win but a few benchmarks which is pathetic. I don't care if its new technology and you should let it break in or not; as of right now, the damn thing costs way too much for such little performance increase.
This is where I get upset; why do people support a processor that is not only slower (clock for clock in todays applications) but also requires specific RAM (which costs a lot by the way) and on top of that, is the most expensive desktop CPU out? It just doesn't make any sense. Not when there are alternatives (AMD) which is cheaper, faster (clock for clock again), and can use either SDRAM or DDR-RAM. At the point I am at right now, there is no reason I would go with a P4 over the T-bird. I can see people supporting the P3, atleast in the benchmarks it seemed comparable to the Athlon, but this? How can people be in such denile about how this processor (at this current time) just sucks?
[align=center]<img src=http://i54.tinypic.com/j9tydf.gif>
<i>
My get up and go
must have got up and went.
</i>[/align]
<i>
My get up and go
must have got up and went.
</i>[/align]
This could just go back to the same argument i used earlier. Why do people buy Lexus when you can get a Toyota? They are the exact same cars, performace is the same, yada yada yada. I don't think anyone is in denial. DDR ram support for the P4 is coming out soon from mobo vendors such as via and the likes. Intel has said they will have DDR mobo's for the p4 in Q1 of '02.
I really don't see how this processor "sucks" as some people think. None of you have actually used it? Just because it doesnt win every benchmark is stupid! How are you going to notice the difference between 2-3 frames per second? How many nanoseconds of MS Outlook opening are you going to notice?
I really don't see how this processor "sucks" as some people think. None of you have actually used it? Just because it doesnt win every benchmark is stupid! How are you going to notice the difference between 2-3 frames per second? How many nanoseconds of MS Outlook opening are you going to notice?
-
Hipnotic_Tranz
- Almighty Member
- Posts: 3750
- Joined: Wed Nov 22, 2000 6:35 am
- Location: Indpls, IN
- Contact:
I think cars are different in the sense that people buy cars because:
1) it shows they have the money for that particular type of car
2) it's a social status thing
As with a CPU, most of the techies don't care which way "car" or CPU they use, as long as it performes well for what they use. I personally am a gamer and an amateur video editor (just started really so this isn't a big deal right now) and if a processor that is 400MHz faster (on top of being more expensive in the long run) is performing just as well as a slower one, common scense would tell me to go with the slower one.
I guess I shouldn't say the P4 "sucks" but it's kinda like Dakitana. It had a lot of hype, but never did real good and there are still those few that wanna defend it. I'm sure Dakitana is a good game in some of the visuals and the time it took (whats on paper) but what it shows in the real world is different.
(I would like to point out I have never played Dakitana or used a P4--my opinions are merely based on what I've heard repeatidly from people which makes my comments "invalid", to a point)
1) it shows they have the money for that particular type of car
2) it's a social status thing
As with a CPU, most of the techies don't care which way "car" or CPU they use, as long as it performes well for what they use. I personally am a gamer and an amateur video editor (just started really so this isn't a big deal right now) and if a processor that is 400MHz faster (on top of being more expensive in the long run) is performing just as well as a slower one, common scense would tell me to go with the slower one.
I guess I shouldn't say the P4 "sucks" but it's kinda like Dakitana. It had a lot of hype, but never did real good and there are still those few that wanna defend it. I'm sure Dakitana is a good game in some of the visuals and the time it took (whats on paper) but what it shows in the real world is different.
(I would like to point out I have never played Dakitana or used a P4--my opinions are merely based on what I've heard repeatidly from people which makes my comments "invalid", to a point)
[align=center]<img src=http://i54.tinypic.com/j9tydf.gif>
<i>
My get up and go
must have got up and went.
</i>[/align]
<i>
My get up and go
must have got up and went.
</i>[/align]
Ok, let's avoid hte bashing and get back to some tech...
My issues with the P4
1 SSE2 - How is this going to save the P4? I don't understand how advanced matrix manipulation will help in microsoft office. I mean, the Athlon destroys the P4 in office benchmarks, and that is a good reflection on intense use of the OS.
2 Bigger bandwidth=better - This is the biggest crock of all. AMD and Intel want us to believe that we need more bandwidth. Bull. We need more general purpose registers. Once you have more than 8 GPR's, memory bandwitdh becomes less of an issue.
3 P4 has some serious ALU issues - Ok, the P4 has 2 double-pumped ALU's. So how the hell can the athlon beat the P4 at all? That makes no sense. I read a test where a guy wrote x86 assembly loops, and the P4 performed horribly. Why is that? Shouldn't it be executing 4 instructions/clock, if the loop is long enough?
If anyone can clear this stuff up for me, show me an error in my thinking, or enlighten me if I am somehow ignorant, I will respect the P4. Until then, I would rather not use a P4.
p.s. - RISC is where it is AT.
edit: Never having used a P4 has no bearing on my judgement of its technical merit. Attacks ad hominem demonstrate a fundamental ignorance of the subject matter at hand.
[Edited by poop on 04-28-2001 at 04:19 PM]
My issues with the P4
1 SSE2 - How is this going to save the P4? I don't understand how advanced matrix manipulation will help in microsoft office. I mean, the Athlon destroys the P4 in office benchmarks, and that is a good reflection on intense use of the OS.
2 Bigger bandwidth=better - This is the biggest crock of all. AMD and Intel want us to believe that we need more bandwidth. Bull. We need more general purpose registers. Once you have more than 8 GPR's, memory bandwitdh becomes less of an issue.
3 P4 has some serious ALU issues - Ok, the P4 has 2 double-pumped ALU's. So how the hell can the athlon beat the P4 at all? That makes no sense. I read a test where a guy wrote x86 assembly loops, and the P4 performed horribly. Why is that? Shouldn't it be executing 4 instructions/clock, if the loop is long enough?
If anyone can clear this stuff up for me, show me an error in my thinking, or enlighten me if I am somehow ignorant, I will respect the P4. Until then, I would rather not use a P4.
p.s. - RISC is where it is AT.
edit: Never having used a P4 has no bearing on my judgement of its technical merit. Attacks ad hominem demonstrate a fundamental ignorance of the subject matter at hand.
[Edited by poop on 04-28-2001 at 04:19 PM]
- p o o p
Just so you know I wasn't supporting the P4; I was simply preventing people from believing stupid reasons for hating Intel. I have a P3 now and my next chip will also be an Intel one simply because I am too cheap to spring for a new motherboard.This is where I get upset; why do people support a processor that is not only slower...
If my board were to die however, I would definitely but an AMD rig right now. Of course I own both stocks, so what do I care as long as you don't by Cyrix! On a long enough time line, the survival rate for everyone drops to zero.
-
Kakarot
- Golden Member
- Posts: 1713
- Joined: Wed Nov 22, 2000 8:26 am
- Location: Chicago Land Area
- Contact:
Uhh... unless you're going to upgrade to a faster p3, no matter what chip you get next you'll have to get a new mobo. Whether its the athlon, p4 or p5 or next athlon or whatever.I have a P3 now and my next chip will also be an Intel one simply because I am too cheap to spring for a new motherboard
But lets go with this... you want to buy a P4. You look into it and find out that you have to buy the CPU, the Rambus RAM, Special Power Supply that has a 12v power connector for the cpu alone, Motherboard of course... oh and lets not forget new case since the way the coolers for P4's are mounted is via screws at the corners that go thru holes on the mobo into the case underneath. hmm yea, good choice considering it offers no real value or performance compared to its competition.
"Why build only one when you can build two for twice the price?"
<a href="mailto:murphy@excaltech.com">Email</a>
<a target=NEW href="http://www.heatware.com/eval.php?id=377">Heatware evals</a>
<a href="mailto:murphy@excaltech.com">Email</a>
<a target=NEW href="http://www.heatware.com/eval.php?id=377">Heatware evals</a>
I'm a poor college kid. -
Hipnotic_Tranz
- Almighty Member
- Posts: 3750
- Joined: Wed Nov 22, 2000 6:35 am
- Location: Indpls, IN
- Contact:
Splitfire - I said "why do people support the P4..." not why does "splitfire support the p4..." It was just a generalization. I don't believe you have to sit at a computer with a P4 to be able to say if it's good or not. Have I driven a Dodge Viper? No, but I still know it's a badass car.
It was just a generalization. I don't believe you have to sit at a computer with a P4 to be able to say if it's good or not. Have I driven a Dodge Viper? No, but I still know it's a badass car.[align=center]<img src=http://i54.tinypic.com/j9tydf.gif>
<i>
My get up and go
must have got up and went.
</i>[/align]
<i>
My get up and go
must have got up and went.
</i>[/align]
Well, in defense of my statement(s).....I buy what works NOW, not what somebody thinks has good "potential". I'll pass on "waiting" for devs to incorporate SSE2 and utilize some serious whoop-ass FPU of the T-birds. The fact of the matter is that *ntel is getting their ASS handed to them in a paper bag RIGHT NOW by AMD.
I freely bash *ntel because without the competition that has been created with the newer AMD parts, you'd be paying out the keester for any crap that *ntel deemed worthy to sell you. So even if you are an *ntel groupie you should fall down on your knees and thank AMD for at least keeping the prices low on the cheez that *ntel is trying foist on you. To which I say......." better you than me" Hell, my retro rig (see sig) outbenches a P4 and this thing is ancient. I'll be unleasing the newest version of "Jersey Devil" later this week that will show you what T-Bird based dominance looks like!
*ntel along with M$ have been gouging peeps for years and I ,for one, am happy to see at least one of them (*ntel) take it up the poop chute.
While maybe Nexus_7 is a bit disoriented ( :lol ) with regard to the financial aspects of the stock world, anyone who is paying attention to chip sector analysts would know that the street advice these days is be out of *ntel and into AMD before the year is out.
Gee, I wonder why????
Just my vidman2 cents
I freely bash *ntel because without the competition that has been created with the newer AMD parts, you'd be paying out the keester for any crap that *ntel deemed worthy to sell you. So even if you are an *ntel groupie you should fall down on your knees and thank AMD for at least keeping the prices low on the cheez that *ntel is trying foist on you. To which I say......." better you than me" Hell, my retro rig (see sig) outbenches a P4 and this thing is ancient. I'll be unleasing the newest version of "Jersey Devil" later this week that will show you what T-Bird based dominance looks like!
*ntel along with M$ have been gouging peeps for years and I ,for one, am happy to see at least one of them (*ntel) take it up the poop chute.
While maybe Nexus_7 is a bit disoriented ( :lol ) with regard to the financial aspects of the stock world, anyone who is paying attention to chip sector analysts would know that the street advice these days is be out of *ntel and into AMD before the year is out.
Gee, I wonder why????
Just my vidman2 cents
AMD Ph II X4 955 BE 3.2 @ 3.8 GHz | Scythe SCSMZ-2000 | ASRock 880GMH/USB3 | 8 GB G.Skill DDR3 1600 | Radeon HD5670 | Kingston 128GB SSD
Wow, this is going to be easy. First,
Now onto the fun stuff...
Sheesh, that makes me wonder what you would think about my 600E that I use everyday!my retro rig (see sig) outbenches a P4 and this thing is ancient...
Now onto the fun stuff...OK, I don't know where everyone on the net is getting these intensely screwed up ideas about what makes a monopoly, but INTEL IS NOT A MONOPOLY!! It never has been and it probably never will be. There are reasons why market share is not the only tool used by the DoJ when evaluating firms for possible nati-trust violations and that is because market share is INCREDIBLY misleading. Despsite popular misconception, MS was not ruled a monopoly because they were the most popular OS in the world. It was ruled a monopoly because it bundled IE with Win98 and made it almost impossible to uninstall. Note that this has absolutely NOTHING to do with putting out bad products or fixing prices. A year and a half ago, Intel was clobbering everyone because it had the best product available (e.g. AMD's K6-x chips blew for all intents and purposes) NOT because it was somehow keeping AMD from developing better chips.I freely bash *ntel because without the competition that has been created with the newer AMD parts, you'd be paying out the keester for any crap that *ntel deemed worthy to sell you...*ntel along with M$ have been gouging peeps for years and I ,for one, am happy to see at least one of them (*ntel) take it up the poop chute.
HAH! Have you been actually LOOKING at the market or are you just guessing? Because this statement shows clearly that you don't know what you are talking about. AMD has been DOWNGRADED along with Intel and the entire chip sector for the last 5 months or more. Intel has been upgraded in the last few weeks along with AMD and teh ENTIRE CHIP SECTOR. No broker in the world would tell you to not own Intel in the long run, nor would they tell you to NOT own AMD in the long run.anyone who is paying attention to chip sector analysts would know that the street advice these days is be out of *ntel and into AMD before the year is out.
On a long enough time line, the survival rate for everyone drops to zero.