Didn't I also read somewhere that the P4 has 20 checkpoints for executions to go through? Double that of a PIII making the time it takes to execute an exectution per clock cycle longer? Intel is making you believe that their P4 is faster by calling it a 1.7ghz, which it is, but it will actually run slower per clock cycle than cpu's with less mhz. The only thing I see the P4 has going is it's memory bandwith adjustsments, and I applaude Intel for addressing this issue. I've always said my dream CPU would be a Ghz with a Ghz FSB. But hey that's just me.
"it will actually run slower per clock cycle than cpu's with less mhz"
Yes, it will not do as much per clock cycle. This is by design, just as 2nd gear in a vehicle does not have as much power as 1st gear when running at the same speed. It sure scales up to much higher speeds though. Since performance is the product of clockspeed and average instructions per clock, if you lower one a little, you must raise the other a little in order to compensate and remain equal in performance. The clockspeed on the P4 is being raised considerably and still has a long way to go. I fully expect the P4 core to see 10GHz.
-Raystonn
= The views stated herein are my personal views, and not necessarily the views of my employer. =
Yes, it will not do as much per clock cycle. This is by design, just as 2nd gear in a vehicle does not have as much power as 1st gear when running at the same speed. It sure scales up to much higher speeds though. Since performance is the product of clockspeed and average instructions per clock, if you lower one a little, you must raise the other a little in order to compensate and remain equal in performance. The clockspeed on the P4 is being raised considerably and still has a long way to go. I fully expect the P4 core to see 10GHz.
-Raystonn
= The views stated herein are my personal views, and not necessarily the views of my employer. =
Maybe your not catching my drift, but apparently you work for Intel so you must know that the IPC x Mhz for the P4 is actually lower than the PIII and only on software that is optizimed for SSE2 does the P4 actually gain in performance. So as it stands right now, under your equation, that's right (mhz x instruction per clock cycle (IPC)) the Athlon is the victor. And that's coming directly from you.
this 70% improvement is with SSE2 code, Adobe must make AMD 3DNow! code and then compare the two. I am sure it would be almost at par with SSE2 if not better.
get real world day to day apps tests, field testing is more important than lab tests.
girish
<font color=blue>die-hard fans don't have heat-sinks!</font color=blue>
get real world day to day apps tests, field testing is more important than lab tests.
girish
<font color=blue>die-hard fans don't have heat-sinks!</font color=blue>
Adobe is already enhanced for 3d Now. Aparently AMD just cant make em like Intel can. Hmm monkey stew.
SPUDMUFFIN
<font color=red>Being Evil Is Good. Cause I Can Be A Prick And Get Away With It.</font color=red> :lol:
SPUDMUFFIN
<font color=red>Being Evil Is Good. Cause I Can Be A Prick And Get Away With It.</font color=red> :lol:
Fltsimbuff
Distinguished
- Mar 30, 2001
- 114
- 0
- 18,680
Using MHz to rate the speed of a CPU is like using rpms to rate the Power of an engine.
--Fltsimbuff
--Fltsimbuff
Raystonn, why do we have to discuss it over and over again? we are doing it already in another thread...
deeper pipelines introduce larger penalties when the instruction stream changes. once the pipes are flushed, the next instruction from the newer instruction stream has to go through all the 17 pipes before it is executed, and thats too much of a latency although the BTB does reduce it by a few cycles. data dependencies between instructions do not help either, one of the execution units has to wait for the previous instruction to complete. this reduces your "average instructions per clock"
if performance is the product of clockspeed and average instructions per clock, and as you say "if you lower one a little, you must raise the other a little in order to compensate and remain equal in performance" then what is the use of all these extra MHz anyway if the performance is to remain the same?
slower on clock but better on performance chips will surely be cheaper. and who doesnt want cheaper chips?
girish
<font color=blue>die-hard fans don't have heat-sinks!</font color=blue>
deeper pipelines introduce larger penalties when the instruction stream changes. once the pipes are flushed, the next instruction from the newer instruction stream has to go through all the 17 pipes before it is executed, and thats too much of a latency although the BTB does reduce it by a few cycles. data dependencies between instructions do not help either, one of the execution units has to wait for the previous instruction to complete. this reduces your "average instructions per clock"
if performance is the product of clockspeed and average instructions per clock, and as you say "if you lower one a little, you must raise the other a little in order to compensate and remain equal in performance" then what is the use of all these extra MHz anyway if the performance is to remain the same?
slower on clock but better on performance chips will surely be cheaper. and who doesnt want cheaper chips?
girish
<font color=blue>die-hard fans don't have heat-sinks!</font color=blue>
the clock for clock i know its not even a clock per clock because intel cant compite clock per clock anymore thats why they need a 1.7 gig to half beet a 1.333athlon
Computer Shop owner and Head tech.
Computer Shop owner and Head tech.
I cant cause i dont care but when photo 6 fires up it say extensions used if you can catch it it says 3dnow mmx sse simd yada yada.
SPUDMUFFIN
<font color=red>Being Evil Is Good. Cause I Can Be A Prick And Get Away With It.</font color=red> :lol:
SPUDMUFFIN
<font color=red>Being Evil Is Good. Cause I Can Be A Prick And Get Away With It.</font color=red> :lol:
Fur a "cumpewter snot owner" and "Hed Teck", can eye recommend a spelled/grammier check the next you posted hear??
<font color=blue>The #1 reason to upgrade your PC - to run faster benchmarks...</font color=blue>
<font color=blue>The #1 reason to upgrade your PC - to run faster benchmarks...</font color=blue>
AmdMELTDOWN
Distinguished
- Dec 31, 2007
- 2,000
- 0
- 19,780
>Adobe must make AMd 3dNow! code and then compare the two.
why, I thought the Athlon had "raw" power? and didn't need "optimization code" like the pee4 does.
"AMD...you are the weakest link, good bye!"
why, I thought the Athlon had "raw" power? and didn't need "optimization code" like the pee4 does.
"AMD...you are the weakest link, good bye!"
AmdMELTDOWN
Distinguished
- Dec 31, 2007
- 2,000
- 0
- 19,780
>tbirdinside your just as retarted as that other dipshit amdmeltdown.. Are you two brothers?
no everett6, you're retarted, you retard!
"AMD...you are the weakest link, good bye!"
no everett6, you're retarted, you retard!
"AMD...you are the weakest link, good bye!"
"the IPC x Mhz for the P4 is actually lower than the PIII"
Not even close. The Pentium III is currently maxxed at 1GHz. The Pentium 4 is currently maxxed at 1.7GHz. That is a 70% higher clockspeed. The IPC of the Pentium 4 is nowhere near 70% slower than Pentium III.
-Raystonn
= The views stated herein are my personal views, and not necessarily the views of my employer. =
Not even close. The Pentium III is currently maxxed at 1GHz. The Pentium 4 is currently maxxed at 1.7GHz. That is a 70% higher clockspeed. The IPC of the Pentium 4 is nowhere near 70% slower than Pentium III.
-Raystonn
= The views stated herein are my personal views, and not necessarily the views of my employer. =
"what is the use of all these extra MHz anyway if the performance is to remain the same"
The IPC was lowered a bit in the core in order to allow the clock speed to be ramped up _much_ more. The Pentium 4's current clock speed already more than makes up for the loss in average IPC. Over its lifetime it will continue to scale to _much_ higher clock speeds, and thus much better performance.
-Raystonn
= The views stated herein are my personal views, and not necessarily the views of my employer. =
The IPC was lowered a bit in the core in order to allow the clock speed to be ramped up _much_ more. The Pentium 4's current clock speed already more than makes up for the loss in average IPC. Over its lifetime it will continue to scale to _much_ higher clock speeds, and thus much better performance.
-Raystonn
= The views stated herein are my personal views, and not necessarily the views of my employer. =
"clock speed is totally irrelevant"
You guys it's not about clock speed or electronics or sparks or silicon! It's not about the technology, its that intel designed the technology around the marketing, not because they wanted to make a fascinating piece of hardware. Why else does the average person think clock speed is everything? Why would a company want to build their product around the assumptions of average consumers when their technicians know they can make something better that has real performance rather than TV commercials and magazine ads. Do you see what we mean it has NOTHING to do with technical stats, it has to do with honesty, personal beliefs, and integrity and because of that intel makes me quite upset like tbird was saying, so please do not try to make this all technical. computers have so many bugs and software problems so that's why we need people to help run them, and sometimes we forget that a team of designers made the motherboard you just bought or the memory or processor. Which team do you want to support? Are they real hardware enthusiasts or just people itching for some cash? Too many people choose money as their main goal, and they end up making lots but the products are so unintelligent and uninspired.. hardware should have a human touch to it, like someone spent time to make something good, or the person who made it is as enthusiastic about it as you are. See what I mean? When your only goal is design the entire chip around the marketing that doesn't work. It's like trying to make up a special scientific theory just because it'll make you famous. No you do it because you believe in it and it's your passion and you love computers or games or programming or whatever.
You guys it's not about clock speed or electronics or sparks or silicon! It's not about the technology, its that intel designed the technology around the marketing, not because they wanted to make a fascinating piece of hardware. Why else does the average person think clock speed is everything? Why would a company want to build their product around the assumptions of average consumers when their technicians know they can make something better that has real performance rather than TV commercials and magazine ads. Do you see what we mean it has NOTHING to do with technical stats, it has to do with honesty, personal beliefs, and integrity and because of that intel makes me quite upset like tbird was saying, so please do not try to make this all technical. computers have so many bugs and software problems so that's why we need people to help run them, and sometimes we forget that a team of designers made the motherboard you just bought or the memory or processor. Which team do you want to support? Are they real hardware enthusiasts or just people itching for some cash? Too many people choose money as their main goal, and they end up making lots but the products are so unintelligent and uninspired.. hardware should have a human touch to it, like someone spent time to make something good, or the person who made it is as enthusiastic about it as you are. See what I mean? When your only goal is design the entire chip around the marketing that doesn't work. It's like trying to make up a special scientific theory just because it'll make you famous. No you do it because you believe in it and it's your passion and you love computers or games or programming or whatever.
So whats your point? Should we boycott all hardware that didnt come from somebody's divine inspiration? In case you didnt know everybody is out to make a buck, or better yet a billion bucks AMD, Intel, Nvidia etc..... Show me an honest company and i'll show you a fool.
>Show me an honest company and i'll show you a fool.
What a sad outlook on life.
I believe it's possible to make a buck honestly.
In theory, there is no difference between theory and practice.
In practice, there is.
What a sad outlook on life.
I believe it's possible to make a buck honestly.
In theory, there is no difference between theory and practice.
In practice, there is.
All companies are honest. They are all honestly trying to make money. That's the point of capitalism and that's the point of business. Making better products usually makes you more money.
-Raystonn
= The views stated herein are my personal views, and not necessarily the views of my employer. =
-Raystonn
= The views stated herein are my personal views, and not necessarily the views of my employer. =
>All companies are honest. They are all honestly trying to make money.
That's a pretty skewed definition of honesty. You could ethically justify pretty much anything with reasoning like that.
And no, not all companies are honest/ethical.
In theory, there is no difference between theory and practice.
In practice, there is.
That's a pretty skewed definition of honesty. You could ethically justify pretty much anything with reasoning like that.
And no, not all companies are honest/ethical.
In theory, there is no difference between theory and practice.
In practice, there is.
Honest and ethical are two completely different words with two completely different meanings. Honesty implies that you disclose your intentions. Ethical implies that, whether you disclose your intentions or not, they do not violate any moral values. Not all companies are honest and not all companies are ethical. However, I don't see any examples of these types of companies in the CPU industry. Everyone pretty much knows where each of the major companies stands.
-Raystonn
= The views stated herein are my personal views, and not necessarily the views of my employer. =
-Raystonn
= The views stated herein are my personal views, and not necessarily the views of my employer. =
lol. Don’t get me wrong. I too believe it's possible to make a buck honestly. But I also believe it's very rare for large companies to stay honest. They all get greedy sooner or later. They're all scamming a buck here and a million bucks there. Ask any small business owner how hard it is to stay afloat without "cheating". That’s why many start-ups go out of business, they're too honest, and I’m not just talking about the computer industry. I know its sad.. but its so true.
There is so much marketing hype in the computer industry and in many cases it's no different then flat out lying about there products to generate good publicity. Unfortunately we let them get away with it, so now its common practice. Reading press statements is sometimes akin to reading a fairy tail. The marketing departments in this industry are quite pathetic.
from what i know the issue on this forum is to talk about computers not your nit picky self. as far as my spelling i fail to see what that has to do with business or knowing anything about computers. stick to the forum not your personal need to talk S**T.
Computer Shop owner and Head tech.
Computer Shop owner and Head tech.
TRENDING THREADS
-
-
-
-
Question Blackscreen and RAM debug light changed to VGA debug light - - - - Is it gone for good ?
- Started by SuperCitizen
- Replies: 5
-
-
Discussion What's your favourite video game you've been playing?- Started by amdfangirl
- Replies: 3K
Facebook
Twitter
Instagram