SpecInt/SpecFP - Intel vs AMD

Page 8 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
I suggest you read up a bit on the differences between Intel's SMP, and the upcoming AMD SMP solutions before you make such claims... AMD's SMP solution will be Far superior to Intel's current P3 SMP solution.

--Fltsimbuff
 
Oh brother, tonestar, of all people, sets himself up as the judge of "profesional".

Profesionals don't need to lie. (That's not directed @ you cyberimage)


In theory, there is no difference between theory and practice.
In practice, there is.<P ID="edit"><FONT SIZE=-1><EM>Edited by ergeorge on 04/17/01 04:20 PM.</EM></FONT></P>
 
That still would not explain why the p3 is beating the p4 as well clcok for clock now would it? Or dows the p3 have 3dnow instructions as well? It is well known that 3dnow has never been widely used our fully implemented, even if 3dstudiomax is somehow takeing advantage in some small way of 3dnow, its SSE coding is much stronger and should give the p4 an edge not a disadvantage, come on you are going to have to do better than that, your argument is becoming weaker and weaker.

p.s btw I have to leave now for a day and a half don;t feel that I have given up on this argument, perhaps another will fill in.

A little bit of knowledge is a dangerous thing!
 
You trust Seagate hard drives? I have a little anecdote for you.

When Seagate first released their Ultra160 drives in 10,000RPM and 7200RPM, they were riddled with problems, especially working with Ultra2 RAID controllers like the DPT SmartRAID V. At the same time, Seagate discontinued their Ultra2 production, which left a lot of people (us in particular) with their asses out in the wind. We were stuck with hard drives that were slow, unreliable, and had a very bad habit of losing or corrupting RAID data. 10K RPM Cheetahs weren't even close to stable until firmware revision 0004; 7200RPM Barracudas would last as long as a month before dying unless they were up to firmware rev 0010. We were stuck with problems for months, and Seagate initially gave us the run-around. Later they admitted that when we first confronted them about the problems, they had just discovered them as well and were frantically writing new firmware.

If you doubt any of this, check out one of MaximumPC's recent "Lemons of the Year" (I believe Lemon of 2000 ). Now you know the full story behind <i>that</i> lemon.

Be careful not trust anyone blindly based on a brand name. That can get you in a lot of trouble.

Kelledin
<font color=red>"Step away from the gimp suit and put your hands on top of your head."</font color=red>
 
You also forget that the executable is built and optimized for the pipelines of the Athlon and P3, and not the P4. The P4 will see many more branch mispredictions and stalls because of this.

If you want to compare, use a test that has been compiled with a P4-aware compiler (in addition to Athlon aware), and either uses both 3DNow _and_ SSE2, or neither.

The optimizations given by the proper ordering of instructions far outweighs the optimizations given by SSE2.

-Raystonn

-- The center of your digital world --
 
Actually, nVidia's drivers take advantage of both 3Dnow! and SSE.

Kelledin
<font color=red>"Step away from the gimp suit and put your hands on top of your head."</font color=red>
 
Like I have said before what good is the p4 if I have to recompile all my current software? If it is truly advanced it must be able to run software that is currently available not wait for every user to completely replace all the software they currently have. Sure, if it was designed right it could still implement sse2 but NOT AT THE COST OF BEING NOT BEING ABLE TO RUN TODAYS SOFTWARE AS FAST IF NOT FASTER than equally clocked cpu's from AMD or even for that matter from itself!

A little bit of knowledge is a dangerous thing!
 
"Like I have said before what good is the p4 if I have to recompile all my current software"

That's completely irrelevant to this discussion. We're discussing the performance of the CPUs unrelated to that of any specific software. In this fashion we'll be able to base comparisons off the results of our discussion and look at the present and into the future to see where we will go. I'd rather not restrict our discussion to only what's relevant in the software available at this precise second. I'm not trying to sell processors. I'm merely trying to analyze what's available and what will be happening in the future.

"NOT AT THE COST OF BEING NOT BEING ABLE TO RUN TODAYS SOFTWARE AS FAST IF NOT FASTER"

Not true. The laws of scaling indicate that as processors get faster, they will not necessarily still retain the ability to do as much _per clock_. This is an inevitable part of scaling up the clock speed. Certainly a vehicle in 1st gear can accelerate more quickly and has more power than a vehicle in 2nd gear that's still running at a speed attainable by 1st gear. In order to ramp up speeds you have to change the gear ratio, removing some of the power per RPM. That is, unless you wish to strip and blow your transmission.

-Raystonn

-- The center of your digital world --
 
well said and true you took the words right outta my mouth
best
CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
you are so LAME<
first you say I did not adress you point about proving the register is full of crap WHICH I THEN DID,
then you replay to that message with quoting tech stuff I quoted that you said you were not talking about and wanted to address the register issue..
make up you twisted mind,,, you sould nothing but confused...

CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
HI,

what you said was true of P3 and Athlon but NOT of P4 that is what makes it soo good..

for the first time, the P4 matches its 3.2 GPS memory bus,
with ram for rambus 3.2 GPS, throught dual channel and 400 mhz bus so it can communicate with system faster and thus is not the bottleneck..
also the P4's 850 MB chipset also supports 3.2 GPS through multiple paths etc

the Athlon and P3 were limited to 800-MPS to maybe 1 GPS,
because of RAM, CPU bus speed, and chipset..

this is what I have tried to explain that makes the P4 different, and faster, but by the same time, you have to
hammer the system to reach beyond the Athlon adn P3's
bottleneck to see how well P4 keeps going and leaves the others behind...

if you test all three with worthless simple looped benchmarks that do not saturate the system and show bottlenecks that the advantages of P4 and not demonstrated..

its like comparing an INDY car and a Toyota both doing 60 MPH onthe freeway and saying the INDY car to to expensive and not any faster,,

but take them up to 200 MPH and the INDY car leaves the TOYOTA behind past 130 MPH..

hope you find this interesting
best
CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
the new MS compilers will use P4 arranged instructions and support them...
them are comming along with Windows XP which is fully supported P4 code..
you are simply incorrect when you say Palimino will have a better structure and branch prediction unit..

Palimino is sole a reduced die size and voltage version of
the Athlon so as not to run so hot and burn up as they do..
there Palimino is merely reduced in size and voltage to as to allow high MHZ which the current athlon cannot do..
no major hardware changes are planned for it..

the P4 includes a hardware prefether, that can move instructions in and out of registers without disturbing cache, and way before it actually needs them !!

the Athlon would and does not dominate at the same clock speed, if you look at tom's tests involving an overclocked 1.6 athlon, some of them show the P4 faster the athlon at slower clock speed,
even though his tests are very flawed and he should know better, and not use dd3d 7 which has np P4 cade in it,
and he used many benchmarks that are a generation old like
SANDRA, and 3dMARK both of whihc contain no P4 code,
and he used games like unreal and toehrs whoses enginers are a year old and have no p4 code, or have timer limits in them unlike quake 3 which scales limitlessly..

so since I have run tests on P4 versions on these programs
I can tell you for a fact that P4 code makes a huge difference.. just as all you athlon nuts screamed when benchamrks showed P3 wailing on Athlon before 3d NOW was in
wide use..

fact is most credible benchamrks that are crossplatform
and apps that are compiled for P4 are much faster die to the superior architecture of the P4, its 400 mhz bus,
and dual channel rambus and its 3.2 GPS bandwidth..

no matter what you say in you opinion, you cannot change computer architecture facts like the ones above which are factually superior in bandwidth and speed
its like saying a V8 is slower than a V4 given equal circumstances
the fact you do not like or have a P4 does not make it true that is it not a better CPU than Athlon, or P3..
fact is I see and test dozens of machines a week,
and I have compared both and I am telling you P4 is the fastest thing I have even seen PERIOD

enough said
CAMERON


CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
that is the stupidest thing I have ever heard,
if something is superior it should not need optimization..

INDY cars are 1 million dollar cars that if the fuel air mixture is not optimized would be slower than a VW,
same with jet planes, and hell we would still be using DOS
if your theory holds true..
technology is one big optimization when compared to the original WHeel or mousetrap..

you are a real LAMER sometimes
CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
well it sure can on the P4,
the P4's memory bus, is perfectly matched with the rambus at 400 mhz and 3.2 GPS, and the dual channel design combined with the 850 chipset , and all matched in speed,
and throughput, and the dual channel design allows huge capabilities in bandwidth when compared to Athlon..
you can see this is memory benchmarks like SANDRA,
MEMTACH, and STREAM al of which I have run and they are 3-6 times faster than Athlon, due to athlon bottleck, of memory,
chipset and bus speed

CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
this is true, good point !!
the reason was simply and the same reason why P4 is so much faster than P3 or athlon..

the P3 and athlon are limited to about 1 GPS bandwidth and bus, as are their chipsets, so the superior rambus was hampered by their bottleneck..

the P4 has a 3.2 GPS bus at 400 mhz matched to rambus at the same 400, 3.2 GPS, matched to a 850 chipset that has dual channel memory paths, and can also support 3.2 GPS

thus the difference..
according to my memory throughout tests on memtach, SANDRA, and STREAM, the P4 is anywhere from 3-6 times faster than P3 or athlon !!

big difference...
best
CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
apparently you have not seen FOSTER..
I have seen a demo, and though I cannot say to much,
FOSTER in the dual P4 CPU that combined with the new 8xx
chipset, will crush athlon's SMP,
having in effect 2 x dual channel rambus or QUAD channel,
for effective 6.4 GPS..

Athlon,s SMP has still one lousy channel which both CPU's must bottleneck on....

Foster will be out in 3 months, Athlons SMP will not ship till 4th qtr at least, and then there is the issue of Windows XP's SMP kernal being optimized for INTEL's P4

not good for Athlon

CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
yeah thanks that is what I read too,
and these people were telling me that there way no
SSE or direct draw 3d in MAX...

jesus people should read more :)) ha ha

thanks for the post, I knew I was not seeing things :))
best
CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
I'm not going to doubt your story or even check out MaximumPC's article. It's unfortunante that your in this situation. However in my firms experience ( We have about 40 computers that are updated every 2 years) we specify seagate's. We have 14 Ultra160 drives and most of the others are Ultra2's, mostly barracuda's. We have seen them die, IBM's, WD's, Quantum's, Fujitsu's etc. We have still to see a seagate go down. I have a a 10k Ultra2 and a 10k Ultra160 in my machines at home and no probs there either. I'm not sure what kinda RAID setups we have at work but I can't remember them having any problems either. I infact had an old Barrracuda SCSI drive when I first started with the company 4 years ago and it use to bust its ass all day everyday (I only had 128MB ram). It's still going today. Admitadly I have always been working at standalone workstations. Why does your RAID controllers spit out so many hard drives?

"Cock-a-doodle-do" is what I say to my girl when I wake her UP in the morning!!
 
Err this is going quite out of the line he he .. Well just to set things straight on Cyberimages account. I just have to say that hmm his Company does not sell highend systems.. Well they do, 4 way servers and stuff, but still the Game Puter with a 1.5Ghz P4 is only supplied with 128mb ram and a measly Geforce2 GTS card. ICK! I wouldn't use less than 256mb ram and an Ultra card if I were to call it highend... I won't nag much about this since also the site is pretty poorly designed and has a lack of professionalism over it. Sorry Cyberimage, I don't do this to .. well I do.. he he .. I don't put much weight on yer arguments anyway. Besides, online computer news magazines, like the register, cnet eenews and so forth change oppinions like the wind changes directions. News are only news fer a limited ammount of time. And if one posts bollocks the others soon follows. Anyway the only reall workstation U at cyberimage sell is the Xeon one.. With elsa Quattro 3, the others are bollocks.. the register btw tries to write things in a funny way... Makes it more interesting to read than crappy cnet or eenews.. which are EXTREMELY boring.. allmost feels like being back to college.. Read read read , bla bla bla.. and the wisdom of today is the joke of tomorrow!

Tho... just to be totally back on the track. I have to agree with Rayston's original post on the SpecInt/FP.. As much as I hate to say it.. Intel has made a good processor, a bold move since it's dependent on recompile and reprogramming of apps to utilize the benifit. P4's are getting cheaper, DDR ram is getting cheaper, RDRAM is getting cheaper, but still there's quite a few bucks difference.. if it's enough to disquallify P4 ? dunno.. my system is fast enough, besides, a GeForce 3 will do more good to yer system than a 1.5 P4. Believe me. that is if yer a gamer, normal users should settle with a 800duron or something.. and a cheap graphics card.. Ohh that was a huge side track.

I'm Outa here..
(Don't think I said something sensible at all 🙂)
 
APPARENTLY
you did not read the web page correctly,
the ones shown are simply examples of some configurations,
the ram and options are limitless, just like DELL , GATEWAY,
HP and others..

we could not begin to list even option , we have build game machines with 1 GIG ram for ID games...

the web page shows machines FROM 128-2 GIGS or above..
we build machines with up to 16 GIGS ram !!

for graphics cards we offer and have built eventhing from GEFORCE 2, which is entry level, to 3D labs rendering cards costing $2000 +

some machines offer a 75 GIG IDE HD, but does that mean you have to get that one with that drive ..no
we can offer up to and routinly sell 1 TERABYTE RAID 5 systems as well, or
15000 rpm Ultra 160 HD's scsi for workstations

how about ultra 320 scsi which we are starting to test now?

you totally misunderstand the abilty of OPEN system design and OPTIONS, and most people are smart enough to realize
if they see a samply system but want more ram or video power or larger disks.

if you do no consider us offering ultra high performance, then you must not know what the word means
or you must just be trying to be a dick,, which is more likely..
anyway, not worth anymore of my time..
good luck in your misinterpretations
CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom