SpecInt/SpecFP - Intel vs AMD

Page 6 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Well then along as Rambus drops its prices, and AMD does not drop theirs, you may have an argument there, otherwise that still will not be the case. But lets address that issue when it, or if it happens.

A little bit of knowledge is a dangerous thing!
 
HI,
a few facts about benchmarks and P4\RAMBUS

the P4 Performance and the Latency rumor about rambus is grossly misreported by many who do not understand the overall picture about the P4 and rambus relationship..

a few interesting facts that the media does not report and tom missed as well..

the clock latency on rambus 800 is CL 2,
DDR is 2.5-3 depending..

also, as you said so well, the rambus has 4 paths,
and is clock doubled so this helps compensate for latency..

also, RAMBUS on the chipset 850 which has eveerything to do with memory performance, chipsets usually do,,
is a dual channel design which means they precharge
and can operate concurrently, so as to make rambus much faster..

lastly the P4 has a very deep pipeline and efficient cache
over the athlon that takes rambus design into account and actually compensates for rambus by its ability to reorder and reexecute instruction depending on the situation.

to give you an IDEA, I just ran our test P4 1.5 machine
on MEMTACH and SANDRA benchmarks,
and the DDR with athlon 1.2

in sandra athlon gets 500 INTEGER and 700 FP with PC2100 ddr
P4 gets 2200 or over 4 times faster in integer memory
and 4400 in FP memory with sse 2 enabled
or over 6 TIMES faster

in MEMTACH the results are similiar

RAMBUS scores over 1000 whereas the Athlon machines
are in the 300's

one of the most respected programs that is cross platform
testing Memory is STREAM by VIRGINA TECH EDU.

here are some posted scores including ours

SGI_Octane_300 375.3

Sun_Ultra60-360 355.2

Apple_PwrMac_G4-500 558.1

Asus_Athlon AMD_800 mhz 600.3

Cray_J932 Supercomputer 1413.6

DELL Pentium 4-1400 1437.2

CYBERIMAGE P4 1.5 Ghz 1544.4

as you can see the P4 ends up being the fastest CPU in the world and its rambus blows away SUN and SGI workstations
as well as a Cray supercomputer

there is alot of BS in the media about the P4 and rambus not being fast,
and as I have said before many times here, you have to use a program that
1- takes advantage of the P4 code and compilers,
with an OS like ME and DD8b that are compiled for P4

2- most importantly, use a program that can hammer the system hard enough to show where the P3 and Athlon start to
get saturated and decline or level off in performance where the P4 keeps on going..

MS OFFICE and simple programs do NOT do this,
most games do NOT, although QUAKE is an exception
as they do not put a clock or performance timer in theor program like others do, so that explains why
Quake runs at 235 FPS compared to Athlon's 150 !!

when you run very complex programs that have the data to saturate most P3's and Athlons, Like Mpeg 4 encoding,
Photoshop calcs, engineering apps, some games like Quake,
Multimedia apps, video , 3D calcs, like MAX, and Lightwave,
DVD,etc , you can clearly see the results of the 3.2 GPS
of the RAMBUS and the 850 chipset working together

the programmer of memtach wrote me recently when I sent him our P4 results and said this ...

"I'd expect the P4 to do well on the streaming tests (fill double / int, sum double / int) etc..
I'd love to work on a P4 review, but Intel has not been very helpful in the past; they have yet to loan me a system. I'm still working on getting a P4
in for testing - it should do quite nicely on multimedia that can utilize SSE2."

believe me I have run several dozen tests in apps and benchmarks, using the proper atmosphere of ME, WIN XE,
and DD8b and the P4 with dual channel rambus is
like 2-3 times faster than a P3 in some instances
as well as an athlon even with DDR

I know its not what people want to hear if they do not have a P4, but we are relatively unbiased as we have to engineer and sell the highest performing workstations and servers to our clients, and if Athlon and DDR was better , they would demand it and we would supply it.. but they do not,
because the P4 and rambus really is faster,
and I do not have a machine ego in owning a P3, or Athlon
to say P4 is not worth it or is slower to rationalize it,
in fact after testing the P4 and seeing the results
I upgraded my dual 933 P3 to a P4 1.5 immediately,
at some expense but it was worth it..

Specmark, as Stream is respected round the workd ans is cross platform, and in both the P4 is much faster than athlon or anything else for that matter..


though the readers in this thread may find this info interesting and keep up the good posts :)

best
CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
Ok then here are the adjusted marks as you deem fair after the 2.3% has been deducted from AMD's scores.

Advanced Micro Devic Gigabyte GA-7DX , 1.33GHz 471 527
Intel Corporation Intel D850GB 1.3 GHz, Pe 473 483

advantage AMD

Intel Corporation Intel D850GB 1.3 GHz, Pe 503 511
Advanced Micro Devic Gigabyte GA-7DX ,1.33GHz 404 435

Now, you are unsure of the testing platform used for the AMD chip as far as the type of RAM and its cas latency. It could be as high as cas3 pc1600, or it could be possibly cas 2.5 pc2100. At any rate I highly doubt it is the fastest cas2 pc2100 that has just become available. Furthermore, the only test platform that is available on the link you provided us with is the gigabyte motherboard. I present you with the following link that shows the gigabyte board is not the best platform for use with the AMD chip:

http://www.tomshardware.com/mainboard/01q1/010314/amd760-09.html

As we can see here the Asus board does appear to be faster. Now couple that with the CAS2 DDR RAM and instead of The athlon being penalized for a higher clock we may now need to actually give it percentage points, do to the fact it was not used in the best possible platform as you agreed would be fair for purposes of comparison. But I digress, this is merly interpolation, and as we have seen earlier this does not always hold true.
So tell me, where can I download this benchmark to try it out to see if what I beleive may be the case is in fact true?



A little bit of knowledge is a dangerous thing!
 
YEs and for those interested ,
the editor of Microprocessor Review , I think his name is Kevin, used to work in engineering at AMD until recently,
and he says in a video interwiew at Cnets site,
that since INTEL is using 300 mm wafers NOW, and P4's
smaller line width size, that INTEL can go well past 2 ghz
with it, and make them 40% cheaper than AMD, which is using still a 200mm wafer, so INTEL gets more chipset per wafer, and thus it cost them less to make the P4

the P4 prices just dropped 40% today because the release of the 1.7 GHZ p4 next week..
the 2 ghz is expected in 3 months..
AMD delayed the future athlon at 1.5 for 3 months,
because it is having problems with making them in bulk because increased frequencies require more pipeline stages
to show gains, and smaller die sizes and voltages
to keep it cool, which is why the Athlons melt and burn up
without very large heatsinks, and why they require chipset coolers, and overload powersupplies..
the P4 has a special 12 volt power supply with 3 leads,
and the P4 has special circutry to slow down and or shut of at a certain temp so it will not burn up..
My 1.5 overclocked to 1.6 is running at 32C !!! with a
$6 heatsink\fan

AMD won the battle a year ago, when it took time for INTEL to move to P4 but they will lose the war as INTEL has better manufactureing to move past 2 ghz

CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
That was some interesting read from an 😉 unbiased sorce. If I understand you correctly Intel uses a 300 mm wafer while AMD uses a 200mm wafer. The last time I checked 200 mm was 33% smaller than 300 mm. So how can intel yield more chips giving them a lower production cost? Also, reliable sources have told us that intel is far behind AMD in there conversion to a .13 micron process in terms of manufactoring. If AMD is having such a rough time launching the 1.5 gig athlon as you suggest why are so many people having great success overclocking there 1.33 gig chips to 1.6 gig? Not to mention the palomino core athlon is supposed to be cable of running cool enough not to require a fan (that I admit even I have to see to believe). Bring to the fold AMD's ( actually IBM's) SOI technology and my friend the war is far from over. As for your glaring love of RAMBUS and RDRAM lets leave that for another thread. The plusses and minus's have already been well debated. I myself am curious to see the p4 coupled with DDR RAM tested in real world apps to see just how significant RDRAM really is.


A little bit of knowledge is a dangerous thing!
 
you misunderstand what a wafer is..

a 300mm wafer or 12 in allows more CPU dies to be lithographed on it, as you said 30-40% more

AMD's 200mm or 8 in wide wafer is smaller and holds less cpu dies Per sq inch
AMD's palimino has been delayed till OCT,
and it is not very different than current athlon
its sole purpose is to reduce voltage and heat so as to allow better yields and high clock speeds otherwise unattainable because of the horrific heat Athlon produces.

Pentium 4 is a ground up new design P7, that allos it to go to 3 ghz.. min.

whoever told you INTEL is behind AMD is .13 process is an idiot..
again, the editor of Microprocessor REport who worked as an engineer at AMD 8 months ago stated INTEL is ready to move to .13 in NOW has whereas AMD will be 12-18 months behind..
AND...
you missed this weeks EE TIMES article
http://www.eetimes.com/story/industry/semiconductor_news/OEG20010406S0062

"Intel claims first blood in 300-mm wafer shift"
By David Lammers
EE Times
(04/06/01, 3:28 p.m. EST)
AUSTIN, Texas — Intel Corp. has fabricated its initial ICs with 130-nanometer (0.13-micron) design rules on 300-mm wafers at its D1C development fab, the company said this week. Intel claims to be the first in the industry to combine the advanced process on the larger wafers."

AND...
"Factory to factory, the larger wafers provide about 1.6 to two times higher output for the same die, and a 30 percent cost advantage," Garrett said, noting that Intel can build fewer factories and get more output from its fab construction resources. "

"For its part, AMD recently announced it will begin 300-mm wafer production in 2004, with partners being sought to share the cost. "

they are the premiere CPU data source for industry news..

and the reason AMD's are overclocked is that the are not clock locked to the degree that INTEL chips are,
INTEL uses a more sophisticated communication regarding identification between CPU, BIOS, and MB that prevents this
without FSB overclocking and even then it is difficult..

this is to protect people from getting riupped off from remarked grey market CPU's that are clocked higher and sold
for more than they really are as in the case of AMD's..

a few facts for you to ponder..

I have several DDR boards at work with P4 from SUpermicro
and they are slower by far in memory than RAMBUS,
as they do not have dual channel or 3.2 GPS bandwidth,
or the 850 chipset to manage the P4..

the VIA chipsets suck, and are not hyperpipelined, dual channel design, and do not have 3.2 GPS bandwidth
as the 850 does, so they are a bottleneck to P4's
3.2 GPS memory bus runing at 400 mhz down to DDRs 250 !

I have tested the same MB makers DDR and RAMBUS chipsets and the 850 and rambus is 2 times faster at least,
in SANDRA, MEMTACH, and STREAM !!

sorry but that is the way it is..
as far as IBM, what technology, they are still trying to make 600-700 mhz PPC chipset because the design is so poor..
what you read they did at their expierimental lab does not translate to production on the line..


enough said ..
CAMERON


CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
It's available at www.spec.org, but it's not free. In fact it's rather expensive as it a professional piece of software with lots of scientific effort put into it.

-Raystonn

-- The center of your digital world --
 
You make some good points, and I favor the Pentium 4 over the Athlon, but I wish you'd try to sound a bit less like a marketing guy. 😉 For a moment I thought I was in a commercial...

-Raystonn

-- The center of your digital world --
 
I am not a marketing guy , I am a computer engineer..
so have no idea what you mean..

I will say this though, marketing often involves staing facts and points that are not widely known to illustrate a point,,, and that is certainly need in this forum where so many AMD Zealots are spewing emotion and rhetoric
instead of facts that exist int eh real world

hope that helps clarify
best
CAMERON

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

We offer both OEM and retail depending if the customer
needs Rdram included.. as they are otherwise identical CPU parts...

the fans we use are ball bearing and from INTEL
regardless of OEM or retail..

the retail is a good deal IF you need 128 megs RDRAM

IF you DO NOT or NEED MORE then the OEM makes sense
the prices listed are retail though..

hope that helps clarify
and thanks for asking..
best
CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
Out of curiosity, have you seen the site www.pricewatch.com? You might want to take a look at some of the prices listed and make sure you're competitively priced. Most of what's sold by the vendors there are OEM.

-Raystonn

-- The center of your digital world --
 
http://www.theregister.co.uk/content/3/18193.html

http://www.theregister.co.uk/content/3/17866.html

Jack Robertson writes that Intel Corp. has not cut back on its incredible $7.5 billion capital spending plan for 2001, despite current dismal condition for chip makers, because it is costing so much to make Pentium 4s. Currently, the P4 die is huge, over 200mm square, and therefore extremely expensive to produce. The solution to that problem is the transition to the new P860 0.13-micron process, but Intel must completely retrofit existing facilities, installing entirely new exposure tools.

Advanced Micro Devices' lithography system, on the other hand, will require relatively minor modification to move to the 0.13-micron process. Additionally, AMD has already integrated copper interconnect production at its Dresden fab, while Intel is only beginning to enable copper deposition.


A little bit of knowledge is a dangerous thing!
 
I have to explain something...
the price cut goes into effect tommorrow..
as an INTEL OEM and DEALER we are not allowed to advertise the price cut until INTEL announces it..

others on prices watch, mainly asian grey market dealers
ignore this or don't care and post them ahead of time..
we can match most legimate retail prices IF
it is from a authorized distributer and dealer..
if not it is greay market or a foriegn import and INTEL may not honor the warranty

we change prices tommorrow or today I guess
so it ok..
check back tommorrow forst thing they should be more accurate..

thanks for asking

CAMERON


CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
Intel is slated to shrink the P4 to 0.13-micron process later this year, allowing it to increase margins and in effect have greater flexibility in pricing. AMD’s response to this will be the Thoroughbred, a 0.13-micron shrink of the Palomino that is slated for the first quarter of next year. The Appaloosa, a 0.13-micron shrink of the Morgan is expected a quarter later.




A little bit of knowledge is a dangerous thing!
 
I have to explain something...
the price cut goes into effect tommorrow..
as an INTEL OEM and DEALER we are not allowed to advertise the price cut until INTEL announces it..
our retail part will sell for $595 tommorrow

ALSO ALL THOSE PRICES ARE OEM NOT RETAIL, and INCLUDE NO 128 MEGS RDRAM


others on prices watch, mainly asian grey market dealers
ignore this or don't care and post them ahead of time..
we can match most legimate retail prices IF
it is from a authorized distributer and dealer..
if not it is greay market or a foriegn import and INTEL may not honor the warranty

we change prices tommorrow or today I guess
so it ok..
check back tommorrow forst thing they should be more accurate..

thanks for asking

CAMERON


CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
Your links don't contain any of the text you're citing.

"Currently, the P4 die is huge, over 200mm square, and therefore extremely expensive to produce."

Actually, larger in the case of wafers is better and cheaper. This is an entirely different subject from the 0.13 micron process. The 200/300mm wafer issue and the 0.18/0.13 micron issue are completely independant from each other.

-Raystonn

-- The center of your digital world --
 
In one quote"

"I know its not what people want to hear if they do not have a P4, but we are relatively unbiased as we have to engineer and sell the highest performing workstations and servers to our clients, and if Athlon and DDR was better , they would demand it and we would supply it.. but they do not"

Then the next:

"as an INTEL OEM and DEALER "

Can both really be possible?



A little bit of knowledge is a dangerous thing!
 
THE REGISTER IS A TOTAL BS GOSSIP WEBPAGE EQUILIVENT TO the INQUIRER
the are the same ones who posted rambus lost the case and a decision was in..

I cited Cnet, and Microprocessor report both industry respected pubs, and from a guy Kevin Krewell who worked at AMD as an engineer last year..

AMD will not move to .13 for over a year or 18 months becasue it requires retrofit and new 300 mm equipment
from 3rd party vendors ..
INTEL will have 3 plants making them 3rd quarter...

you info is bogus
CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
YES they can,
as we choose freely to be becasue of INTEL superior suppport
3rd party relationships in the industry
and performance and quality..
if AMD has this we would be an AMD dealer...

we only offer the #1 part manufacturers in terms of quality, regardless of price, no exceptions..
this way no clients are ever dissatisified
sort of like what Mercedes and the Germans do..

Cameron

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
"Can both really be possible?"

I would hope so. In fact, why should you trust those who don't tell you what they do for a living _over_ those who give full disclosure. At least the latter are being honest.

Noone knows what you or I do for a living.

-Raystonn

-- The center of your digital world --
 
Then prove it! Cnet err isn't that more likely Intelnet?
Easy to point as one source that is favorable to what your claims true and respectable then totally discredit one that you don't like as bogus. Does not intel have an investment in cnet? And to that you call unbiased reporting?

A little bit of knowledge is a dangerous thing!<P ID="edit"><FONT SIZE=-1><EM>Edited by Ncogneto on 04/17/01 05:29 AM.</EM></FONT></P>
 
For anyone to see what I do for a living they simply need to look at my profile, before you make a claim as to someone alledged bios or lack therof perhaps you should look into there past post on this very message board and make that decision for yourself.

A little bit of knowledge is a dangerous thing!
 
you should study you sources better..

CNET is the most respected internet news source for IT
and the 3 major news tv networks and CNN often use them as a source..

Microprocessor report, EE TIMES, CNET
have all reported teh same story,
INTEL already is making .13 P4's on 300 mm wafers...
and guess what
AMD is not, and will not according to the companies own sources for 12-18 months on 300 mm dies.

and on 200 mm not until next year
that is a fact, they have delayed Palimino because of this

how many CPU trade pubs do you get or IT pubs in general..
I read some 25 a month

CAMERON

CYBERIMAGE
<A HREF="http://www.4CyberImage.com " target="_new">http://www.4CyberImage.com </A>
Ultra High Performance Computers-
 
You should heed you own signature..
you have a little knowledge and it appears to be dangerous to you and others on this forum....
:)
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