I need Kobe Tai to come over and bite my tongue.

Printable View

Show 100 post(s) from this thread on one page

06-09-2005, 07:53 PM
Zebo

I gave you the easy answer.:)
06-09-2005, 08:00 PM
Hans.Gruber

Quote:

Originally Posted by Zebo

I gave you the easy answer.:)

Ok, now you answer to this.

Why Intel stats are increasing without it having a network but AMD's don't ??

http://www.tomshardware.com/stresstest/intel.html
06-09-2005, 08:09 PM
Zebo

Quote:

Originally Posted by Hans.Gruber

Ok, now you answer to this.

Why Intel stats are increasing without it having a network but AMD's don't ??

http://www.tomshardware.com/stresstest/intel.html

That's easy too. Toms HW a liar making this whole test up after initial failures.

Intel used to have a network.
I got all this archived...(need IE)
http://www.csc.calpoly.edu/~smanning/44.mht
06-09-2005, 08:12 PM
Lithan

Lol, he's right. I just verified it and Intel says no net, but it's stats continue going up. If I was Intel I'd be pissed off that my puppets were making it look like I needed them more than I actually did.

Edit: I honestly think this has got to be a joke. There is no way someone could &(*^ up a test so badly, try to cheat in it so badly, and then cover it up so poorly. And if it turns out this is a joke, I'm sure it'll infuriate Intel that a site that gets as much traffic as Tom's makes them look so bad.
06-09-2005, 08:56 PM
pha|anx

I wonder what Mr. Pabst thinks if he's reading this, gurus and legends and such cutting him up good lol

:wierd: :bounces:
06-09-2005, 09:10 PM
Lithan

Probably too busy trying to scrub the stink of whoredom off himself to be reading this.
06-09-2005, 09:55 PM
XtremepH

nvm
06-09-2005, 09:59 PM
Lithan

Yeah was posted on page 5 I think. Sorry.
06-09-2005, 11:09 PM
MaxxxRacer

Guys lets stay calm and not have this turn into the flamewar that it is moving towards.
06-09-2005, 11:39 PM
OC Detective

Yeah lets get back on track to the burning issue - namely THG's shenanigans.
He claims not to be touching priority selection in Windows for AMD but how can he explain the following:
Approx 3:20pm my time (+8 GMT) the readings were as follows:
Lame Encoder Intel = 301 AMD = 256
Winrar Intel = 1634 AMD = 2072
Farcry Intel = 27FPS AMD = 35FPS
Divx Intel = 1380 minutes AMD = 80 minutes

About 10 minutes later the following:

Lame Encoder Intel = 303 AMD jumps up to 311!!!
Winrar Intel = 1643 AMD jumps to 2321!!!
FarCry Intel = 27FPS AMD drops to 15FPS!!! (edit: latest now is Intel at 37 and AMD at 35!)
Divx Intel = 1390 AMD jumps to 130

No explanation for the Farcry drop other than priority adjustment is there?
06-09-2005, 11:43 PM
MaxxxRacer

i think their benchmark numbers are TOTALY fubared. there is no ryhme or reason to them. at some points it shows intel doing better in one area and then a few hours later it shows amd doing better.. they are totaly useless sets of data imo..
06-10-2005, 12:13 AM
OC Detective

I am looking forward to seeing what kind of rational conclusions they are going to come up at the end of this!
If it were me, I would be saying "well basically we f###ed this test up big time - sorry!"
06-10-2005, 12:23 AM
trance565

lol, i think they shoulda used a dbl bar graph for the far cry section, one bar showing the amount of runs, one bar showing the fps, cuz the fps is kina misleading, like when the bar shows intel winning, ppl are like ooh yay, but they forget to read amd has been faster for more time, making it finish 300 more runs than intel

oh ya, lithan, u get my pm :)
06-10-2005, 07:54 AM
saaya

Quote:

Originally Posted by FUGGER

Yes Hans, that is true

The side panel I picuted is from a reference BTX machine I own and it was provided by Intel. The Heatsink fan I pictured is the original BTX design that was delivered the above said BTX case.

Yes Intel shipped a lot of these and yes Intel has changed the spec on BTX.

You guys can argue with an Intel developer that has the reference designs on hand? Try and open your eyes and look at the picture, it is obvious that the fan has no shroud. The shroud I pointed out (granted it is hard to find on the shelf) is the correct hardware for that setup.

FYI 955 is a BTX board, last ATX board was the 875 chipset. BTX boards are exact to ATX boards less the percise placement of CPU in relation to the shroud I posted a pic of.

My Intel reference machine (BTX standard) will fit any 915, 925, 945 and 955 mobo perfectly.

charles, i think you might have received a very early idea of btx, probably just an atx heatsink with a shroud to suck cool air from outside of the case. it must have been very very early because they changed the btx idea and went fopr a different mainboard design and a tunnel copu heatsink that pushes the air through it from one side to another:
http://pcstats.com/articleview.cfm?a...id=1712&page=2

this is from january 2005 and shows and early btx heatsink sample, wich looks identical to the final version of the btx heatsink that is now available in retail. somebody posted a pic of it earlier in this thread:
http://www.watch.impress.co.jp/akiba...btxp4560j.html

and btx mainboards have a totally different layout than atx mainboards, a real btx heatsink will just not fit in an atx mainboard.

left is atx right is a btx mainboard

http://img226.echo.cx/img226/7051/btx12mz.jpg

here you see what the idea of btx is. on the left is an atx case, theres acase fan in front of the hdds that blows in fresh air on the bottom, and a psu fan on the top that sucks tha air out. then you have a fan on the videocard and cpu and most of the times also on the northbridge. but each of those fans is not moving the air from the intake to the outtake of the case but only moving it around creating turbulences and heatink up the case air.
on the right you see a btx case, there are only 2 fans, one that takes the air in at the front and one that takes it out on the back. maybe an extra fan for the videocard, but thats it, there is only one direction the air is moving, from the front to the back.

http://img226.echo.cx/img226/6513/btx29fa.jpg

thats why i dont think the heatsink is a btx heatsink, btx heatsinks dont have a fan on top of them, the idea of btx is to move the air from right to left, a fan on top would only create turbulences and reduce the cooling effect. and the heatsink isnt shaped like a btx heatsink either taking advanctage of the airflow from right to left.
i think your heatsink and shround are intels first steps towards btx but dont have anything to do with the current btx standard.
they probably sent you an atx heatsink with a shroud, wich is how the idea to btx started. thats probably why it is a btx heatsink, but then again its not. :D

Quote:

Originally Posted by FUGGER

Hans, I am a Intel developer and that is from my BTX refence system. The spec has changed since then but many of those heat sinks hit the retail market and were shipped with the processors to the reviewers.

I have expressed the problem to Intel about giving reviewers the incorrect supporting hardware a few times now.

My case was designed for that sink and that is obvious.

Trust me, I have seen every revision and know that heat sink well. open your eyes and you will see it has no shroud, use common sense and figure out that without the shroud the air movement down is reduced greatly.

If you had the case like mine (sold at Frys and Compusa) you would know it make a huge difference in the operation of that heatsink. Night and day difference.

Intel website has been update to show current BTX. I was on BTX so long ago and I tested the sinks you are seeing now on all those review sites. I expressed my concerns about this problem and un educated people running the open fans (exp this stress test crap) Without it, it is going to fail over and over again under load. Why do you think so many people complained of Prescott temps?

All motherboards 875 and newer support BTX standard as well as ATX. AMD guys are clueless and I am speaking jibberish.

so your saying the idea of having one airflow to cool the cpu nb sb memory and videocard are now changed? the idea of btw WAS to have one wind tunnel to cool all hot parts of the pc. and intels btx site still says so:
http://www.intel.com/cd/channel/rese.../eng/95347.htm

Quote:

Superior Thermal Environment
*High power components use the same in-line, high velocity, low temperature airflow
*Above and below motherboard airflow helps improve voltage regulation and socket capability

Superior Acoustic Performance
*Only two fans required - even in high performance configurations
-Thermal module fan and PSU fan create and manage simplified system airflow
*Lower impedance and high velocity, low temperature air allows lower fan speeds

and

http://www.intel.com/cd/channel/rese...hermal_Modules

Quote:

BTX-based Motherboards

The BTX-based motherboard layout differs significantly from ATX/microATX, and requires designs built specifically for BTX. BTX also allows for multiple board sizes utilizing a common core:

picoBTX: maximum width 203.20 mm, up to 1 add-in card slot.
microBTX: maximum width 264.16 mm, up to 4 add-in card slots
BTX: maximum width 325.12 mm, up to 7 add-in card slots

BTX-based Chassis

Because of the changes in motherboard layout, the new form factor will require unique chassis designed specifically for BTX. It will be critical to select a chassis that supports the chosen board size. A BTX-compatible chassis will require built-in features to attach the Support and Retention Module (defined below). Additionally, a Thermal Module Interface may be required. This is a ducting feature connecting the front of the Thermal Module to a front vent. It is needed to ensure that airflow exiting the Thermal Module cannot re-circulate and re-enter the Thermal Module Interface opening.

Support and Retention Module (SRM)

The SRM is a metal plate that is assembled to the chassis beneath the motherboard to provide structural support for the motherboard and retention for the thermal module. The SRM is expected to be shipped with BTX-compatible chassis.

so that explains the btx design as having a duct on the FRONT of the case sucking in air ,blowing it over the cpu and then exhausting it in the rear.
the case also needs to support this.
your case and duct are on the SIDE of the case, not in the front.

so either you got an early btx design before the standard was set or intel changed the btx standard and hasnt updates their site yet?
and i cant believe that intel went from the btx with a windtunnel to a simpler btx with just a duct from the side of the case. this simple duct from the side of the case doesnt have all the benefits intel advertised btx with, bigger and heavvier heatsinks possible and cooling all system components. the duct you use only cools the cpu.

and your mainboard is not btx, so did intel scrap the whole idea of btx mainboard layouts? :confused:

this doesnt make any sence, i think you have a very early btx prototype.

Quote:

Originally Posted by Noldor

I think we all can, and should, agree that the heatsink used in the test is NOT the heatsink Intel meant users to combine with their dualcore processors. This chip apparantly requires BTX to run properly, and I guess Intel should advertise that more and demand that their resellers inform about this on the spec-sheet at the retailers webpages. 99% uf all users buying this chip will not own BTX cases and will run into some serious problems. Too bad. Yonah seems extremly promising though and I can't wait to buy one. But even those who prefer Intel have to agree that if you ARE going to spend 1000$ on a processor, the Athlon X2 IS a better choise. Which processor is best for budget systems will of course depend on money and where and how its gonna be used. But if you MUST spend 1000$ on a CPU, I can't see ONE reason why the Intel one should be recommended.

if it requires a btx case to work then why doesnt intel say anything about it?
while this might be an error of communication, the following certainly isnt:
all dual core reviews were done on mainboards intel sent the hardware sites. they sent not just dual core cpus but testing systems, some sites even received an entire case already assembled to test the dual cores and it also had the same heatsink on it!
dual core cpus dont need a btx mainboard to run, what they need is a better heatsink! theres no need for a new case with shroud and whatnot, all you need is a better heatsink!

Quote:

Originally Posted by FUGGER

The Intel is superior when more than two threads are used. It is superior at encoding and multimedia streaming.

It is a superior product when selling to big business and is accepted into military, government and educational sales. This is where AMD is rejected the most.

People who demand higher quality machines buy Intel

Professionals who need un comprimised stability buy Intel.

AMD is fine for the gamer and casaul home user who can overlook minor stability issues and willing to beat head against the wall over NF3/NF4 reboot issue for hours to eventually give up to only see the Windows splash screen.

The members of this site has killed hundreds of FX processors over the past few months. Some members killing upwards of a dozen. Weather it was too much vcore or too much vdimm they are still deaded.

You would be digging deep to find one dead P4 and near impossible to find two dead ones by the same person. Give it a rest about AMD being superior. Its all in you feeling you have ownage with your AMD cpu. :party3:

no amd vs intel discussion please :D :toast:

i dont understand why intel wants to save 5$ on each cpu wich a better heatsink would cost them at most... i hope they finally start equipping their boxed cpus with better heatsinks.

Quote:

Originally Posted by Noldor

Well, this debate could go on for years and years. Toms test hold no real relevance to this and I'm sure Intel's processor is just as stable as the AMD one when used like it's intended to be used, with the BTX case/fan.

there are intended to be used on atx mainboards and atx cases, and they dont need a btx mainboard and case tow rok, they only need a good heatsink.

Quote:

Originally Posted by Noldor

Many of us have been running AMD for years without problem. FUGGER has ran Intel without problems. So which is best? I game mostly and do some "internetting", the choise is obvious. However, should I go into some serious encoding i'd look into an Intel rig. And to be honest, 100 or 110 FPS in a game, or 5 CD's in an hour instead of 6 CD's. LOL wth are we arguing about anyway? As long as the processor is fun to overclock we should all be happy. We can use intel for PI and AMD for 3D. YAY!, fun for everyone.

exactly! :toast:

please stop the intel vs amd discussion, this thread is about thgs incapability of getting a cpu stable, what brand that cpu is doesnt really matter, id laugh my 4ss of as well if it was an amd cpu they wouldnt manage to get stable and kills 5 mainboards trying to get the system working :lol:

Lithan and everybody else, please stop the amd vs intel discussion!
im going to remove all intel vs amd comments and posts to keep this thread clean. if you want to start an intel vs amd thread please do so in the chat section.

Quote:

Originally Posted by Zebo

Some drunk in toms "lab" prolly fell over the network cable and disconneted it.

tom got frustrated and got drunk and then fell on the server :lol:
06-10-2005, 08:14 AM
Gorgeous

Quote:

Originally Posted by saaya

Lithan and everybody else, please stop the amd vs intel discussion!
im going to remove all intel vs amd comments and posts to keep this thread clean. if you want to start an intel vs amd thread please do so in the chat section.

tom got frustrated and got drunk and then fell on the server :lol:

Super Saaya to the rescue! :p: Glad you cleared up that mess.
06-10-2005, 08:21 AM
saaya

LOL super saaya yin :lol:

*looks in mirror* meh, still no blond hair
gotta try harder.... hrrrnnngggmmmm
bah i give up...

btw, i compared the performence stats at 4pm and 6pm central european time:

340 cds
342 cds

1844 archives
2576 archives

37fps 1152runs
35fps 1450runs

1550 mins
130 mins

2 hours later:

351 cds
351 cds

1906 archives
2655 archives

37fps 1198 runs
35fps 1495 runs

1600 mins
140 mins

this means the intel system performs as following:
it encodes ~5.5cds/hour
zips 31 archives/hour
does 23 runs of far cry/hour
encodes 25mins of a dvd/hour

and the amd system performs as following:
it encodes ~4.5cds/hour
zips 39.5 archives/hour
does 22.5 runs of far cry/hour
encodes 5mins of a dvd/hour

lets see if this will change in future.
does anybody have several screesnshots of the same day at different times to compare how the systems performed before?
06-10-2005, 08:35 AM
Lithan

Nah, a decent copper heatsink would cost $10-15 more to make I figure. A better aluminum heatsink would mean more noise (since cpu manufacturers are hardly free to create high performance sinks, limited by the need for 100% compatibility and little R&D).

Trance, I got it. The only thing I can figure is now that memory controller is onboard, games might stress it more than most cpu loaders do and it might be the reason your games heat up the cpu more than cpu loaders. I'm not exactly sure how much heat output A64 mem controllers have.
06-10-2005, 08:40 AM
Charles Wirth

This thread was flying yesterday :D

I dont make any $ from Intel directly LOE. I stand by the product that serves me well. Others stand behind AMD

Anyway, it will interesting to see where the stress test goes on THG. :lol:
06-10-2005, 08:44 AM
saaya

yeah! i hope they blow up some more mainboards, a psu and maybe some other stuff :lol:
this is really fun to watch :rofl:

i hope epox gigabyte and intel will send them their special "thanks" and support thg with the hardware they "deserve" in future :D
if i was in charge at intel id send them network cards to test to see if they are capable of doing a good review of that before i send them anything complex again. i bet they would kill 2 boards with that network card as well though :rotf:
man i was close to laughing tears when i first read the thg article updates about all the boards they killed :D
06-10-2005, 08:51 AM
trance565

saaya, the link
http://www.watch.impress.co.jp/akiba...btxp4560j.html

of the six pics on the right, u can see the middle left one shows an actual btx design, the fan is on the side, so it is btx, reviewers just must not have been informed that they need to keep the fan on the side

i guess most the review sites pulled a tom, and took the shroud off, and mounted a fan on top, like they are used to doing, and didnt do what intel wanted them to
06-10-2005, 08:53 AM
MaxxxRacer

THG with a network card:
It seems the heatsink on the NIC was not large enough and it overheated frying the memory power regulators on the mainboard. We tried a second motherboard and found the southbridge to be faulty which caused more fialures. We finally resolved the problem by installing a XP-120 with a 220CFM delta fan on the NIC. Currently this setup is working nicely with the fan being powered off of the mainboard.

:rolleyes:
06-10-2005, 09:02 AM
Zebo

Quote:

Originally Posted by MaxxxRacer

THG with a network card:
It seems the heatsink on the NIC was not large enough and it overheated frying the memory power regulators on the mainboard. We tried a second motherboard and found the southbridge to be faulty which caused more fialures. We finally resolved the problem by installing a XP-120 with a 220CFM delta fan on the NIC. Currently this setup is working nicely with the fan being powered off of the mainboard.

:rolleyes:

GAWD what a bunch of liars... How the heck you install an XP120 *AND* a 220CFM fan on a nic !??!?! I gotta see this before believing.

220 CFM means a huge denki or panaflo 120x38mm. XP120 masses about 1 Kg or two pounds. Nic's have no holes in them to "attach" anything. You drill holes and PCB is fried.
06-10-2005, 09:07 AM
Ref

Quote:

Originally Posted by Zebo

GAWD what a bunch of liars... How the heck you install an XP120 *AND* a 220CFM fan on a nic card!??!?! I gotta see this before believing.

220 CFM means a huge denki or panaflo 120x38mm. XP120 masses about 1 Kg or two pounds. Nic cards have no holes in them to "attach" anything. You drill holes and PCB is fried.

lol it was a joke :ROTF:

Show 100 post(s) from this thread on one page