AMD X2 A Failure, Multitasking, Others

Page 3 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
[[[...HT has always been a strong feature and now that AMD is finally learning what multithreading is...]]]

For Fugger:

You're another dumbass.

When the P4 3.0GHZ first came up with HyperThredshit cappabilities it sucked like hell. Even A Barton processor would run all over it in scientific aplications, workstations apps and file compressing (where HT was supposed to excell).

For Porky:

All you're trying to do is to find flaws in the AMD64 architecture where there's none at all. AMD64 is far more advanced architecture than Intel's Netburst bullshit. The reason Intel came up with HT it's because they wanted to "patch" a flawed architecture that stays idle most of the time wasting processor cycles. That's the reason AMD never needed HT (and will never haver to) because the processor excutes more IPC than any Intel processor (even the Pentium M).

You want an advice: Don't smoke it green. :)


My Beloved Rig:

ATHLON 64 FX 55
2X1024 CORSAIR XMX XPERT MODULES
MSI K8N DIAMOND (SLI)
2 MSI 6800 ULTRA (SLI MODE)
OCZ POWERSTREAM 600W PSU<P ID="edit"><FONT SIZE=-1><EM>Edited by bullshitter on 06/06/05 12:13 PM.</EM></FONT></P>
 
For Fugger:

You're another dumbass
Click to expand...
Is he really? Hmmm I wonder how he can do <A HREF="http://www.futuremark.com/community/halloffame/" target="_new">this</A>.

Scroll down to the PcMark04 top 20.

<font color=green>NED FLANDERS FOR PRESIDENT</font color=green> Its justa nother gansta PARTY!
Intel P4 Extreme(3.73)@<font color=green>5.6Ghz</font color=green>
Asus P5AD2-E-Premium
Crucial Ballistix DDR2 667@DDR2 855
 
For some percentages as of, GMT-10 20050607:05.55am

Thread 1 : AMD leads by 19.12% (lite task)
Thread 2 : AMD leads by 49.01% (lite task)
Thread 2 : AMD leads by 10.52% (moderate-heavy task)(fps, not interrupted session loop count!)(AMD 38fps, Intel 34, usually neck and neck.)
Thread 3 : Intel leads by 94.44% (heavy task)

See the clear picture now? If AMD was doing even half of the Divx thread the Intel would be mince-meating it in the others scores. Sorry, but there is something very wrong with the AMD X2.

Any issues with the Intel that could happen will most likely be the case the test is being contucted in, as it is not to Intel specs regarding the CPU manual. You're supposed to use an air duct style case that draws air from CPU. This factor may cause a higher than expected temperature.

Note :

A (Lite Task) is an application that isn't sophisticated in computation, but will try to obtain as much CPU time that is free.

A (Moderate-Heavy Task). The game thread is fairly moderate in CPU use as most of the game work is done in the GFX cards. The demand is on memory being supplied and basic gaming logic.

A (Heavy Task). The divx encoding requires alot of computation, so is classed as a heavy load.

.
 
See Here where Porkster Lost it. He only looks at one item. He dont look at the other stuff there testing. Where Amd Win.

So Porkster quit being a Moron.
 
Seems to me that Porkster still hasn't seen the threads at the very very beginning that says that thread priority is controlled by the operating systems. It is a known fact that the cpu timeshare of windows is very inefficient. If you just search in google, there are plenty of research documents that explain better ways of timeshare.

IMHO, Intel's hyperthreading is just a hack to bypass this weakness. After all, if Microsoft fixed this problem, I would bet that hyperthreading would become useless.

Also, the Intel machine has rebooted 3 times. I guess that thread on if AMD chips are business reliable is proven.


Spelling correction.... "light" not lite.
 
When the P4 3.0GHZ first came up with HyperThredshit cappabilities it sucked like hell. Even A Barton processor would run all over it in scientific aplications, workstations apps and file compressing (where HT was supposed to excell).
Click to expand...
You must mean a 3.06 I am assuming, since last I checked the Barton’s had a hard time keeping up with the Northwood P4's.

AMD64 is far more advanced architecture than Intel's Netburst bullshit.
Click to expand...
AMD64 is x86-64 which is a IA originally 8bit and carried along to 64bit, while Net burst is a name Intel tagged onto the P4 to describe the series of technologies onboard.

The reason Intel came up with HT it's because they wanted to "patch" a flawed architecture that stays idle most of the time wasting processor cycles.
Click to expand...
Actually Intel wanted to push multithreading and what’s the easiest way to do it other than to use 2 CPU's on die? Well the answer is clear thread-level-parallelism or Hyper Threading.

As well as an AMD user you would be completely unaware of the additional responsiveness in multi tasking that Hyper Threading can deliver. So your point of view is flawed, now for the arguments sake of course you will say "oh I have used a HyperThredshit P4 and they don't handle multitasking as well as a Athlon 64" but alas P4's are technically better in that aspect(In single CPU applications).

But you are right on the fact the P4 wastes a lot of cycles but that’s more to do with the fact it has a very long pipeline nothing to do with the fact its a POS that cant think since its prefech engine is still significantly better than that of the Athlon 64 (better by 3-5%).

That's the reason AMD never needed HT (and will never haver to) because the processor excutes more IPC than any Intel processor (even the Pentium M).
Click to expand...
I would have to beg to differ since each chip reaches ideal IPC in specific software, the fact of the matter is the Athlon 64 is pretty darn good in just about every type of software, but both the P4 and the PM have their niche's PM having more than the P4.

IMHO, Intel's hyperthreading is just a hack to bypass this weakness. After all, if Microsoft fixed this problem, I would bet that hyperthreading would become useless.
Click to expand...
I am glad that’s your opinion and you are completely oblivious to the nature of an OS and the code that goes into it.

-Jeremy Dach
 
quote:
---------------------
All these years on-line and you guys haven't figured out yet that when you feed a troll it grows? -DAPUNISHER
---------------------
QFT

LMFAO!!!
 
Since I ask you to list your computer knowledge background, and you didnt, Then I assume that you have none. So in that case, I have a link for you, to explain what is Multitasking:

<A HREF="http://www.newton.dep.anl.gov/askasci/comp99/CS027.htm" target="_new">Multitasking</A>

Where you can read:"That is another great question.
Let me define the two forms of multitasking. The first form, which is most common is, preemptive multitasking. This is achieved by the OS of the computer giving a program a given amount of time to get a task done."

So, there is no mention of the CPU controling the priority here..

And here:"One example of a preemptive OS is Windows NT. NT has the ability to multitask/ multithread. And the ability to assign a higher, lower, or real-time priority to the process to get it done with the most efficiency."

As you can see, it mention that Windows NT set the priority, as it is with 2000 and XP

This is a simple explanation. If you want a more complex, or if your brain is able to multitask, I mean reading and understand what you are reading at the same time, I suggest <A HREF="http://www.embedded.com/story/OEG20030306S0037" target="_new">Introduction to Preemptive Multitasking</A> to have some idea about how multitasking is used in apps..

You can also read <A HREF="http://burks.brighton.ac.uk/burks/foldoc/42/92.htm" target="_new">pre-emptive multitasking</A> to have information about the scheduler. On that link, you can read about the <A HREF="http://burks.brighton.ac.uk/burks/foldoc/31/5.htm" target="_new">Amiga</A> computer, the one that was multitasking in 1985 and that I bought in 1990. I've been multitasking with the Amiga until I switch to Windows machine in 2001. AmigaOS did not BSOD.. it was <A HREF="http://www.elook.org/computing/guru-meditation.htm" target="_new">meditating</A>. That was good souvenirs to remember..

But this short definition from <A HREF="http://whatis.techtarget.com/definition/0,,sid9_gci212823,00.html" target="_new">Whatis.com</A> say it all..

I hope you enjoy this short lesson to multitasking. If you can understand all that, you will now understand that this is not the CPU that define priority and the behaviour you see is perfectly normal for a dual core processor that execute 4 tasks with one in low priority and a dual core processor with HT that process 4 tasks, even with one in low priority in 4 threads..




<font color=red>Sig space for rent. make your offer.</font color=red>
 
I know, but I feel due for some trolling... But the fact is that I'm on a therapy.. Anger management.. how far can I go with him before I start becoming mad...

I could do some spamming too. I'll see..

<font color=red>Sig space for rent. make your offer.</font color=red>
 
In reply to:
--------------------------------------------------------------------------------

For Fugger:

You're another dumbass



--------------------------------------------------------------------------------

===========================================================
Is he really? Hmmm I wonder how he can do this.

Scroll down to the PcMark04 top 20.
===========================================================

So. who cares????

I really apologize for calling that dude a dumbass, but it just pist me off see people talking about something they don't know and keep fighting about it.


===========================================================
Actually Intel wanted to push multithreading and what’s the easiest way to do it other than to use 2 CPU's on die? Well the answer is clear thread-level-parallelism or Hyper Threading.
===========================================================

Well, I just think HT is a bad way to achieve multithreading.
I don't believe in HT.

I believe in SMT which is a way better approach than having 2 pseudo-processors that don't do a [-peep-] (only 5%-10% performance gain, if you can consider that a gain.) :)

I believe in multicore technologies such as AMD's Athlon X2 processor, the Cell (from IBM and Sony) and the Power 5 (from IBM).

===========================================================
You must mean a 3.06 I am assuming, since last I checked the Barton’s had a hard time keeping up with the Northwood P4's.
===========================================================

I was talking about the first P4 that had HT (I beleive it was the 3.06GHz).



And for Porky:

AMD isn't losing on nothing. The problem with you is that you're so obsessed with Intel that you even think that a PII 233MHz is better than an FX55. :)

The X2 doesn't has the fault of running in a dumb OS like WinXP, since Tom could set the thread priority to normal for the DivX test.
I believe such benchmarks should've been run on Linux. Linux is far superior to any Microsoft OS when it comes to thread handling. A benchmark under Linux Would still give the X2 the lead.

Anyhow, talking to you is worthless so all I can say is good luck with your crusade agains't the X2. :)






My Beloved Rig:

ATHLON 64 FX 55
2X1024 CORSAIR XMX XPERT MODULES
MSI K8N DIAMOND (SLI)
2 MSI 6800 ULTRA (SLI MODE)
OCZ POWERSTREAM 600W PSU
 
In reply to:
--------------------------------------------------------------------------------

For Fugger:

You're another dumbass



--------------------------------------------------------------------------------

===========================================================
Is he really? Hmmm I wonder how he can do this.

Scroll down to the PcMark04 top 20.
===========================================================

So. who cares????

I really apologize for calling that dude a dumbass, but it just pist me off see people talking about something they don't know and keep fighting about it.


===========================================================
Actually Intel wanted to push multithreading and what’s the easiest way to do it other than to use 2 CPU's on die? Well the answer is clear thread-level-parallelism or Hyper Threading.
===========================================================

Well, I just think HT is a bad way to achieve multithreading.
I don't believe in HT.

I believe in SMT which is a way better approach than having 2 pseudo-processors that don't do a sh*t (only 5%-10% performance gain, if you can consider that a gain.) :)

I believe in multicore technologies such as AMD's Athlon X2 processor, the Cell (from IBM and Sony) and the Power 5 (from IBM).

===========================================================
You must mean a 3.06 I am assuming, since last I checked the Barton’s had a hard time keeping up with the Northwood P4's.
===========================================================

I was talking about the first P4 that had HT (I beleive it was the 3.06GHz).



And for Porky:

AMD isn't losing on nothing. The problem with you is that you're so obsessed with Intel that you even think that a PII 233MHz is better than an FX55. :)

The X2 doesn't has the fault of running in a dumb OS like WinXP, since Tom could set the thread priority to normal for the DivX test.
I believe such benchmarks should've been run on Linux. Linux is far superior to any Microsoft OS when it comes to thread handling. A benchmark under Linux Would still give the X2 the lead.

Anyhow, talking to you is worthless so all I can say is good luck with your crusade agains't the X2. :)






My Beloved Rig:

ATHLON 64 FX 55
2X1024 CORSAIR XMX XPERT MODULES
MSI K8N DIAMOND (SLI)
2 MSI 6800 ULTRA (SLI MODE)
OCZ POWERSTREAM 600W PSU<P ID="edit"><FONT SIZE=-1><EM>Edited by bullshitter on 06/06/05 09:59 PM.</EM></FONT></P>
 
Porkster you cannot say that one task is lite, moderate or heavy unless you have calculated just how many flops or what ever is needed on each to finish.

Adding to this is the fact that saying recall due to the fact that the x2 computes the 4 th task slower is like saying that I can get the money back on my p3 550 because the guys selling it to me said it would run the newest games...

It doesnt anymore... but its just as expectable as the results sofar in this "stress test".

A thought -> What would have happened to both the results and your poor argumentation if thg had been running FIVE (5) intensive tasks or even more? At some point the intel is going to give up too...

I say we test with atleast 7 applications because thats what I can multitask in a kitchen. 4 pans, the oven, coffemachine and the dishwasher... Mebbe even the fridge and a few cabinets but then i'd leave it to my gf. Snicker
 
Just a qick question: are you either stupid or blind? Or both? Because you seem to completely misunderstand the fact that the os kernel handles the cpu load. Really, that's why you can run to kernel os's at one in a dual-core x86 environment (maybe you can with more expensive processors with virtualization).

So really, I'd rather look at the reboot counter, and the amount of power required.
 
I think it would be great fun and make me very happy if some fluke occured like the encoding program is messing up on the AMD system, or it's not refresing the stats properly, or something that would mean all this BS argueing is for nothing.. I mean look at the numbers people... 110 minutes vs 1640? So in all the time that computers been on it only managed to compress like half a movie? SOMETHING is going on here but all anyone can do right now speculate, and I dont see how anyone could begin to do that.
 
BePe86,

I make it 7 (Intel) vs. 0 (AMD).

However, I think they discount reboots if it's for a hardware change (the Intel is on it's third motherboard; one Gigabyte and two ASUS).

When they get the graphs sorted out, and not several hours / days behind (24 and weekly graphs for Intel and weekly graph for AMD) we should see how many reboots there have been.

Pete
 
More Shocking News has come to light. It appears the AMD system is using more wattage of the two systems being tested.

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

You examine the test page above you will notice that power consumption is marked at 345watts all whilst the Intel system is off and being rebuilt. Before the rebuild the total wattage of both systems was 600watts +/- about 80watts. Doing some simple math one can see the AMD system is the higher abuser of making a high power bill.

This makes the Intel are far better system for large office structures.

.
 
OMG the Intel system is back up and it has added 211 watts to the experiment. Shocking that the AMD is using 345 constantly, when the Intel is using 211 on a static OS.

.
 
I have to disagree with your statements. the Intel machine is currently using about 215 watts while idle with only one VGA card in it. The AMD machine is using 345 watts at full tilt with two VGA cards in it. Further, when both machines were running flat out the total power usage was actually around 850watts (if I remember rightly, when the Intel machine is back up and running properly we'll see) so, simple maths says the AMD machine uses 345 watts (we know this as a fact) and the Intel machine uses over 500watts.

Does anyone else remember the total wattage used to back either cliam up?


---
<font color=green>AMD</font color=green> Athlon64, Abit AV8.
<font color=blue>Intel</font color=blue> Dual PIII, Asus P2B-DS.
<font color=blue>VIA</font color=blue> Epia M9000.
<font color=red>I</font color=red><font color=green>B</font color=green><font color=blue>M</font color=blue> Thinkpad 570E.
 
I think the intel was idle thats why it was only using 211watts. When it is on full load it used 500+ watts. Yesterday it was running at 800+ watts I remember. It was probably the graphics card.

Is the live charts accurate? how can it display the "Current Rate" while its offline?
 
The AMD was using more than 345watts, just I took the score on a low. I'm trying my best to over look the flaws on the new AMD X2.

.
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom