AMD files anti-trust suit vs. Intel

[Jammrock]

well benchmarks don't lie unless he is fudging results...

While AMD is drastically ahead of Intel in the gaming arena, it's quite the opposite in areas like video/graphics production, which is a huge market right now.

So if you want a powerhouse workstation, the recommendation is alway Intel, it's got faster memory bandwidth and the architecture allows for faster rendering capabilites.

You should see my Dual XEONs render in after effects!!!
Actually the only AMD processor I liked from Sharky Extreme's review was the Athlon 64 X2 4800+ it rocked in almost every benchmark but probably worth over a grand.

http://www.sharkyextreme.com/hardware/cpu/article.php/3261_3515996__1

Cheers,
Elie

The problem is that Tom's IS comparing an AMD Athlon 64 X2 4800+ to a Pentium-D 840 EE. The DivX numbers are a joke. Yes, Intel has always done DivX encoding better than AMD. That's a known fact ... it's just the way DivX is. However, most benchmarks put the X2 4800+ at about 25%-35% slower than the 840 EE. Tom's has the 4800+ at 93% slower than the 840 EE. 93%!!!!! That's a big enough difference that you have no choice but to take Tom's benchmark results with a grain of salt.

At least in my book.

Jammrock

[Elie]

If you look at the second link, it was Sharky Extreme's review also that shows the weekness of AMD in video aps.

Like I said otherwise the AMD rules in the gaming department.

Regards,
Elie

[Jammrock]

You mean these video encode test, which show the X2 4800+ beating or neck and neck with the 840 EE:

http://www.sharkyextreme.com/hardware/cpu/article.php/3261_3515996__6

Remember, low is better when video encoding. And the X2 4800+ only lost out to the 840 EE by a whopping 2 seconds, or by 2.2%, in DivX encoding. The problem with Tom's review is sumed up in this quote:

Our observations suggest that the AMD system can't distribute the CPU's computing power as well among the four single applications as the Intel system does.

This "conclusion" is laughable at best. The CPUs have ZERO say in "distributing the CPU's computing power." All the CPU does is take the information provide by the operating system and crunch numbers. It can play with information given to it, i.e. out-of-order execution, pipelining, and all sorts of other tricks, but in the end it's just a giant calculator. In a multiprocessor configuration it is the operating system which decides what application is processed by what thread on which processor/core, NOT THE CPU!!!! An application can be designed to operate on multiple processors, but all that application will do is split up the work and tell the operating system to put Task 1a and 1b on seperate CPUs. It is up to the OS to decide which CPU does what task in an SMP/load-balance intelligent OS, which Windows NT based OSs are.

But without seeing a snapshot of the Task Manager in Windows, showing a full list of processes and which processor the process is being processed on (talk about a tongue twister), it is impossible to say if the error, and it is obviously an error as other indepented tests are no where near a 93% difference in DivX encoding speeds, is a fault of Tom's or of the CPU. Though my money is on Tom's.

Jammrock

[Sasq]

ok guys, before we get too excited Toms has for a long time had questions hanging about the review quality.

Each proc has strengths and weaknesses, get the proc that suits your needs.

now look at the bunnies

[Jammrock]

now look at the bunnies

/Jammrock's head explodes.

You know we hardware geeks love taking our shots at Tom's whenever we can I'm honestly looking at getting a Pentium-D because I can get one for $250, while the cheapest X2 is $585. It wouldn't be nearly as fast as other CPUs I could get, but it may be worth it for all the multi-tasking my computer does these days.

*shrugs*

But that decision is still a few more months away.

Jammrock

[dshumake]

According to Tomshardware, the AMD chip was better at every task when run alone (see link below). The Intel chip was better, significantly, when four processes were being run on a dual core with hyperthreading --> acting like for cpus, without hyperthreading Intel was again below AMD. The question toms hardware didn't ask is what happens on each machine if more programs are running, 5-x or if 2 or 3 programs are running. Obviously the number of combinations is large and Tomshardware made a decision.

dshumake
_________________

http://www.tomshardware.com/cpu/20050603/stresstest-03.html

Now that we have all the data from the individual tests in the last few days, what can we say about performance? AMD is in the lead for all four applications.
Data compression with WinRAR: AMD system has 29.5% better performance;
MP3 encoding of the CD: AMD system has 4.7% better performance;
DivX encoding of the DVD: AMD system has 28.2% better performance;
3D game Farcry: AMD system has 21.4% better performance.

[ALBPM]

Bottom line is AMD has failed to deliver consistantly. This isn't new, it's plagued them all through the 90's and still continues. Their problem has been bad management at the hands of "King" Jerry Sanders.

Jerry Kayoed by the K6

http://www.businessweek.com/1997/42/b3549066.htm

How to get paid well for performing badly.

http://slate.msn.com/id/2688/

This is the same guy who put murals of himself as "King" in the AMD Lobby and ran the company according to his whims.

.....It's everyone elses fault

[Wombat]

Bottom line is AMD has failed to deliver consistantly. This isn't new, it's plagued them all through the 90's and still continues.

I dunno, I've been pretty impressed with all of the K8s so far, which is more than I can say for the P4s. I've never liked the P4s as an architecture, but at least Intel is finally wising up a little bit and bringing the P3 line back. I think that Intel burned up a lot of its reputation for quality with the Rambus/MTH issues, and the lack of killer processors. Don't even get me started on the Montecito thing.

For the past couple of years, Intel's performance has levelled off while their quality has gone down, and AMD's performance, quality, and value have gone up, up, up.

[Jammrock]

According to Tomshardware, the AMD chip was better at every task when run alone (see link below). The Intel chip was better, significantly, when four processes were being run on a dual core with hyperthreading --> acting like for cpus, without hyperthreading Intel was again below AMD. The question toms hardware didn't ask is what happens on each machine if more programs are running, 5-x or if 2 or 3 programs are running. Obviously the number of combinations is large and Tomshardware made a decision.

dshumake
_________________

http://www.tomshardware.com/cpu/20050603/stresstest-03.html

Now that we have all the data from the individual tests in the last few days, what can we say about performance? AMD is in the lead for all four applications.
Data compression with WinRAR: AMD system has 29.5% better performance;
MP3 encoding of the CD: AMD system has 4.7% better performance;
DivX encoding of the DVD: AMD system has 28.2% better performance;
3D game Farcry: AMD system has 21.4% better performance.

Yes, yes, yes ... but the comment I quoted made me wonder about their ability to test, troubleshoot and draw conclusions on an SMP testbed. Oh well. Both of those CPUs are far our of my reach, so in the end it doesn't really make much difference. ...and Tom's still sucks.

Jammrock