AMD dual core reviews

Toggle sidebar Toggle sidebar
D

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
First off who has a list of all sites that did them...

IMO the two best review sites on the web did them in Anandtech and TechReport.....But who else did them. I haven't seen any others to date:

secondly what can we tell from this limited review by a few sites???

1) Are these sites that did them viewed by AMD in the "most favored" status?

2) Other sites such as Tomshardware, and a few others who I wont name whio in the last few months have been repeatedly caught running AMD systems at less then spec or severly crippling them with other components in reviews against 6xx series and dual core INtel chips...IE via boards and cas 3 ram....

Thirdly what about the "big snub" of Tomshardware?? Are they just slow in getting a review out or is their bias so obvious they chose not to run any test to embarass the "silver bullet". That was the title to their Pentium D reviews...
 
D

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
Thanks guys keep them coming....Then comment on how Toms (even if you dont read him his site is HUGE) is absent from this later round of reviews......
 
D

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
Originally posted by: PetNorth
Another

http://www.extremetech.com/article2/0,1558,1788701,00.asp
Click to expand...



I read the review...My thoughts...

1) Too many synthetic test...

2) he uses POV-ray 3.6 which is not multithreaded unless you run more then on thread by opening more then one app of it...Most are smart enough to do this and not put out confusing data such as that.

**note: most reviews need to note there are many apps that can handle 2 threads but do not handle 4 threads efficiently as well...In the rendering field many of the apps are very good performers with more then 2 cores...TechReport does this....


3) his Divx test shows he used an encoder that is not very multithreaded. There are some that gain far more i(both the PD and X2) then what he has listed...Again be smarter then this...

4) I like his multitasking graphs.....


EDIT: I do have to say MAJOR :thumbsup: for stating the inherent performance degradation by using the 875 opteron as a reflection of a X2 of equal speed in terms of performance...some reviews did not touch on this and that makes it a bit deceiving...better for Intel nonetheless!!!
 
D

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
"It?s clear that dual core Opteron works very successfully as the basis for an incredibly fast PC, but that completely misses the point as this is technology that cries out for multi-threaded work so we decided to let it stretch its legs.

We fired up the POV-Ray (Persistence of Vision Raytracer) rendering benchmark which is a free download from www.povray.org

POV-Ray version 3.6 doesn?t multi-thread so it only uses a single CPU and took 29 minutes five seconds in 32-bit Windows XP

Switching to XP 64-bit with the same version saw the time drop to 22 minutes 55 seconds.

We then installed the version 3.7 Beta which does use multiple CPUs and the time came crashing down to 2 minutes 50 seconds using 32-bit Windows XP. This was quite staggering but when we then moved to Windows XP 64-bit the time came down still further to 2 minutes 14 seconds."


I liked this statement in the trustedreview link....Dont care much for the webformat and no direct comparison benches but it tiouches on the powe rof threads when configured correctly can really deliver some power...and in the field I want to see it CAD rendering....
 
Zebo

Zebo

Elite Member
Jul 29, 2001
39,398
19
81
WOW!! Thanks D for taking the time to point this stuff out. Definity on that.:)
 
R

rgreen83

Senior member
Feb 5, 2003
766
0
0
Duvie: I havent seen where toms was crippling the amd platforms, do you have a (or more?) links to that?

I really think the reviews should be done by benchmarking the systems, then putting them in ratios of performance to power usage. Some have done this to an extent, but I havent seen it enough. Also, I dont see any discussion of throttling on high end cpus anymore, is this not an issue anymore?
 
D

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
I dont think Toms did it...it was the "others" I am talking about...We must have had 2-3 threads about reviews around the time the P4 6xx series came out regarding this....What they mainly were doing was forcing the cas timings to 3 (which is like low end value sticks of pc3200) to compare to higher latency DDR2 (which by the way cas 3 DDR2 is the high end performer)...The other site didn't have an NF3 or 4 chipset so they ran an older Via KT800 chipset....

Toms is more in his article favortism, botched IMO stability test he ran with a P4 next to the AMD64...Do you remember that??? he screwed up so much there it wasn't even funny....



It is still a problem....the 6xx series helped it a bit, but temperature problems are still an issues for many with prescott chips and using stock coolers....I expect to see more come summer time when it starts to heat up....My northwood would swing 6-8c in the summertime and I had very good cooling. Most prescotts dont have that headroom....
 
R

rgreen83

Senior member
Feb 5, 2003
766
0
0
I guess its good that I never read toms, just cause the site is Fugly and hard to navigate.

My suggestion: We should have a site akin to resellerratings, but to rate hardware review sites. If a site recommends a piece of hardware that you go out and buy on their suggestion and it performs like a leper, they should be punished with a lower rating!
 
S

Suspicious-Teach8788

Lifer
Feb 19, 2001
20,155
23
81
Tom's hardware is ok. I mean all sites have some bias, but I must read all sites in order to decide what I will buy. However, it's not cool that they never had a Winchester core review and their interactive CPU chart has Winchester on the KT890 chipset. Hi, I would like to know Winchesters on NF4 vs P4s on 925. That would be a fair benchmark. Otherwise, it's lame using their cpu chart when friggn Newcastles dominate Winchesters. LoL. Did Tom's hardware receive Opterons anyways? I mean after all we don't see many Opteron reviews. Nor do we see many Xeon reviews. This goes for ALL tech sites.
 
D

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
I dont know good point....Some of the sites that did reviews dont normally do opteron reviews as well, and one was a well known gaming site....


Toms is fine...I like his wide range of testing...he just doesn't do all the same test as everyone else....That being said he sometimes has greatt difference in same test versus other sites which makes you wonde. Also his cpu showdown (live thing he did) with AMD 4000+ versus a 3.4ghz P4 or something should that a bunch of his lab boys lack any real commonsense...
 
I

irwincur

Golden Member
Jul 8, 2002
1,899
0
0
I just want to know why there were only three review?

I am guessing that Intel is once again cutting checks and threatening sites. Nothing new here. It is just obvious that the chip that absolutely embarrasses Intel does not get reviewd - at all.

When Intel release DC, there were literally 20+ reviews.
 
A

ahock

Member
Nov 29, 2004
165
0
0
I dont think Intel is involve on this..... The only reason I can think of is I think AMD had just limited supplies. Intel did provide all of the new hardwares for the review.... This does shows that Intel have a lot of money in its pocket....
 
P

PetNorth

Senior member
Dec 5, 2003
267
0
0
I count eight reviews in this thread (anand, firings, techreport, linuxhardware, techannel, extremetech, trusted, hexus).

ahock, are you joking or are you telling seriously that there are not more reviews because AMD hasn't enought stuff to supplied it to web pages? LOL

let's be serious please...
 
A

ahock

Member
Nov 29, 2004
165
0
0
PetNorth.... What Imean is compare to Intel they can supply all the review sites for their free hardware.... Unlike for AMD which based from the other guys who posted is limited only
 
D

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
Originally posted by: ahock
PetNorth.... What Imean is compare to Intel they can supply all the review sites for their free hardware.... Unlike for AMD which based from the other guys who posted is limited only
Click to expand...



This number is comparable to the numers that did review the PD and PD EE in the first week....So was Intel having shipping issues??? NO, it was probably Intel knew it would favorable reviews from those sites...How did TechReport (one of the finest) get the shaft by Intel???
 
Zebo

Zebo

Elite Member
Jul 29, 2001
39,398
19
81
Originally posted by: ahock
PetNorth.... What Imean is compare to Intel they can supply all the review sites for their free hardware.... Unlike for AMD which based from the other guys who posted is limited only
Click to expand...

Your point is obsurd. Even if AMD could only make or afford to loose a limited number of chips you think Toms Hardware who is number 1 so far as tech traffic would'nt have one? Riiiiight. My bet is toms is still "gaming" the DATA.;)
 
The Boston Dangler

The Boston Dangler

Lifer
Mar 10, 2005
14,647
2
0
I've read Tom's for years. Please explain what they (or anyone) would have to gain by favoring 1 chip company over another.
Example: Tom's says P4s suck. Intel pulls their advertising. AMD or newegg or MS buys the ad spots. Problem solved.

This shouldn't be confused with unknowingly running tests that favor one or the other, or not testing a device adequately.

I do think the Intel versus AMD stability test was weak tho. I'm not even sure of it's purpose, as both chips demonstrate nearly 100% stability all over the world, every day.
 
I

imported_g33k

Senior member
Aug 17, 2004
821
0
0
Sometimes I don't like Tom's conclusions though, like this one for the 3400+ review.

That leaves us with a clear description of the Athlon64 3400+: it's a top quality CPU that's especially suitable for games and that also lives up to its model name - albeit only in this category. At the end of the day, it still lags slightly behind the Pentium 4(3.2ghz), a deficit that the 64-bit architecture could compensate for in the medium term, however.
Click to expand...
Tom's Article


If you look at Anand (and others) they have the 3400+ winnning in lots of other benchmarks besides just games. It could be that Tom's uses more synthetic benchmarks, which tend to favor intel?
 
Keysplayr

Keysplayr

Elite Member
Jan 16, 2003
21,219
54
91
Originally posted by: The Boston Dangler
I've read Tom's for years. Please explain what they (or anyone) would have to gain by favoring 1 chip company over another.
Example: Tom's says P4s suck. Intel pulls their advertising. AMD or newegg or MS buys the ad spots. Problem solved.

This shouldn't be confused with unknowingly running tests that favor one or the other, or not testing a device adequately.

I do think the Intel versus AMD stability test was weak tho. I'm not even sure of it's purpose, as both chips demonstrate nearly 100% stability all over the world, every day.
Click to expand...

And I suppose its your job to observe all of these said chips? Watch the blanket statements. They get you in hot water in here. ;)

 
I

Insomniak

Banned
Sep 11, 2003
4,836
0
0
Wow that POV ray stuff is impressive.

And to think that K8 will still only get better when 64 bit programs start showing up....it's hard to imagine that as well as Athlon 64 has done so far that it still has lots of unlocked potential...

God, Prescott was BORN at the end of its rope, and K8 just keeps getting better and better...AMDs engineers hit one hell of an elegant home run this time. This architecture just reeks of good design.
 
Keysplayr

Keysplayr

Elite Member
Jan 16, 2003
21,219
54
91
Originally posted by: Insomniak
Wow that POV ray stuff is impressive.

And to think that K8 will still only get better when 64 bit programs start showing up....it's hard to imagine that as well as Athlon 64 has done so far that it still has lots of unlocked potential...

God, Prescott was BORN at the end of its rope, and K8 just keeps getting better and better...AMDs engineers hit one hell of an elegant home run this time. This architecture just reeks of good design.
Click to expand...

Is it that AMD engineers hit one hell of a home run? Or more likely, Intel's engineers hit a few foul balls. Did one excel? Or did one stumble? Hmmmm. Ok, lets say a few strikeouts.

 
Z

ZobarStyl

Senior member
Mar 3, 2004
657
0
0
Originally posted by: keysplayr2003
Is it that AMD engineers hit one hell of a home run? Or more likely, Intel's engineers hit a few foul balls. Did one excel? Or did one stumble? Hmmmm.
Click to expand...
I've thought that for the last year or so it's been both: every AMD homerun is answered by an Intel foul ball (although we're talking desktop only). I mean, look at each step made recently, it hasn't been just Intel holding still and AMD stepping forward...90nm transition makes Intel chips hotter and scale no further, but AMD's 90nm run cooler and can clock higher. Intel's DC needs a new mobo, doesn't take full advantage of both cores (see Anand's Doom 3 bench while multitasking) and is clocked low compared to desktop, meaning that making the jump means you lose single thread performance, while AMD's just needs BIOS update, can feed both cores fine with memory controller, and are clocked just below their single core brethren. AMD has really pushed things forward recently, but with Intel it's either a step back or simply to the side (DDR2 anyone?)

 
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom