Tom's 3.06 GHz HT Review

As the title says , Tom has a review of a 3.06 hyperthreading P4 :

http://www.tomshardware.com/cpu/02q4/021114/index.html

hey is th 3.06 ghz chip still on the .13 micron process??

Yup it is still the .13 micron .

Here is Anand's : http://www.anandtech.com/cpu/showdoc.html?i=1746

EDIT :

Firingsquad : http://firingsquad.gamers.com/hardware/p43000/default.asp

Gamer's Depot : http://www.gamersdepot.com/hardware/cpus/intel/p4_3.06ghz/001.htm

i wonder if they'll change the 3D mark results like they are on [H].

for those that don't know what i'm talking about, [H] have also reviewed the 3.06 and somehow they were the only ones able to get an improved 3D mark score when they enabled HT.

a 1000 point improvement. everyone else had no change.
all of a sudden the [H] review has changed.
the difference became 450 points. now it's suddenly 879 points.

you have to wonder about the quality of these review sometimes, don't you?


EDIT: http://www.hardocp.com/article.html?art=Mzg4

looktall said:

i wonder if they'll change the 3D mark results like they are on [H].

for those that don't know what i'm talking about, [H] have also reviewed the 3.06 and somehow they were the only ones able to get an improved 3D mark score when they enabled HT.

a 1000 point improvement. everyone else had no change.
all of a sudden the [H] review has changed.
the difference became 450 points. now it's suddenly 879 points.

you have to wonder about the quality of these review sometimes, don't you?


EDIT: http://www.hardocp.com/article.html?art=Mzg4

Click to expand...

Odd happenings... I thought the HardOCP review was rather positive myself but I attributed it to 'real' performance... maybe that was too soon then.

Believe it or not Tom has the best review out there and coupled with his video gives a good picture of hyperthreading's gains and losses . The key to this is showing the chip with and without HT enabled . It allows me to summarise nicely the salient points :

1/ HT causes for the most part a small performance hit in single threaded apps ( most apps and games ) . Much smaller than the massive drops in the earlier versions of hyperthreading .

2/ Only things like sysmark 2002 , synthetic mem bandwidth and 3d studio max show some hyperthreading gains .

3/ HT is still a good tech and in his video shows that when doing more than one thing at a time the HT enabled 3.066 P4 eclipses his unlocked 3.6 chip !

4 / The athlon 2700 + 2800 + are still excellent performers and since they can be easily unlocked will also do very well .

5/ The ATI 9700 Pro offers very good cpu scaling , just look at the 3.6 ghz 3dMark 2000 score !

All in all HT is a good tech for multitasking which we all do but don't expect the gains when just doing one thing . The next step I think is to compare the performance vs single and dual cpu systems in apps known to have SMP support . And also to show the HT performance in chips like the 2.4 B . IMHO of course .

EDIT :BTW I am not really sure what is going on over at [H] , but mistakes do happen . In fact one or 2 places in Tom's charts I see mislabelling ( once ) or the wrong colour used for the wrong cpu .......... will probably be fixed soon .

It seems Sharky has some different review results and the HT there seems to outperform the chip without HT in most tests .

http://www.sharkyextreme.com/hardware/cpu/article.php/3261_1500631__7

Say what you want, but I like tom's., And I LOVE that video LoL! Too bad my XP won't do that LoL!

FYI the age of Apple owning Adobe products was already dead, but now the coffin is nailed shut!

-Toysrme

Anandtech reports a loss of performance in virtually no single threaded applications, adn so does extremetech. I think multi-treading is the wave of the future, it will be interesting to see how this type of technology progresses
________
Ferrari tipo 500 specifications

If I recall anand did not show the effects in the benches with HT turned off . That is why it isn't an ideal review . HT has improved a great deal since it was first shown on Xeons . Back then the performance hit was so huge in some apps that a cpu of given speed could fall below the performance of a cpu clocked lower . When you run this new HT it cuses a much smaller performance drop relative to the 3.066 with HT disabled . This clearly shows there is still a drop in performance though it has improved .

Only in around 2 graphs at Tom's for example does the performance drop cause the 3.066 P4 to dip below 'slower' P4 CPUs .

totally disappointment 4 me
another "new tech bla bla bla" from intel,
i don't like intel's marketing skills,

tons of (c)'ed tech, (like Hyperthreading, etc...)
bad price/perf rate (that's why i buy amd)

another similar example from ati vs nv:

tons of (c)'ed tech, (like lightspeed mem, infinite fx, etc...)
bad price/perf rate
(that's why i'll buy 9700pro instead of nv30)
(some day 9700pro price'll drop down or i'll gettin' rich lightyears later! )

"dual performance on single cpu"
not yet, we still need dual cpu 4 "real dual performance"

Yes but did you see how much faster multitasking on the toms video the HT 3.06 was compared to the unlocked at 3600?

Even if there was a loss in performace with single applications. How many of you ALWAYS have background programs running or multiple tasks running at the same time? I know I always do LoL!

Also you set your sites to high. Am I the only one who realized that both computers are running through some of the benches they use AT THE SAME TIME.

First part I can make out:
File compression(compressing), Adobe photoshop, MS Powerpoint in the first half

Second part includes:
Pinnical Studio(encoding),3D Studio Max, UT2k3(Running the Antalus flyby benchmark)

Now I don't know about you... but how much performance are you asking for? jesus I mean the 3.6 get's slaughtered. The 3.06 is stuck waiting on the 3.6 before the test at one point.

386 what kind of performance were you expecting??? If you ask me they delivered. And I've got just as many intel boxes as I do AMD boxes. Maby you should set yours sites to a more realistic goal, or get off your 386 and download the video for yourself. j/k

-Toysrme

ok, i know there's a hi perf@lower mhz, huh?
and intel says "that's a new great technology 4 cpu's history",

i'm usin' 1800xp@1533mhz, and it faster than p4@2ghz!

ok that's a positive feature of intel, (improved multitasking)
but we still remember "old motorola cpus" and "amiga"

"HyperThreading speeds up single threaded applications a little bit by handling the OS tasks in the background on the second logical CPU (See Comanche framerates for example: +2.6%)
HyperThreading speeds up multiple single threaded applications quite a bit (See Comanche (+7.1%) and RAR compressing results (+ 7.5%))
HyperThreading speeds up multithreaded applications a lot (Virtual Dub MPEG4 encoding + 20%)
But seems to have a little trouble with a multi thread applications and some single threaded applications at the same time. "

"If you like to compile, Animate, Encode MPEG4 or render on the same desktop system on which you play games, HyperThreading as a lot to offer"

Both from Ace's Hardware. For those of you who don't like Anand and TH
That should say it all.
-Toysrme

ok, i have no problem with intel and hypertreading...

i just disappointment multi-processing performance,
intel sells his products with "mhz" and "some HT like features"

but i say again, that's not a real "multiprocessin',
and intel r liar again...

not sure what to make of hyper threading for the average user. doesn't seem like right now the average dumb user would benefit at all from this.

i'm guessing the 3.06 is the first HT enabled chip. but would a lower clocked base chip which would overclock better perform better in the same battery of tests?

i'd like to see that.

The benefits to multitasking and multithreading are obvious but there are still only a few single threaded apps eg games that benefit from it . Most get a drop .

The thing I put to you Toysrme is that the main benefit is running very intensive apps along with other stuff . But most ppl will still choose to run that intensive app by itself . Not with other hungry tasks .

Task A+B may take 10 mins without HT enabled but 8:30 with it on .......... very good . But running task A by itself without HT is 6 mins . What I'm saying is that many ppl will take task A and run it by itself esp if it is important and task B is winamp or a dvd movie . Many will still rather 6 mins than 8 1/2 .

here is the anand graph, they do compare HT to HT http://www.anandtech.com/cpu/showdoc.html?i=1746&p=6

how do i attach images bigger than 300x300?
________
PRETEEN ANGELS

maybe just give us the link