AMD is very "accurate" wit their pricing.
i still remember wen phenom launch, dey calculated how many % it is slower den q6600 & price phenom exactly dat percentage bellow q6600 price........

Haha..Look at the graphics counterpart already know..
But i hope the pricing is right

3DMark does not have linear scaling.

With respect to a 3DMark shootout from Computerbase, the Ph-II 940 is 2.5% faster than the 9550, but 1.5% slower than the 3Ghz Q9650. It's about 7% faster than the Phenom I @ 3Ghz.


And Phenom II prices are revised. Now it's an Italian source, but we remove the VAT.



This post has been edited by X.E.D: Nov 27 2008, 10:18 AM

OT abit..maybe due to P-II hype, amd decides to do price cut


http://www.techpowerup.com/77249/AMD_Price..._Announced.html
still unconfirm for msia

If AMD price their new CPU higher than i7, they are digging their own grave...

huhuhuhu im waiting and saving my money... now almost rm2k huhuhu....

amds_upcoming_phenom_ii_cpus_get_priced

haha..digging their own grave..lol
i think the price should be around c2q only...more than that...maybe people may jumpship to intel

Actually, it does have almost linear scaling. Look at the Q9450, Q9550 and QX9650..

You mean this one? http://www.pcgameshardware.com/&menu=brows...1&show=original



That article does not have Q9400 but it has the Q9300 and Q9450.

From the PCGH graph.... Q9450 = 4081 and Q9300 = 3811, that's only 7.08% difference. However the AMD graph shows around 15.2% difference between Q9300 and Q9400, and the Q9450 is slightly faster than the Q9400. So obviously the scaling in the AMD graph is different, more than twice that of PCGH graph.

Lets assume Q9450 as Q9400.. If 4081 at 91%, to 3811 at 79% is (91% - 79%) / 79%= 15.2% on the AMD graph, so obviously 12% difference equals 15.2%. This 9% difference from Q9400 to Phenom II 940 would be around (9% * 15.2%) / 12% = 11.4%.

Thus 11.4% / 15.2% * 7.08% = 5.31% on the PCGH graph. Total difference from Q9300 to Phenom II 940 is 5.31% + 7.08% = 12.39%. That puts the score around...

3811 * 112.39% = 4283

That puts the Phenom II 940 a bit behind Q9550 at 4284... Much more than 1 point because Q9400 is a bit slower than Q9450.

I guess you saw this http://theovalich.wordpress.com/2008/11/26...-less-than-400/

This post has been edited by lex: Nov 27 2008, 08:31 PM

OMG

Who would fall for this trap?

I took the performance per clock scaling from Q9550 to Q9650. 6+ % clock difference, but ~ 5% perf difference.
So I applied it to Q9400 from Q9300 scores, and calculated a Q9400 3DMark score. It's about 4001.

The score is in line with most sightings of Q9400 vs Q9450, only 100 points more at max. (Q9450 is 4081 here, 80 points, close to that)


If 4001 = 91% (AMD graph), 100% = 4397. Q9550 is 4284.


If you use the Q9300 graph it's even crazier, 940 is at the same level with QX9770. But let's have a level headed estimation.

This post has been edited by X.E.D: Nov 27 2008, 10:14 PM

but 1.5v for 4.0 is abit high le ba....

The DELTA scaling on that graph is different..

Example 1: If you take Q9300 as 3811 and upscale from 79% to 91% for Q9400, that would put Q9400 at 4389... which is incorrect!

Example 2: If you use 4001 for Q9400 and down-scale from 91% to 79% for Q9300, that would put Q9300 at 3473... which is again incorrect!

My calculations are much better..

This post has been edited by lex: Nov 29 2008, 08:09 AM

Your calculations have an assumption (that is considered not true already).


Occam's razor. )))))))))

This post has been edited by X.E.D: Nov 27 2008, 10:39 PM

My assumption is way more accurate than your simple assumptions (calculation).

Due to Q9400 missing in that PCGH graph, I used the nearest processor which would be the Q9450..

The Q9400 is a little slower than the Q9450. If I had a real Q9400 score, the performance scaling from Q9300 to Q9400 much lower, and putting the PII 940 further behind the Q9550.

This post has been edited by lex: Nov 27 2008, 10:52 PM

You're re-averaging the graph, because Q9300 does not match with Q9400/9450.

You're assuming both datasets are wrong because they are wrong relatively (true, but one should be correct objectively, or this graph is all lies)

I took into account the DELTA scaling on both graphs. That is how I calculate the outcome. Didn't you notice the percentages are different.

Even the steeper "slope" (delta) from Q9300 to Q9400 can be seen, which means the AMD graph has used smaller range scaling between PII 940 as 100% and Q9300 as 79%, unlike the "infinite" scaling on the PCGH graph.

Since the PII 940 score is unknown, I took the Q9450 score as Q9400 (both same clocks, just different cache thus scores are very close) and scale the score from there using the Q9300 as the base.

Hmmm... Phenom II.... where are youuuu....? We all really miss youuuu...

Phenom II 2.8Ghz priced at USD266 is sweet! By the time it hits the market... It'll cost lesser... My budget is RM600-700... So, I'll just wait for the durian to fall la...

http://theovalich.wordpress.com/2008/11/26...-less-than-400/

That would put right smack in Q9400 to Q9450 territory.. We already guessed the prices much earlier in this thread. See...

Not so fast yet, you may want to wait till January and see...

Already posted.. See few posts above...

i want more.... more...... im hungry..... huhuhuhuhuhuhu