X-bit labs

Started by: dudde | Date 07/15/07

Comments: 24 | Last Comment:  02/27/08

[1-17]

1. these new chips coupled with Intel's price cuts should tighten the grip around AMD's complacent neck!

2. nice 10/10 review!

I cant wait until the QX6600 drops to $266 at the end of july.

3. 1333 MHz- 1066MHz != 66MHz

Your talking about effective FSB, the real FSB frequency is quad pumped.
1333 MHz- 1066MHz = 266MHz
66MHz * 4 = 266MHz

In regard to the article, I was hoping for a stronger emphasis on multi tasking scenarios and not just multi threaded aware apps.

4. Hmmm...I thought these 3GHz Quad-Cores were already available in the Mac Pros for a month or two.

Those are Clovertown for Dual Socket motherboards. These is for Single Socket motherboards.

The ones in the Mac Pro are also unofficial ones on the old stepping with 150W TDPs. The official X5365 will have the newer stepping and have a 120W TDP.

5. Does Prime95 really Burn the CPU?
Does Prime95 really load the 4 cores? Or even 2?

Prime95 is one of the best CPU burner in large FFTs mode.
It supports multicore CPU from version 25.

6. keep up the good work

7. Can I learn, what the stepping of q6600, q6700 and qx6800 is ?

8. Well, this sucks. My Pentium 4 Northwood is going to Pentium 1 world now

there was no pentium 1. i think the closest thing to that was the 286 and 386 lines.

There was a pentium 1 ..only called "pentium"

chips...
8086,8087,8088,286,386,486,pentium,peniumII,pentiumIII,pentiumIV ...then to numbered names like AMD also strangel to know TI (texas instruments) also made 486 chips

9. I'm still happy with my QX6800 running at 3.2GHz and doing it very cool. It only got 1.5deg C warmer than the stock speed and is stable as can be. Also, it was only $200 Dollars more than the new Kid on the Block and has a reputation behind it. I really doubt I'd see much improvement in the new CPU but Kudos to Intel for making it cheaper.

10. Chat with me
www.olegator.com
i have a webcam!

11. Matlab results are very strange, as usual... When running version 7.1 (apparently old now), everything was running faster when the process was attached to only one core (even the bench routine)... While this may be blamed on the Windows XP scheduler, I don't really think that this is the reason for the results we see here.

The engineers at MathWorks SHOULD have began parallelizing the application years earlier, their efforts to catch the last train now lead to comic situations... It's by no means an easy task, but it must be accomplished if they want to retain their position. It's a shame that one of the most widely used engineering applications is still in the stone era, unable to utilize hardware that has been around since years now...

Just in case - were the tests conducted with the proper BLAS versions? (I'm not sure if the optimized versions are installed by default now, as of version 2007a...)

12. I would very much like to see some video transcoding tests with different video editors (e.g., Vegas 7, Video Studio 11, ...). Many editors have already been modified to make efficient use of multiple processors. Useful transcodings would be Mpeg2 HDV to highly compressed Divx HD or downrezzed to standard DVD quality. Another interesting test: AVCHD to Mpeg2 HDV program stream.

FYI, I recently suffered thru a 12 hour transcode of a 40 minute HDV video to Divx HD (using old 3GHz P4-HT), I would really appreciate a computer that could lop off 9 hours from this process...

13. Nice

14. what is diffrent between 965 cpu and core 2
extreme qx6850.which one is faster?

15. i am waiting for respones!

i guess you mean the pentium extreme edition 965.

the core 2 model is likely about 50% faster in single

threaded and dual threaded apps, and depending on

how well the 965's ht works in 4 thread apps, the core

2 is probably between 50% faster and 3x as fast.

16. Core 2 Extreme QX6850 is very strong !

17. Excellent. Well Done

[1-17]