| View previous topic :: View next topic |
| Author |
Message |
jack
Joined: 27 Jun 2007 Posts: 359
|
Posted: Tue Mar 25, 2008 9:04 am Post subject: |
|
|
Technically, Itanium also was always "on schedule"....
There were several roadmaps in last Autumn that showed 2.6GHz Phenom X4 scheduled for Q4 2007.
For example:
"Ever wonder why we didn't have an early look at Phenom like we did for every Core 2 processor before the embargo lifted? Not only are CPUs scarce, but AMD itself didn't really know what would be launching until the last moment.
At first Phenom was going to launch at either 2.8GHz or 2.6GHz; then we got word that it would be either 2.6GHz or 2.4GHz. A week ago the story was 2.4GHz and lower, then a few days ago we got the final launch frequencies: 2.2GHz and 2.3GHz."
http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3153
|
|
| Back to top |
|
 |
Phenom
Joined: 22 Mar 2008 Posts: 4
|
Posted: Tue Mar 25, 2008 9:23 am Post subject: |
|
|
| jack wrote: | Technically, Itanium also was always "on schedule"....
There were several roadmaps in last Autumn that showed 2.6GHz Phenom X4 scheduled for Q4 2007.
|
…non official. Everything else is just hear-day "investigation" of the yellow web IT press… and yes that includes Anand's lamenting over initial K10 frequencies in a range of 2.6-2.8 GHz! :roll:
|
|
| Back to top |
|
 |
jack
Joined: 27 Jun 2007 Posts: 359
|
Posted: Tue Mar 25, 2008 11:51 am Post subject: |
|
|
| Phenom wrote: | | jack wrote: | Technically, Itanium also was always "on schedule"....
There were several roadmaps in last Autumn that showed 2.6GHz Phenom X4 scheduled for Q4 2007.
|
…non official. Everything else is just hear-day "investigation" of the yellow web IT press… and yes that includes Anand's lamenting over initial K10 frequencies in a range of 2.6-2.8 GHz! :roll: |
That's the point. Officially, nothing is ever delayed, also different version of Itanium were "officially" on schedule.
Anand didn't make those launch frequencies by himself, he was simply relaying information that AMD gave to him.
By the way, did you forget that AMD made a lot of presentations last Autumn with 2.5GHz Barcelona results (there were some SPEC submissions also) and even Johan mentioned in his article, that 2.5GHz Barcelona will be launched in Q4 2007. But I guess, "officially" there is no delay in 2.5GHz Barcelona shipments....
|
|
| Back to top |
|
 |
Johan
Joined: 23 Jul 2007 Posts: 162 Location: Belgium
|
Posted: Wed Mar 26, 2008 9:25 am Post subject: |
|
|
| jack wrote: | | Phenom wrote: | | jack wrote: | Technically, Itanium also was always "on schedule"....
There were several roadmaps in last Autumn that showed 2.6GHz Phenom X4 scheduled for Q4 2007.
|
…non official. Everything else is just hear-day "investigation" of the yellow web IT press… and yes that includes Anand's lamenting over initial K10 frequencies in a range of 2.6-2.8 GHz! :roll: |
That's the point. Officially, nothing is ever delayed, also different version of Itanium were "officially" on schedule.
Anand didn't make those launch frequencies by himself, he was simply relaying information that AMD gave to him.
By the way, did you forget that AMD made a lot of presentations last Autumn with 2.5GHz Barcelona results (there were some SPEC submissions also) and even Johan mentioned in his article, that 2.5GHz Barcelona will be launched in Q4 2007. But I guess, "officially" there is no delay in 2.5GHz Barcelona shipments.... |
To be fair, I should have added that AMD didn't expect mass shipments of 2.5 GHz in Q1 (but that was made clear just after the article). The plan was to give the OEMs a lot of 2.5 GHz in Q4. Well, that is what AMD (different people) said to us.
Now the official line is 2.5 GHz SE (120W) in "late" Q2. Which is disappointing of course.
|
|
| Back to top |
|
 |
Groo
Joined: 22 Jul 2007 Posts: 178
|
Posted: Wed Mar 26, 2008 5:21 pm Post subject: |
|
|
[quote="Johan"] | jack wrote: |
Now the official line is 2.5 GHz SE (120W) in "late" Q2. Which is disappointing of course. |
That 'official' line is way way way off. Think next week or before for a more realistic time schedule.
-Charlie
|
|
| Back to top |
|
 |
LiamC
Joined: 23 Jul 2007 Posts: 82
|
|
| Back to top |
|
 |
jack
Joined: 27 Jun 2007 Posts: 359
|
Posted: Thu Mar 27, 2008 9:17 am Post subject: |
|
|
[quote="Groo"] | Johan wrote: | | jack wrote: |
Now the official line is 2.5 GHz SE (120W) in "late" Q2. Which is disappointing of course. |
That 'official' line is way way way off. Think next week or before for a more realistic time schedule.
-Charlie |
Keep in mind that we were talking about availability of 2.5GHz Barcelona and not Phenom X4. Are you sure that 2.5GHz Barcelona will be available next week?
|
|
| Back to top |
|
 |
jack
Joined: 27 Jun 2007 Posts: 359
|
Posted: Thu Mar 27, 2008 9:28 am Post subject: |
|
|
Q6600 seems to win more CPU limited benchmarks in that test.
Xbitlabs was able to overclock 9850 to 2.7GHz with raising voltage from 1.3V to 1.4V: http://www.xbitlabs.com/articles/cpu/display/phenom-x4-9850.html
However, in that becnhmark suite 9850 lost to Q6600 in every benchmark and Q9300 was even more ahead.
|
|
| Back to top |
|
 |
Johan
Joined: 23 Jul 2007 Posts: 162 Location: Belgium
|
Posted: Thu Mar 27, 2008 11:36 am Post subject: |
|
|
| jack wrote: |
Q6600 seems to win more CPU limited benchmarks in that test.
Xbitlabs was able to overclock 9850 to 2.7GHz with raising voltage from 1.3V to 1.4V: http://www.xbitlabs.com/articles/cpu/display/phenom-x4-9850.html
However, in that becnhmark suite 9850 lost to Q6600 in every benchmark and Q9300 was even more ahead. |
I really would like to understand the effects of the cache a bit more. Considering that each core in Barcelona has only a 512 KB cache before it has to go to a slow L3, I can imagine that is far from optimal for most desktop apps. Even Shangai will not change this. Penryn with it's huge low latency L2-cache is really ideal for desktop apps.
|
|
| Back to top |
|
 |
Pjotr
Joined: 06 Aug 2007 Posts: 159
|
Posted: Thu Mar 27, 2008 1:16 pm Post subject: |
|
|
| Johan wrote: | | I really would like to understand the effects of the cache a bit more. Considering that each core in Barcelona has only a 512 KB cache before it has to go to a slow L3, I can imagine that is far from optimal for most desktop apps. Even Shangai will not change this. Penryn with it's huge low latency L2-cache is really ideal for desktop apps. |
And where would that put Nehalem, going away from the big L2 to smaller L2 with a separate L3 instead? Perhaps it is a better design when you are not limited by a FSB.
|
|
| Back to top |
|
 |
jack
Joined: 27 Jun 2007 Posts: 359
|
Posted: Thu Mar 27, 2008 1:43 pm Post subject: |
|
|
| Pjotr wrote: | | Johan wrote: | | I really would like to understand the effects of the cache a bit more. Considering that each core in Barcelona has only a 512 KB cache before it has to go to a slow L3, I can imagine that is far from optimal for most desktop apps. Even Shangai will not change this. Penryn with it's huge low latency L2-cache is really ideal for desktop apps. |
And where would that put Nehalem, going away from the big L2 to smaller L2 with a separate L3 instead? Perhaps it is a better design when you are not limited by a FSB. |
That will depend on L3 latency. However, based on previous designs, it should be quite decent. Certainly better than Barcelona's L3 which latency is quite horrible.
Does anyone remember L3 latency of Tulsa?
|
|
| Back to top |
|
 |
Johan
Joined: 23 Jul 2007 Posts: 162 Location: Belgium
|
Posted: Thu Mar 27, 2008 2:58 pm Post subject: |
|
|
| jack wrote: | | Pjotr wrote: | | Johan wrote: | | I really would like to understand the effects of the cache a bit more. Considering that each core in Barcelona has only a 512 KB cache before it has to go to a slow L3, I can imagine that is far from optimal for most desktop apps. Even Shangai will not change this. Penryn with it's huge low latency L2-cache is really ideal for desktop apps. |
And where would that put Nehalem, going away from the big L2 to smaller L2 with a separate L3 instead? Perhaps it is a better design when you are not limited by a FSB. |
That will depend on L3 latency. However, based on previous designs, it should be quite decent. Certainly better than Barcelona's L3 which latency is quite horrible.
Does anyone remember L3 latency of Tulsa? |
Yes...
http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=2897&p=2
109 cycles.
|
|
| Back to top |
|
 |
Groo
Joined: 22 Jul 2007 Posts: 178
|
Posted: Thu Mar 27, 2008 3:07 pm Post subject: |
|
|
| jack wrote: |
Keep in mind that we were talking about availability of 2.5GHz Barcelona and not Phenom X4. Are you sure that 2.5GHz Barcelona will be available next week? |
Oh, sorry, I thought you meant the core. Sorry about that.
-Charlie
|
|
| Back to top |
|
 |
EaS
Joined: 09 Sep 2007 Posts: 8
|
Posted: Thu Mar 27, 2008 4:41 pm Post subject: |
|
|
Some of the first details on IPC gains that AMD expects going from Barcelona to Shanghai:
"AMD is expecting, clock for clock, 45nm Deneb based Phenom cores to offer up to 15% more performance."
Reminds me of something:
So essentialy AMD expects Shanghai to deliver the same performance improvement as K8 -> K10.
|
|
| Back to top |
|
 |
Alberto
Joined: 04 Sep 2007 Posts: 111 Location: Italy
|
|
| Back to top |
|
 |
|