-
Should you upgrade your GPU or CPU for faster gaming? We tested many hardware combos to find out.
MAKE THE BAD MAN STOP. I WANT MY MOMMY.
\
<= Serf Pope
Hint:
(A) Visit a few Intel overclocking forums to establish Conroe's overclockability. Oh ... wait ... my bad. AMD fanbois think overclocking means +10% clock speeds :lol:
(B) Re-read the articles describing 45 nm transistor technologies (which AMD/IBM are madly scrambling to catch up with now) ... you may have to turn off your AMD fanboi Reality Distortion Field (RDF) to get the gist of the artcle.
(C) A + B makes 3.3-3.7 GHZ launch clocks seem conservative.
ok so again this is speculation and opnion.... u have no facts to prove penryn will clock at 3.3 - 3.7 ...if u have proof please post links
I mean how dare anyone want to see working silicon on an unreleased processor... i mean that makes them a AMD fanboi
http://www.zdnetasia.com/news/hardware/0,39042972,61984216,00.htm
"At a given clock frequency, a Barcelona core outperforms a current Opteron core by a factor of 1.8."
Is english your first language? If so I suggest you sue your teachers
:lol:
It is - and I'm an English major. I simply hadn't read that quote. I was thinking back on Randy Allen's interview and how he compared Barcelona to Clovertown and not K8.
MAKE THE BAD MAN STOP. I WANT MY MOMMY.
\
Hint:
(A) Visit a few Intel overclocking forums to establish Conroe's overclockability. Oh ... wait ... my bad. AMD fanbois think overclocking means +10% clock speeds :lol:
(B) Re-read the articles describing 45 nm transistor technologies (which AMD/IBM are madly scrambling to catch up with now) ... you may have to turn off your AMD fanboi Reality Distortion Field (RDF) to get the gist of the artcle.
(C) A + B makes 3.3-3.7 GHZ launch clocks seem conservative.
ok so again this is speculation and opnion.... u have no facts to prove penryn will clock at 3.3 - 3.7 ...if u have proof please post links
I mean how dare anyone want to see working silicon on an unreleased processor... i mean that makes them a AMD fanboi
It seems pretty clear that Intel will be able to do this. Doesn't make much sense to dispute it.
My impression of Barcelona is that it's meant for servers, not the desktop. That falls in line with an AMD statement last year revealing that they are now primarily after enterprise sales because of the sheer margins. In the server world, the processor logic core is a secondary consideration to the rest of the system architecture - caching, memory bandwidth, busses, storage, interconnects. In the enthusiast desktop world, it's almost the opposite.
The model numbers, TDP, and clock speeds listed on TrackingAMD seem quite realistic and within most expectations for the 65nm node. Those are quad-core devices, which are subject to increased heat and yield restrictions relative to dual-core parts.
For K8L, there is ample talk about L3 caching, doubling FPU resources, moving to a 128-bit SSE cycle, and widening the HT links. This really beefens up Barcelona for server and heavy FPU (high-performance computing) demands, but it does very little for the desktop market. The improvement in SSE is merely following in the steps of Conroe, and while the FPU is new, most specialized FPU code is found in the HPC market, not on desktops.
I feel I need not waste my breath explaining why desktop A64 does not need wider HT links or an additional level of caching. The IMC already communicates directly and quickly.
Based on the official statements, I believe that on a per-clock basis K8L will perform in between K8 and Conroe for most consumer applications while exhibiting considerable gains over both previous architectures in floating-point code. But while the processor may be churning through floating point 80 or 100% faster, one would have a hard time finding a practical desktop benchmark to show anywhere near those gains. (The only thing I foresee is that a K8L should make for an excellent folding machine.)
For 2S and 4S server applications, I have much less doubt that K8L will wipe the floor clean on introduction. The robust caching system and wide interconnects are just what large database servers have been asking for, and the strong FPU and bandwidth should find K8L a home in scientific computing. Ever notice the 40% and 80% performance improvement claims from AMD are referring to server benchmarks?
The model numbers, TDP, and clock speeds listed on TrackingAMD seem quite realistic and within most expectations for the 65nm node. Those are quad-core devices, which are subject to increased heat and yield restrictions relative to dual-core parts.
For K8L, there is ample talk about L3 caching, doubling FPU resources, moving to a 128-bit SSE cycle, and widening the HT links. This really beefens up Barcelona for server and heavy FPU (high-performance computing) demands, but it does very little for the desktop market. The improvement in SSE is merely following in the steps of Conroe, and while the FPU is new, most specialized FPU code is found in the HPC market, not on desktops.
I feel I need not waste my breath explaining why desktop A64 does not need wider HT links or an additional level of caching. The IMC already communicates directly and quickly.
Based on the official statements, I believe that on a per-clock basis K8L will perform in between K8 and Conroe for most consumer applications while exhibiting considerable gains over both previous architectures in floating-point code. But while the processor may be churning through floating point 80 or 100% faster, one would have a hard time finding a practical desktop benchmark to show anywhere near those gains. (The only thing I foresee is that a K8L should make for an excellent folding machine.)
For 2S and 4S server applications, I have much less doubt that K8L will wipe the floor clean on introduction. The robust caching system and wide interconnects are just what large database servers have been asking for, and the strong FPU and bandwidth should find K8L a home in scientific computing. Ever notice the 40% and 80% performance improvement claims from AMD are referring to server benchmarks?
MAKE THE BAD MAN STOP. I WANT MY MOMMY.
\
Hint:
(A) Visit a few Intel overclocking forums to establish Conroe's overclockability. Oh ... wait ... my bad. AMD fanbois think overclocking means +10% clock speeds :lol:
(B) Re-read the articles describing 45 nm transistor technologies (which AMD/IBM are madly scrambling to catch up with now) ... you may have to turn off your AMD fanboi Reality Distortion Field (RDF) to get the gist of the artcle.
(C) A + B makes 3.3-3.7 GHZ launch clocks seem conservative.
ok so again this is speculation and opnion.... u have no facts to prove penryn will clock at 3.3 - 3.7 ...if u have proof please post links
I mean how dare anyone want to see working silicon on an unreleased processor... i mean that makes them a AMD fanboi
It seems pretty clear that Intel will be able to do this. Doesn't make much sense to dispute it.
It seems pretty clear that Intel will be able to do this. Doesn't make much sense to dispute Precotts power. It will scale to 10 ghz.
Excuse me for not accepting that a TOTALLY NEW PROCESS will go over smooth with no problems and scale again with no problems and be on time.
I know that makes me a fanboi to want to wait and see and no accpet whatever Intel says. Just like with the camino chipset.
Should be the new forum MANTRA " Accept whatever INTEL says.... Makes no sense to dispute them"
MAKE THE BAD MAN STOP. I WANT MY MOMMY.
\
Hint:
(A) Visit a few Intel overclocking forums to establish Conroe's overclockability. Oh ... wait ... my bad. AMD fanbois think overclocking means +10% clock speeds :lol:
(B) Re-read the articles describing 45 nm transistor technologies (which AMD/IBM are madly scrambling to catch up with now) ... you may have to turn off your AMD fanboi Reality Distortion Field (RDF) to get the gist of the artcle.
(C) A + B makes 3.3-3.7 GHZ launch clocks seem conservative.
ok so again this is speculation and opnion.... u have no facts to prove penryn will clock at 3.3 - 3.7 ...if u have proof please post links
I mean how dare anyone want to see working silicon on an unreleased processor... i mean that makes them a AMD fanboi
It seems pretty clear that Intel will be able to do this. Doesn't make much sense to dispute it.
It seems pretty clear that Intel will be able to do this. Doesn't make much sense to dispute Precotts power. It will scale to 10 ghz.
Excuse me for not accepting that a TOTALLY NEW PROCESS will go over smooth with no problems and scale again with no problems and be on time.
I know that makes me a fanboi to want to wait and see and no accpet whatever Intel says. Just like with the camino chipset.
Should be the new forum MANTRA " Accept whatever INTEL says.... Makes no sense to dispute them"
You have some angst. It doesn't make sense to doubt the 45nm transition with high-k and its ability to scale clockspeed well.
MAKE THE BAD MAN STOP. I WANT MY MOMMY.
\
Hint:
(A) Visit a few Intel overclocking forums to establish Conroe's overclockability. Oh ... wait ... my bad. AMD fanbois think overclocking means +10% clock speeds :lol:
(B) Re-read the articles describing 45 nm transistor technologies (which AMD/IBM are madly scrambling to catch up with now) ... you may have to turn off your AMD fanboi Reality Distortion Field (RDF) to get the gist of the artcle.
(C) A + B makes 3.3-3.7 GHZ launch clocks seem conservative.
ok so again this is speculation and opnion.... u have no facts to prove penryn will clock at 3.3 - 3.7 ...if u have proof please post links
I mean how dare anyone want to see working silicon on an unreleased processor... i mean that makes them a AMD fanboi
It seems pretty clear that Intel will be able to do this. Doesn't make much sense to dispute it.
It seems pretty clear that Intel will be able to do this. Doesn't make much sense to dispute Precotts power. It will scale to 10 ghz.
Excuse me for not accepting that a TOTALLY NEW PROCESS will go over smooth with no problems and scale again with no problems and be on time.
I know that makes me a fanboi to want to wait and see and no accpet whatever Intel says. Just like with the camino chipset.
Should be the new forum MANTRA " Accept whatever INTEL says.... Makes no sense to dispute them"
You have some angst. It doesn't make sense to doubt the 45nm transition with high-k and its ability to scale clockspeed well.
It may scale very well... but we are assuming that the 45 process is going to be problem free..... we are now hearing trickles of clock speed limitations on Penryn and delays (overclockers.com)... Intel has no tried and true 45um process to lean on like it did with Conroe. With new metals...and this is all going to work flawlessly and ramp up to 4 ghz beacuse INTEL said so huh?
so yes i doubt INTEL can make it happen till i see scores...... same with barcelona
MAKE THE BAD MAN STOP. I WANT MY MOMMY.
\
Hint:
(A) Visit a few Intel overclocking forums to establish Conroe's overclockability. Oh ... wait ... my bad. AMD fanbois think overclocking means +10% clock speeds :lol:
(B) Re-read the articles describing 45 nm transistor technologies (which AMD/IBM are madly scrambling to catch up with now) ... you may have to turn off your AMD fanboi Reality Distortion Field (RDF) to get the gist of the artcle.
(C) A + B makes 3.3-3.7 GHZ launch clocks seem conservative.
ok so again this is speculation and opnion.... u have no facts to prove penryn will clock at 3.3 - 3.7 ...if u have proof please post links
I mean how dare anyone want to see working silicon on an unreleased processor... i mean that makes them a AMD fanboi
It seems pretty clear that Intel will be able to do this. Doesn't make much sense to dispute it.
It seems pretty clear that Intel will be able to do this. Doesn't make much sense to dispute Precotts power. It will scale to 10 ghz.
Excuse me for not accepting that a TOTALLY NEW PROCESS will go over smooth with no problems and scale again with no problems and be on time.
I know that makes me a fanboi to want to wait and see and no accpet whatever Intel says. Just like with the camino chipset.
Should be the new forum MANTRA " Accept whatever INTEL says.... Makes no sense to dispute them"
You have some angst. It doesn't make sense to doubt the 45nm transition with high-k and its ability to scale clockspeed well.
It may scale very well... but we are assuming that the 45 process is going to be problem free..... we are now hearing trickles of clock speed limitations on Penryn and delays (overclockers.com)... Intel has no tried and true 45um process to lean on like it did with Conroe. With new metals...and this is all going to work flawlessly and ramp up to 4 ghz beacuse INTEL said so huh?
so yes i doubt INTEL can make it happen till i see scores...... same with barcelona
Alright, fair. Go with it.
My thoughts:
K8L will perform very well, will dominate most server spaces (and this is a very widely held opinion), but merely make life more difficult for Intel the desktop / mobile space, where things will be a pretty haggard contest between the two in terms of overall performance, IPC, different benchmarks, etc. (kind of like the K7/P4 days). Intel will probably hold the mobile space, though, unless AMD manages to sneak some nice stuff in there. The 65nm process for AMD is still in question (with the potential to either be very good and scale above 3Ghz or poor and go no higher than 90nm), and 45nm for Intel is probably going to turn out very well assuming baaaaad things don't happen.
A brief rundown of what's in my head.
K8L will perform very well, will dominate most server spaces (and this is a very widely held opinion), but merely make life more difficult for Intel the desktop / mobile space, where things will be a pretty haggard contest between the two in terms of overall performance, IPC, different benchmarks, etc. (kind of like the K7/P4 days). Intel will probably hold the mobile space, though, unless AMD manages to sneak some nice stuff in there. The 65nm process for AMD is still in question (with the potential to either be very good and scale above 3Ghz or poor and go no higher than 90nm), and 45nm for Intel is probably going to turn out very well assuming baaaaad things don't happen.
A brief rundown of what's in my head.
I can accept that... because you atleast entertain the thought that bad thigns CAN happen to Intel 45um process.... people here just assume INTEL will execute pefectly on thier TICK TOCK plan
i wanna wait and see.... so that makes me a AMD paid shill
I'm with Intel in this battle tbh.
Current core 2 duos have plenty of headroom, the 45nm shrink should in theory increase the headroom. OBV we will have to wait and see on this one.
Lets put something else straight, the Core2 clock speeds are only as low as they are so Intel can make money.
Think about it, release a new cpu, sell it so its better than any competitors cpu. After 6 months change your FSB from 1066 to 1333. Release what is essentially the same cpu and extend product lifetime / profitability.
Bare in mind, the mass "dumb" consumer does not overclock!
They can easily up the multipliers too. Current Core2's are sold with overclocking headroom, they could quite easily sell them without that headroom. They could do that now, but no one would need to buy an Intel processor again for a good while and they wouldn't be able to sell anything better in their range until they increased their performance/headroom again.
Current core 2 duos have plenty of headroom, the 45nm shrink should in theory increase the headroom. OBV we will have to wait and see on this one.
Lets put something else straight, the Core2 clock speeds are only as low as they are so Intel can make money.
Think about it, release a new cpu, sell it so its better than any competitors cpu. After 6 months change your FSB from 1066 to 1333. Release what is essentially the same cpu and extend product lifetime / profitability.
Bare in mind, the mass "dumb" consumer does not overclock!
They can easily up the multipliers too. Current Core2's are sold with overclocking headroom, they could quite easily sell them without that headroom. They could do that now, but no one would need to buy an Intel processor again for a good while and they wouldn't be able to sell anything better in their range until they increased their performance/headroom again.
tamalero
Distinguished
- Oct 25, 2006
- 1,145
- 149
- 19,470
so you admit you have no facts or penryn scores to back u up ...just specualation...so u give up
fair enuff
no I never said that, I said I'm getting bored of this endless discussion, so once again, I'm out for now
I wonder how many times you have said that as for now..
15%?
stop smoking weed...
in this post you're not better than shakirou and baron, you keep tossing stuff out of your hat. pure speculation wich you seem to call them "FACTS"
so you admit you have no facts or penryn scores to back u up ...just specualation...so u give up
fair enuff
no I never said that, I said I'm getting bored of this endless discussion, so once again, I'm out for now
I wonder how many times you have said that as for now..
15%?
stop smoking weed...
in this post you're not better than shakirou and baron, you keep tossing stuff out of your hat. pure speculation wich you seem to call them "FACTS"
thanks for picking that up... lot of so called FACTS in this thread
dasickninja
Splendid
- May 16, 2006
- 7,557
- 0
- 30,810
dasickninja
Splendid
- May 16, 2006
- 7,557
- 0
- 30,810
BaronMatrix
Splendid
- Dec 14, 2005
- 6,655
- 0
- 25,790
so you admit you have no facts or penryn scores to back u up ...just specualation...so u give up
fair enuff
no I never said that, I said I'm getting bored of this endless discussion, so once again, I'm out for now
I wonder how many times you have said that as for now..
15%?
stop smoking weed...
in this post you're not better than shakirou and baron, you keep tossing stuff out of your hat. pure speculation wich you seem to call them "FACTS"
I'll excuse your ignorant little barb and say that if these model numbers are true AMD is going right after Kentsfield and the enthusiast market, not to mention Clovertown and the 2P market.
That's going to change the landscape significantly by the end of summer when AMD is set to be 50% 65nm.
TRENDING THREADS
-
-
Question Can you increase the size of an existing partition in Windows without losing data?
- Started by Balic Blackthorn
- Replies: 11
-
-
Question Why does my GPU not work in a PCIe x16 (at x4) slot?- Started by Arturas Kulickis
- Replies: 8
-
Question Sabertooth x79 Rebuilt PC Won’t Start Properly – Output Issues After Running for a Long Time / Turning It Off- Started by Changali
- Replies: 2
-
Facebook
Twitter
Instagram