CPU Power
CPU Power
well lets see whos got what........
when you vote, post your actual speed
OC People post your stock speed and then OC'ed speed
3.06 GHz HT .....for me
when you vote, post your actual speed
OC People post your stock speed and then OC'ed speed
3.06 GHz HT .....for me
- Admiral LSD
- DBB Admiral
- Posts: 1240
- Joined: Sun Nov 18, 2001 3:01 am
- Location: Northam, W.A., Australia
- Contact:
- STRESSTEST
- DBB DemiGod
- Posts: 6574
- Joined: Sun Nov 21, 1999 3:01 am
- Will Robinson
- DBB Grand Master
- Posts: 10135
- Joined: Tue Mar 07, 2000 3:01 am
- CDN_Merlin
- DBB_Master
- Posts: 9781
- Joined: Thu Nov 05, 1998 12:01 pm
- Location: Capital Of Canada
- Vindicator
- DBB Benefactor
- Posts: 3166
- Joined: Mon Dec 16, 2002 3:01 am
- Location: southern IL, USA
- Contact:
- STRESSTEST
- DBB DemiGod
- Posts: 6574
- Joined: Sun Nov 21, 1999 3:01 am
- whuppinboy
- DBB Benefactor
- Posts: 725
- Joined: Sun Jun 03, 2001 2:01 am
- Contact:
- DarkShadow
- DBB Admiral
- Posts: 1371
- Joined: Thu Apr 29, 1999 2:01 am
- Location: Pittsburgh, PA, USA
- Contact:
- Mobius
- DBB_Master
- Posts: 7940
- Joined: Sun Jun 03, 2001 2:01 am
- Location: Christchurch, New Zealand
- Contact:
Sort of a pointless poll really - because even Intel now admit that the "Megahertz Myth" was true all along. IPC, power requirements and heat are all more important than actual clocks.
For example, my old T-Bird 1.0Ghz overclocks to 1.5Ghz, and it beats the original P4 1.8Ghz hands down in pretty much every benchmark.
Currently, AMD64s handily dispatch the very fastest Pentium 4s when they are a full 1000MHz slower than the "fastest" Intel chips.
Personally, I am happy with this shift in paradigm. It means it's going to be much easier explaining to a client why they should go with a mobile chip "2GHz slower than the Pentium 4" when model numbers reflect performance, and not clock rates.
It's been coming for a while now really. Remember when clock rates hit 3.0 GHz? Hell, that was a long time ago now - and in all this time Intel have only succedded in adding a measley 400Mhz to their fastest cores.
They have however dones some good stuff with Hyperthreading, cache-size, better data prediction and branch prediction, memory subsystem speed and data rate.
So - a P43.4GHz EE (Or even a Northwood 3.4) would beat the living snot out of an original Williamette 3.0 overclocked to 3.4.
This is only to be expected given the problematic state of 90nm designs and production. It'll only get worse at 65nm, and a great deal of time and money is going to be spent in the next few years wringing every last IPC (Instruction Per Clock) out of the existing processes.
This is going to mean double and quadruple cores per die, massively increased on-die cache sizes, extreme optimization of execution engines, and huge surges in data throughput.
While the boffins slave away at creating the next paradigm for computing, the rest of the boffins will be preparing the path for ultra-efficient computing. Silcon is the fourth paradigm for computing. The first was mechanical, the second was valves, the third was transistors and Silicon chips are the fourth.
The fifth paradigm for computing will likely be largely self-assembled nano-scale processors incorporating carbon nanotubes, Tri-gate transistors, and other currently nonexistent stuff.
When the fifth paradigm is enabled, computing speeds will rapidly escalate again until they reach "Moore's Law" as plotted from the 60s.
Make no mistake; the current lack of increase in clock speeds is only a temporary hiccup on the exponential-computing-power road. The lessons we learn during this lull will allow the ensuing paradigm shift to make extraordinary gains in computer power.
For example, my old T-Bird 1.0Ghz overclocks to 1.5Ghz, and it beats the original P4 1.8Ghz hands down in pretty much every benchmark.
Currently, AMD64s handily dispatch the very fastest Pentium 4s when they are a full 1000MHz slower than the "fastest" Intel chips.
Personally, I am happy with this shift in paradigm. It means it's going to be much easier explaining to a client why they should go with a mobile chip "2GHz slower than the Pentium 4" when model numbers reflect performance, and not clock rates.
It's been coming for a while now really. Remember when clock rates hit 3.0 GHz? Hell, that was a long time ago now - and in all this time Intel have only succedded in adding a measley 400Mhz to their fastest cores.
They have however dones some good stuff with Hyperthreading, cache-size, better data prediction and branch prediction, memory subsystem speed and data rate.
So - a P43.4GHz EE (Or even a Northwood 3.4) would beat the living snot out of an original Williamette 3.0 overclocked to 3.4.
This is only to be expected given the problematic state of 90nm designs and production. It'll only get worse at 65nm, and a great deal of time and money is going to be spent in the next few years wringing every last IPC (Instruction Per Clock) out of the existing processes.
This is going to mean double and quadruple cores per die, massively increased on-die cache sizes, extreme optimization of execution engines, and huge surges in data throughput.
While the boffins slave away at creating the next paradigm for computing, the rest of the boffins will be preparing the path for ultra-efficient computing. Silcon is the fourth paradigm for computing. The first was mechanical, the second was valves, the third was transistors and Silicon chips are the fourth.
The fifth paradigm for computing will likely be largely self-assembled nano-scale processors incorporating carbon nanotubes, Tri-gate transistors, and other currently nonexistent stuff.
When the fifth paradigm is enabled, computing speeds will rapidly escalate again until they reach "Moore's Law" as plotted from the 60s.
Make no mistake; the current lack of increase in clock speeds is only a temporary hiccup on the exponential-computing-power road. The lessons we learn during this lull will allow the ensuing paradigm shift to make extraordinary gains in computer power.
900mhz duron
OCed to 1ghz.
since the chip's actual speed is prettymuch whatever it's running it. i voted for 1ghz.
coz there is really no such thing as an actual speed. chips are badged, but it means nothing concrete to me (and shouldn't to you either) since it often has little to do with what speed the chip is designed for.
OCed to 1ghz.
since the chip's actual speed is prettymuch whatever it's running it. i voted for 1ghz.
coz there is really no such thing as an actual speed. chips are badged, but it means nothing concrete to me (and shouldn't to you either) since it often has little to do with what speed the chip is designed for.
- KompresZor
- DBB Captain
- Posts: 919
- Joined: Wed Jul 31, 2002 2:01 am
- Location: Clearfield, Pennslyvania
- Mr. Perfect
- DBB Fleet Admiral
- Posts: 2817
- Joined: Tue Apr 18, 2000 2:01 am
- Location: Cape May Court House, New Jersey.
- Contact:
- Aggressor Prime
- DBB Captain
- Posts: 763
- Joined: Wed Feb 05, 2003 3:01 am
- Location: USA
Athlon XP 1800 (T-Bred) @ 10X 200MHz FSB (2.00GHz)
Athlon XP 2100 (T-Bred) @ 1.73GHz
Athlon XP 2100 (Palomino) @ 1.73GHz
Pentium 3 Celeron @ 667MHz (Server)
Pentium 3 Celeron @ 500MHz (Not Active)
Pentium @ 100MHz (Not Active)
Pentium 4 Mobile @ 1.7GHz (Work)
We got the Intels before we started building PCs (Except the Laptop).
Athlon XP 2100 (T-Bred) @ 1.73GHz
Athlon XP 2100 (Palomino) @ 1.73GHz
Pentium 3 Celeron @ 667MHz (Server)
Pentium 3 Celeron @ 500MHz (Not Active)
Pentium @ 100MHz (Not Active)
Pentium 4 Mobile @ 1.7GHz (Work)
We got the Intels before we started building PCs (Except the Laptop).