StarDestroyer.Net BBS

http://www.hexus.net/review.php?review=443


And thats Aircooled and default Voltage... And its not alone, he mentions several reports of people getting that high easy and stable and I know somone who managed as well

For those in the computer know, Prehaps Intel has graced us with another Cerlon like the 366?

Muwhaha, my dual proccessor system under fundraising support, too Pentium IVs too be prcise shall be slower or faster then this?

An I correct in reading that as 3 Gigahertz.... Fuck me.

According to the add-up he had it at 3470 with Watercooling


Oh and fokes keep in mind, that 3GHTZ?

Its off a Cerelron 2 Gigahtz
Thats right over an 1.5 Gigs of Overclocking room in them

Consider the Fact some people still don't even have 1GHTZ + Computers and this thing can OC that much...

Pretty fast.


Who knows how fast computers will get by october of next year?

it has one critical flaw (other than the fact it's a celeron): it's an intel chip, intel sucks.

it's an intel chip, intel sucks.

Brillant Deducative reasoning as usual Hyprion

Tell me did you know there is a 2.6 GHTZ Celeron Avaible(Not publicly but I've seen it) that can be run at full-speed WITHOUT a Heatsink?(120*F and as far as I could tell it tops out heat wise there as it was running a typical Quake III torture demo)

Tell me what pray tell happens to anything over an AMD 500 Althon if the Heatsink is removed?

Your brillant reasoning for Intel being bad is... what?

Price?


Guess BMW's must suck too eh?

Of course if you have some valid critsims shoot, I'm ready

Mr Bean wrote: Tell me did you know there is a 2.6 GHTZ Celeron Avaible(Not publicly but I've seen it) that can be run at full-speed WITHOUT a Heatsink

Gotta give that one a big giant WTF.

You have an article handy?

Gotta give that one a big giant WTF.

You have an article handy?

No saddiy its one of those things they don't just toss around(Like the 5.5 GHTZ Chip Demosrated at Comdex)

I saw this one at a small private demostration it seems to be a brand new method of CPU Packaging. I'm not sure of the exact details however it was slighty overlylarge(tall, same demensions of course) maybe .5-1Cm taller than normal, No idea if they will ever bring it to makert but its designed as a new method for mobile apps

This was a small private demosration a few months back, week or two after the last show that I got to sit in on purley by coincidence

if you got that data from tom's hardware, don't beleive it. i ran the same tests they did (with help and donations from local stores who wanted to prove it), the intels slag very quickly. to recreate the exact effects seen in that test, the intel chips are underclocked by nearly 50%, and the AMD overclocked at least 100-200%. even then the temp readings were NOT 600+F they slagged but didn't get that hot.

with proper setup (no OC), the AMD chip will overheat and the bios will shut down the comp, if not, the framerates for the testgame drop by like 75%. on the intel it locked up quickly, it dropped out much faster than the AMD chip did.

2.9GHZ on an intel, woofuckinghoo, a 1.83GHZ AMD athlonXP 2200+ would still outpace the fucker under load.

key differances between intel and AMD

pipeline:
intel: 20 stage
AMD: 8 stage

commandset:
intel: CISC x86
AMD: RISC w/ x86 CISC command interpreter



those are the main reasons AMD is better, the pipeline being shorter and the fact it's a RISC core make it faster than the intel even at lower clockrates. it also gives the AMD systems more raw processing power. i know from experience that i can do stuff on my ancient AMD athlon-4 1100 system that you couldn't even hope to do on an intel P4 2.4GHZ without lugging it down to the point of being unusable. i'll post screenshots of this later on once i bring up my website.

a note on pipelines: if there's an error at any of the stages of the pipeline, the error is propogated down the line, causing more errors, witha 20 stage pipeline the chances of a catastrophic crash are much higher, whereas AMD's 8 stage propogates the error out much faster and before it can do too much damage.

if you want more data on the architecture and/or commandsets, or other data, just ask, i've got a stack of CS books in front of me.

if you got that data from tom's hardware, don't beleive it. i ran the same tests they did (with help and donations from local stores who wanted to prove it), the intels slag very quickly. to recreate the exact effects seen in that test, the intel chips are underclocked by nearly 50%, and the AMD overclocked at least 100-200%. even then the temp readings were NOT 600+F they slagged but didn't get that hot.

Funny, As the Murder of four AMD chips who's fans died and took the board with em I tend to belive Tom, To be fair however I don't own any p4s or Celeron p4s, just my two Celeron 550's as a Linux Server, besides that I've only bough AMD and I've killed alot of em

First one died was an 850 running at 933(Board screamed when it hit 190F shut down, CPU died board survied)
The other four were all over 1GHTZ and took the board with them(1.1, 1.4, and a pair of 2000XP ones)

with proper setup (no OC), the AMD chip will overheat and the bios will shut down the comp, if not, the framerates for the testgame drop by like 75%. on the intel it locked up quickly, it dropped out much faster than the AMD chip did.

Personal experance speaks diffrently here, however it should be quested what boards where used? Two of mine where Asus, one Soyo and one Abit

those are the main reasons AMD is better, the pipeline being shorter and the fact it's a RISC core make it faster than the intel even at lower clockrates. it also gives the AMD systems more raw processing power. i know from experience that i can do stuff on my ancient AMD athlon-4 1100 system that you couldn't even hope to do on an intel P4 2.4GHZ without lugging it down to the point of being unusable. i'll post screenshots of this later on once i bring up my website.

a note on pipelines: if there's an error at any of the stages of the pipeline, the error is propogated down the line, causing more errors, witha 20 stage pipeline the chances of a catastrophic crash are much higher, whereas AMD's 8 stage propogates the error out much faster and before it can do too much damage.

if you want more data on the architecture and/or commandsets, or other data, just ask, i've got a stack of CS books in front of me.

Muuch Better Hyprion, I never said I was aurging with you, Mearly asking you to try a bit harder

(Like I said I buy AMD myself and will contiue to do so at least for the next year)

When I built this one I put an AMD 1700+ XP in it, which I think they clock it at 1400 something. I figure on getting one that's at least 2.5 soon.

first thing on keeping AMD chips happy: don't OC'em they die fast when OC'd, it's better to just save up and buy the next highest.

btw, i've only owned 4 intel based systems, 1 i386, 1 p-166, a p-120 laptop (still have it but it's dying after many long years of faithful service), and a p2-400 *shudder*. i have built a good number of modern intel based systems, i was never able to get the performance out of'em, even when OC'd that i could get out of even the old K6-2 AMDs. (btw, i own 17 computers, only 2 are intel based, the p2-400 and the laptop, the rest are AMD.)

boards used to date: 8 ASUS skt7's 7 failed from MFGR defects. one Epox skt7 still running in #4 system (aka, "shitpile"), one ASUS k7m partial failure after 18 months due to MFGR defect (solder bridge in ramslot caused a progressive I/O failure), one ASUS K7M still running #2 system, one AMD mainboard running so far for 2 years and counting with no failures or any sort of problem (gateway machine). one ECS/Elitegroup K7S5A MFGR defect in I/O controller computer was never brought online litigations with the company are pending on this one.

and of course a whole host of others that i'm not bothering to list.

if you want more specific data on any of those comps or failures, just ask.

No need, like yesterday's forcast I would quickly forget them in anycase

something of interest regarding AMD vs. Intel:

next gen chips:

Intel Itanium: 64-bit CISC core, incompatible with anything below 64-bit, new software is needed to run on this.

AMD Clawhammer/Sledgehammer (MP): 64-bit RISC core w/ x86 CISC command interpreter, compatible down to 8 bit. all existing software will work on this.



i wonder which one i'm going to use...

Mr Bean wrote:No need, like yesterday's forcast I would quickly forget them in anycase

lol, i know the feeling...

it took a hell of a long time before i trusted ASUS again after that week of bad boards, including 4 which were professionally set up. (all were verifyed by ASUS corporate as having been properly set up and the failures were verifyed to be MFGR caused) the only reason i went back was because i got a package deal on the K7M with an athlon 500MHZ when the p/s in the (then) main system blew up taking nearly all the hardware with it. (yes, smoke and fire, and the whole 9 yards) that board did extremely well with the bad ramslot for that long that i figured they were worth giving another chance to, been using ASUS ever since with no problems.

Hyperion wrote:Intel Itanium: 64-bit CISC core, incompatible with anything below 64-bit, new software is needed to run on this.

Doesn't the Itanium provide hardware assistance for (software-based) IA32 emution? It'll need a new OS but it should be able to run current IA32 software.

AMD Clawhammer/Sledgehammer (MP): 64-bit RISC core w/ x86 CISC command interpreter, compatible down to 8 bit. all existing software will work on this.

With integral Fritz (TCPA/Palladium) components, from what I hear....

Hyperion wrote:something of interest regarding AMD vs. Intel:

next gen chips:

Intel Itanium: 64-bit CISC core, incompatible with anything below 64-bit, new software is needed to run on this.

AMD Clawhammer/Sledgehammer (MP): 64-bit RISC core w/ x86 CISC command interpreter, compatible down to 8 bit. all existing software will work on this.



i wonder which one i'm going to use...

Neither, yet. The market for 64-bit CPUs kinda evaporated when the economic boom did. And Intel's Itanium has already had an embarassingly rocky start. (The first versions were completely humilated in benchmarks by 32-bit PIIIs with similiar clock speeds.)

Hyperion wrote:if you got that data from tom's hardware, don't beleive it. i ran the same tests they did (with help and donations from local stores who wanted to prove it), the intels slag very quickly. to recreate the exact effects seen in that test, the intel chips are underclocked by nearly 50%, and the AMD overclocked at least 100-200%. even then the temp readings were NOT 600+F they slagged but didn't get that hot.

with proper setup (no OC), the AMD chip will overheat and the bios will shut down the comp, if not, the framerates for the testgame drop by like 75%. on the intel it locked up quickly, it dropped out much faster than the AMD chip did.

2.9GHZ on an intel, woofuckinghoo, a 1.83GHZ AMD athlonXP 2200+ would still outpace the fucker under load.

key differances between intel and AMD

pipeline:
intel: 20 stage
AMD: 8 stage

commandset:
intel: CISC x86
AMD: RISC w/ x86 CISC command interpreter



those are the main reasons AMD is better, the pipeline being shorter and the fact it's a RISC core make it faster than the intel even at lower clockrates. it also gives the AMD systems more raw processing power. i know from experience that i can do stuff on my ancient AMD athlon-4 1100 system that you couldn't even hope to do on an intel P4 2.4GHZ without lugging it down to the point of being unusable. i'll post screenshots of this later on once i bring up my website.

a note on pipelines: if there's an error at any of the stages of the pipeline, the error is propogated down the line, causing more errors, witha 20 stage pipeline the chances of a catastrophic crash are much higher, whereas AMD's 8 stage propogates the error out much faster and before it can do too much damage.

if you want more data on the architecture and/or commandsets, or other data, just ask, i've got a stack of CS books in front of me.

Actually, IIRC, the Athlon uses the same x86 CISC instruction set. Where it gets it's performance boost is, among other things, the shorter pipeline and the various tricks it employs to reduce it's CPI (cycles per instruction.)

it is a RISC core, that's well documented. the architecture is NOT x86, so it requires a command interpreter aspect anyway, they just took it a bit farther.

and yes, the *greatly* reduced pipeline is probably the biggest help.

IA64 will emulate IA32 code, though not with the best efficiency. It also requires a very, very efficient compiler to get the best out of it. Otherwise it's EPIC architecture will be hamstrung. I've heard varying reports of excellent performance and horrid performance.

Most processors today don't fit exactly under RISC or CISC - they're more hybrids than anything.

Yes, the K7 and K8 interpret the x86/IA32 command set into their own native instructions. The PPC970 and POWER4 do the same thing with the PPC instruction set.

64-bit processors have their use, but not in the mainstream.

Both the next-generation Intel and AMD CPUs will have Paladium support built-in. At least in the case of the latter processor it can be disabled.

Here's my experiences:

We used to have a bunch of P166C machines. Very stable (they used Intel motherboards). Same with a group of P3/800EBs we have right now (also using Intel motherboards). We've also had a P3/550 and a P2/450 that were very unstale - the former used a PC Chips motherboard and the latter was Compaq Presario. The laptops in the house are also Intel-based and stable.

I had a K7/700. Stable, fast. I also have Palamino 1.53GHz - also stable. We had a K62/450, quite unstable.

k6-2 450's had some issues on certain boards... mainly Epox and PC chips, ironically that epox board's never given me an ounce on trouble.


i've had exceedingly bad luck with the modern intels.