P4 = WORLDS FASTEST CPU!!!

[AmdMELTDOWN]

<A HREF="http://www.tomshardware.com/cpu/02q1/020128/index.html" target="_new">P4 vs AXP</A>

"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"

 

[AMD_Man]

CORRECTION: P4= WORLD'S FASTEST PC PROCESSOR!

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor

 

[bdaley]

Yeah, the P4 OCed to 3 gig barely beats the XP OCed to 2.2 gig. What an efficient design that P4 is!

I guess I'd better ditch my XP rig now and get a P4, huh meltdown?

"There's no such thing as gravity, the Earth just sucks"

 

[tnadrev]

it does look as such, there is no denying it... the P4 (OC'ed) is quite a bit faster than the AXP (OC'ed) when the 0.13ì process TBred comes out i would like to see a new version of this article... especially if you can still unlock the AMD CPUs multipliers...

However, as a student, the AXP still wins in the most important benchmark for me... Priceerformance...

Engineering is the science of making life simple, by making it more complicated.

 

[AMD_Flinkster]

the athlon xp design is way older than the northwood, and is still competition. why doesnt northwood pick on someone his own size and try TBred for a change (as soon as its released in Cebit 2002 in march ) ?

 

[Makaveli]

About time Intel made a decent CPU. After AMD holding this title for the last year, Intel finally decides to get off there asses. There is still no doubt the Original P4 is a POS performance Wise!

Now if only they could have Amd's price/performance ratio I would consider buying one!

Also u did notice on the same Memory the P4 did lose alot of benchmarks. RDRAM is the only thing helping the P4 to stay in the lead.

 

[killall]

its not at 2.2ghz its at 1.87ghz

if in doubt blame microsoft...

 

[Guest]

"Yeah, the P4 OCed to 3 gig barely beats the XP OCed to 2.2 gig. What an efficient design that P4 is!

I guess I'd better ditch my XP rig now and get a P4, huh meltdown?"

I must disagree, the memory is the bottleneck, that is why RDRAM beat out DDR-SDRAM. The board also did not have the 3:4 ratio, which would have greatly overclocked the RAM.

"Heat is a form of energy, thus the Law of Conservation of Energy holds!" - James Prescott Joule

 

[Copenhagen]

I would really like to see Abit allowing a higher Vcore than 1.65V on the RDRAM based TH7-II RAID. The old BIOS allowed 2.20V, so the board is definitely capable of giving a lot more that 1.65V. But for some reason Abit has this time chosen to be political correct and follow Intel's specifications. I hope they will change this soon.

Other than that I think this is bad news for Intel. They desperately need a faster FSB in order to withstand the attach from the 0.13 micron throughbred core.

If the old 0.18 micron Palomino can do 1866MHz watercooled and 1933MHz is all it takes to reach the performance of a P4/3000MHz on DDR RAM (845D), then a watercooled Throughbred will simply crush the P4/3000MHz.

<i>/Copenhagen</i>

<b><i>Seagate Barracuda IV.
Bad performance in RAID setups!
</i></b>

 

[bdaley]

1912.23, which makes the P4@3000 look even worse.

"There's no such thing as gravity, the Earth just sucks"

 

[eden]

Finally someone who understands.
Sure the P4 is officially the most impressive OCing CPU ever, and fastest, but certainly not impressive performance-wise. I would never buy one for the sake of OCing it and get near-results of AXPs.
Simply put Intel is not doing what we want. They once were awesome people, but now once AMD is leading, they put their hands in their asses and ignore that... How deceiving...
Face it people, even if you believe the P4 is better, fact is it's a mere improvement in performance when moved to 512K, and not too far will AMD reclaim an ever more impressive power. Let us wait and see but for the love of god wake up and smell the coffee, P4 IS NOT impressive performance-wise!! If it took 1133MHZ to be ON PAR with the AXP 2300, then that's just sad to me and many peeps I know...

--
The other day I heard an explosion from the other side of town.... It was a 486 booting up...

 

[bront]

In comparison, today the average system is equipped with a 900 MHz processor as well as an MX2000 graphics card.

Forget the CPU war, what is an MX2000? Maybe the average computer user has a beter computer than we thing

RAM Disk is not an instalation step.

 

[Guest]

yes, and when comparing price we should remember that you would either have to get that "specially selected Pentium 4 processor that Intel itself was unable to provide for us."
Or if you just invest "at least 200 dollars" in a water cooling system, "you can use almost any P4 Northwood at 2.2 GHz or higher."
And still they had to "put your heat exchanger outside a window! "
Sure, a cpu running at >3Ghz is impressive. But it isn't interesting to ordinary users until it becomes a little more available to us, right?


old yoga proverb:
"when the dog farts,
it's time to fire up your incense."

 

[IIB]

in order to pump the bus above 533mhz Intel will need to reduce the limitations on higher bus speeds - maintaining a unified signal with slow PCI buses is an issue. And high latency low clock speed memory access – memory need to get faster.

AMD deals with the PCI issue with a new form of bus – HyperTransport, Now accepted by most of the industry, will be the new standard. Some say Intel is to license it from AMD.
And the integrated memory controller cuts Latency issues enabling higher bus speeds while maintaining a unified signal.
Some say the Hammer will have 800 MHz bus. The white paper states that the integrated memory controller will first work with duel channel DDR333 (5.3Gb of bandwidth) so bus numbers around 800 MHz are needed.

Though I'm sure Intel got some of her own tricks up the sleeves.
AMD is also limited in bus speed – by packaging technology which Intel seems to get around with their BBUL packaging technology.

 

[killall]

the XP 2300+ is at 1866 thats what was tested, the p4@3000 with DDR did look quite poor performance for clock speed, but who cares about the clock speed... other than most people of course

if in doubt blame microsoft...

 

[tnadrev]

agreed, i just wish i could wait until march for my new system... reaching 3000+ (PR) on a TBred might be feasible with air cooling...

Engineering is the science of making life simple, by making it more complicated.

 

[TheAntipop]

it took 1133MHZ to be ON PAR with the AXP 2300

1133Mhz?? That's more speed than my processor has in it total

Hard work often pays off in time, but laziness always pays off now.

 

[texas_techie]

Ah yes, bask in the glory of having a 1,100 mhz lead. I look forward to December. I plan on learning photshop, so that when december comes and with it the Hammer. I can post a picture of the Intel Doll's head being crushed by a giant hammer. Maybe I'll throw in the intel doll crapping its pants just as the hammer pierces that gay space suit.

 

[bront]

Well, the good news is, that Intel finaly found a test that justifies it's use of RDRam. However, they tested the DDR ram using the worst performing DDR ram chipset for the P4, and a slower MB for the AMD Athlon. I don't know why they didn't use the Asus board for the Athlon, and the Via P4x266 chipset for the P4.

With the speed increasing, will be interesting to see what happens to the PCI bus as well. There are some issues already developing, and perhaps they will eventualy come out with some new expansion slot standard that supports faster than 66 Mhz. Memory speed will become more important too, and the battle between high speed DDR ram and RD ram will definately be interesting.

BTW, can anyone develope an RD ram system for the Athlon? or is there some reason they couldn't?

RAM Disk is not an instalation step.

 

[charliec2uk]

Its a clean-room suit to prevent contimination of the silicon.

Never spit into the wind. Never eat yellow snow.

 

[Civilized]

yea you can always leave it to convinced intel people such as my roommate who is a intel sales rep to totally believe no matter what intel is the best eventhough he knows jack about computers hehe.......time will tell the difference....
and besides anything over one ghz is hardly noticable except at very high resolutions in gaming and not noticable at all to the average user...

 

[Olfin_Bedwere]

RDRAM (Dual Channel) offers more bandwidth than the AXP can use. The high latency would actually reduce the performance of the AXP.

 

[Matisaro]

Its a clean-room suit to prevent contimination of the silicon.

Although the suits in their commercials are not like most cleanroom suits.(worked in a cleanroom untill my fab was shut down, will work in one again). They dont look like virus proof suits from outbreak, but each company may use different types, so I am not positve.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!

 

[Guest]

Thing about Intel is that their chipsets are still way better than AMD's..(VIA...bleh)

 

[Matisaro]

Thing about Intel is that their chipsets are still way better than AMD's..(VIA...bleh)

Thats in no way true, intels new chipset has the same type of bug that vias has(the pci bandwith bug)(I believe its the i850 , but it may be the I845 I forget ATM.)

If you dont want via, thats fine, there are plenty of athlon chipsets which are via free, and perfectly stable, nforce, ali magik, and the amd760. Intel chipsets in the past may have been more stable, but they are just as prone to issues now as any other chipset, issues which are not important if one does a little research before buying their motherboards.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!