P4's 533MHz FSB.
- Thread starter _KD_
- Start date
Ummmm, why? It works and it provides the highest bandwidth PC FSB on the market - what the issue?
The Pentium 4's front side bus is essentially a 133MHz one, but it can send four signals per cycle. That's how they work it out from my knowledge (133MHz FSB x 4 signals per cycle = 533MHz FSB). It's just a purposely mis-calculated number like the XBox's Fill Rate.
A 'normal' 533MHz FSB would far outdo the quad-pumped one from what I can tell. The stat is just another number that Intel can trick consumers on and gain more money...
P.S: I'm definately not an Intel hater, infact I prefer them over Mac or Athlon. I just tell things the way I see 'em.
EDIT: When I said Athlon, I meant AMD. lol.
A 'normal' 533MHz FSB would far outdo the quad-pumped one from what I can tell. The stat is just another number that Intel can trick consumers on and gain more money...
P.S: I'm definately not an Intel hater, infact I prefer them over Mac or Athlon. I just tell things the way I see 'em.
EDIT: When I said Athlon, I meant AMD. lol.
Technically speaking there would be no difference in terms of theoretical data throughput. I'm sure in the real world there may be timing issues and the like but in fact the quad pumped bus is actually operating relatively well.
The real problem is that motherboards just cannot facilitate these types of speeds - it not really possible, which os why you still see bus speeds so low (AMD are only looking to 166MHz at the moment). I remember at uni being told of the electromagnetic interferance generated when old motherboard were push to high frequencies (we're talking about 30MHz here!) - I guess this type of thing is still partically the issue, although boards have now now have smaller trace lines and more layers etc, which is why they can get them up to the 100-166MHz ranges now.
arjan de lumens
Veteran
Ahhh. I remember some 486s that used to cause visible interference on nearby TV sets - and how the interference went away when I OCed a 40MHz processor to 50 MHz.
AFAIK, the material that motherboards are made of can easily support data rates in the GHz range, like PC1066 RDRAM, Serial ATA, AMD Hypertransport, even the Rambus Yellowstone bus at 3.2 GHz. It's mostly a matter of controlling crosstalk, skew, SSO noise, reflections, and at >2 GHz skin effect, that sort of stuff. Radiation can be kept down by sandwiching signal traces between ground layers if needed.
AFAIK, the material that motherboards are made of can easily support data rates in the GHz range, like PC1066 RDRAM, Serial ATA, AMD Hypertransport, even the Rambus Yellowstone bus at 3.2 GHz. It's mostly a matter of controlling crosstalk, skew, SSO noise, reflections, and at >2 GHz skin effect, that sort of stuff. Radiation can be kept down by sandwiching signal traces between ground layers if needed.
mat said:
I very much looking forward to see this combo in action...
Granite Bay with it's true dual channel (two 64-bit controllers) and being Intel's second-generation dual DDR implementation should set a new standard for Intel/DDR-performance. According to a post on aceshardware forum the Granite Bay chipset should begin shipping from Intel in the first half this month while boards from Asus, Abit etc. should be launched sometime in November. (I understand that the P4 3.06 should be released in the beginning of November).
The E7205 mobo + P4 3.06 will be quite expensive but at least people won't need to buy Rambus PC1066. For anyone considering buying a P4 system this fall I would recommend that people wait a bit and see the performance of the Granite Bay chipset. I know I will.
Oh, and Kyle from hardocp got some benchmarks on a Asus sample of the granite bay. You might wanna check this out:
http://www.hardforum.com/showthread...readid=506435&perpage=15&pagenumber=1
http://www.hardforum.com/showthread...readid=506435&perpage=15&pagenumber=1
Similar threads
