That 133 MHz bus is odd...would Apple move to DDR and keep the that same damn 133 MHz bus speed? Since Apple likes to milk mobo designs for so long, it seems like they would want to use a faster bus so it wouldn't be obsolete the day it came out.
But then, this is the new Apple, where performance doesn't matter.
<strong>not for the system bus</strong><hr></blockquote>
Yes for the system bus... they could build leading/trailing edge transfers into the MPX protocol, which might be easier than doubling the clock rate of a bus with all of those lines.
<strong>Undoubtedly I made several errors in the illustration. But I should say that it does no justice to the 3-dimensional world we live in. Remember that the heat sink could easily extend 4 inches up from the board itself since there appears to be nothing blocking it from entering the depths of the case. </strong><hr></blockquote>
True, but what I am saying is that the heatsink I held will never fit in the current G4 case. 4 inches up from the board is not enough, let alone the additionaly length and width. These processors ran HOT, hot enough to discolor the bottom of the sink. IMO this prototype is not ready for primetime.
[quote]Crucify me if I'm wrong but, I know color alterations of images, and this has tell-tale signs of fakery.<hr></blockquote>
Ain't gonna crucify you, Maskr, since you probably work with high end imagery and not crappy squashed stuff like us web guys.
JPEG compression works by taking chunks of pixels and reducing the palette within those chunks. In a large field of noisy pixels, it's usually relatively benign. It's on the edges between two colors where you get very distinct artifacts as the algorithm tries to find a happy medium.
See any anomalous pixels in my example below?
(I originally saved it with agressive JPG compression-10 out of 100 in Photoshop Save For Web- but resaved the example as a GIF to avoid recompressing and creating more, unnecessary artifacts).
Even if the color is fake, the hardware obviously isn't, nobody would spend so much time to draw all the lines, wires, place all the chips, connectors etc, and have it stil look that real.
Even if the board could initially have been green, that hardware is real, and it's totally different to any Mac mainboard out there.
I'd have to give BIG kudos to the man who faked this picture. Frankly, I don't think it's fake.
(Which doesn't mean I think its the thing that's going to ship soon)
AFAIK, the 533MHz bus is clocked by two 133MHz clock signals 180° out of phase, with data transmitted on the rising and falling edges of both clock pulses.
look those fake shadows from the ram and PCI slots !
the lightning from the table is not the same from the board.. the board is much brighter. just compare those real Mobo pics and you will see that this is a fake.. totaly fake
and those tin can capacitors are totally out of place.. the power regulators too..
every athlon mobo that has DDR-266 has 133mhz double pumped.. it means the signal is double used.. both up and down transients are used as a bit.. and the bus is 32 bits...
the ddr-266 bus is a 64-bit bus, and those nVidia chipsets can pair 2 slots and make a virtual 128-bit DDR slot...
on the p4, the RDRAM is only 16bits... so 1066mhz x 16bits equals a 266mhz x 64 bits. the bus is 533 mhz, but I don't know details from this bus.... sorry
what I am trying to tell is that intel is pushing high numbers, and not performance. the boards using DDR-266, at 64-bits is equal to a 1066mhz, at 16 bits !.. and the nvidia board that can do 128-bits x DDR-266 as double the memory performance, with lower numbers... intel fools everyone with fake numbers... and it hurts apple too..
Re. the iWalk. They fooled noone. From it came up on spymac until the first doubt it took about 30 seconds. And until it was declared fake by anyone it took less than 15 minutes. Go find the old thread and discover that the "production time/total time of believing among all" ratio must have been >1. Well done but noone can fool the AI collective <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />
Comments
But then, this is the new Apple, where performance doesn't matter.
<strong>The bus speed isn't odd. If the board uses DDR 266, then it must have a 133 Mhz bus.</strong><hr></blockquote>
not for the system bus
That's 4x133 bus right there, maybe we'll get that too, in ten years or so.
BTW, the P4 beats the Athlon XP 2100+ everywhere, now there's another company palying catch-up, w're not alone
G-News
<strong>not for the system bus</strong><hr></blockquote>
Yes for the system bus... they could build leading/trailing edge transfers into the MPX protocol, which might be easier than doubling the clock rate of a bus with all of those lines.
<strong>Undoubtedly I made several errors in the illustration. But I should say that it does no justice to the 3-dimensional world we live in. Remember that the heat sink could easily extend 4 inches up from the board itself since there appears to be nothing blocking it from entering the depths of the case. </strong><hr></blockquote>
True, but what I am saying is that the heatsink I held will never fit in the current G4 case. 4 inches up from the board is not enough, let alone the additionaly length and width. These processors ran HOT, hot enough to discolor the bottom of the sink. IMO this prototype is not ready for primetime.
Here's why...
Crucify me if I'm wrong but, I know color alterations of images, and this has tell-tale signs of fakery.
Shrug. PS In the image I meant to say "six PCI slots."
MSKR
[ 05-06-2002: Message edited by: Masker ]
[ 05-06-2002: Message edited by: Masker ]</p>
Ain't gonna crucify you, Maskr, since you probably work with high end imagery and not crappy squashed stuff like us web guys.
JPEG compression works by taking chunks of pixels and reducing the palette within those chunks. In a large field of noisy pixels, it's usually relatively benign. It's on the edges between two colors where you get very distinct artifacts as the algorithm tries to find a happy medium.
See any anomalous pixels in my example below?
(I originally saved it with agressive JPG compression-10 out of 100 in Photoshop Save For Web- but resaved the example as a GIF to avoid recompressing and creating more, unnecessary artifacts).
Good observation, though.
Even if the board could initially have been green, that hardware is real, and it's totally different to any Mac mainboard out there.
I'd have to give BIG kudos to the man who faked this picture. Frankly, I don't think it's fake.
(Which doesn't mean I think its the thing that's going to ship soon)
G-news
G_News
Also Intel today just came out with The new P4 at 2.53GHz and a 533MHz bus speed, 1066MHZ RDRAM speeds.
That's 4x133 bus right there, maybe we'll get that too, in ten years or so.
<hr></blockquote>
It's not really a 533 Mhz bus. It's actually a 133 Mhz bus that was "quad-pumped". This means that data is read and/or written 4x per cycle.
Is this also how AGP 4x works? How about AGP 8x?
the componets placement has no order !
those ide slots are totally stupid !
the power connector is out of place
you can see dropshadows of every item !
this is a fake.... a fake I tell you !
look those fake shadows from the ram and PCI slots !
the lightning from the table is not the same from the board.. the board is much brighter. just compare those real Mobo pics and you will see that this is a fake.. totaly fake
and those tin can capacitors are totally out of place.. the power regulators too..
please, forget about this hoax.. it's a fake.
every athlon mobo that has DDR-266 has 133mhz double pumped.. it means the signal is double used.. both up and down transients are used as a bit.. and the bus is 32 bits...
the ddr-266 bus is a 64-bit bus, and those nVidia chipsets can pair 2 slots and make a virtual 128-bit DDR slot...
on the p4, the RDRAM is only 16bits... so 1066mhz x 16bits equals a 266mhz x 64 bits. the bus is 533 mhz, but I don't know details from this bus.... sorry
what I am trying to tell is that intel is pushing high numbers, and not performance. the boards using DDR-266, at 64-bits is equal to a 1066mhz, at 16 bits !.. and the nvidia board that can do 128-bits x DDR-266 as double the memory performance, with lower numbers... intel fools everyone with fake numbers... and it hurts apple too..