Teron PX ATX PowerPC motherboard

[downix]

800mhz, Teron PX motherboard, $499

How many HUNDREDS of dollars are you sending Genesi and/or Eyetech in premium?

Actually, please check out http://www.pegasos-usa.com for pricing

The retail price on a Pegasos is *trumpet please* $499.

Funny, same price for a higher-end motherboard seems like a good deal to me.

[downix]

MarkTime wrote:
The board you mentioned is a G3/600, not a G4/800, so you aren't comparing apples to apples.

However, all of this is vapor so we'll have to meet again and compare shipping products (if that ever happens again)

But the pegasos is cheaper than the A-one, at least in the states.

But I am.  The G4/800 listed there runs without an L3 cache, which makes it equal in performance to the G4/600 which, in non-Altivec mode, is nearly equal in performance to the G3-600.  So we're still comparing similarly speced machines.  Only the Pegasos comes with a built-in audio system, a mini-ATX form factor, firewire support, and actually can use the AGP slot.

And I'd hardly call it vapor.  Please, order a board from us and in 10-20 business days it should be on your doorstep.

[downix]

He's just running that up the pole to see if it flies...there are a lot of reasons why the G4 is faster than a G3...and no reason why an 800mhzG4 would be slower than a 600mhzG3.

Providing we're discussing the same G3.  If you're comparing the Motorola MPC750 series of processors (what Apple called the G3) then you'd be correct.

But we're not, we're discussing the IBM 750 series of PowerPC processors.  These are very different beasts from Motorola's lineup.  Better pipeline setup, better heat dissipation, better processes used, all make them a stronger beast than Motorola offers.  The performance jump comes out to about 20% last time I checked... which is about the same boost that you get on a G4 w/o using the Altivec.

Now, once you add-in Altivec, it gets into L3 availibility.  If it's availible, then Altivec gives a serious boost to performance.  if not, you don't get much of one.

I'm not saying that a Pegasos is the best deal on the planet, but we're doing our best and delivering an honest product at an honest price.  Before you dismiss us due to a CPU spec at a pricepoint, how about you look at the rest of the solution.

The Teron board is a full ATX, Pegasos is a smaller Micro-ATX.

Teron doesn't include firewire while the Pegasos does

The teron uses the VIA 686B southbridge, which has known DMA bugs that are well documented.  The Pegasos uses the 8231 southbridge, which is a more modern chipset.  In fact, VIA is listing the VT6x series of southbridges as end-of-life when I checked their website.  While the VT82xx series has new versions still arriving.

then there's the quality of construction.  The Teron board is using every approach to lower the cost of the design.  The Pegasos uses higher-quality connectors in almost all cases.  Means the Pegasos will last longer in many situations.  Also, the positioning of the components on the motherboard can be an issue.  The Teron board has plug interfaces at the end of where the PCI slots run, which means if you get a long card, you can't use the hard drives, for example.  (before you get into it, I have a pro-studio AGP card here that is from the backplane to the front-brackets in my case, so it does actually occur)

Then there's the AGP/PCI-66Mhz slot issue.  If you run a 66Mhz card in that slot, AGP turns into a 33Mhz PCI slot.  Oops!

Of course, the Teron does support that funky new winmodem port.....

[downix]

Herewegoagain wrote:

The Teron board is a full ATX, Pegasos is a smaller Micro-ATX.

Works great if you are a midget with tiny hands to work in that tight cramped up case with hardly any way to get air flow through it.

Give me a full size board and case anyday.

I happen to like the smaller board, as I have a full-sized ATX motherboard in the guise of my Athlon and it's size is causing problemswith poor airflow, causing overheating being the biggest problem.

The teron uses the VIA 686B southbridge, which has known DMA bugs that are well documented. The Pegasos uses the 8231 southbridge, which is a more modern chipset. In fact, VIA is listing the VT6x series of southbridges as end-of-life when I checked their website. While the VT82xx series has new versions still arriving.

Well this has been debated to death, and I believe it has been pointed out in other threads that the 8231 has some issues of it's own....So let's not call the kettle black yet.

I didn't bring up the issues, I brought up lifespan.  The chip series that the 686B was a part of is end-of-life.  This means eventually they'll run out of parts, and a whole new board will have to be engineered.  The 8231 has a drop-in replacement already availible, the 8235, which means supplies of either the original, or a replacement, will not be an issue in the future.

then there's the quality of construction. The Teron board is using every approach to lower the cost of the design. The Pegasos uses higher-quality connectors in almost all cases. Means the Pegasos will last longer in many situations.

 :lol:  :lol:  :lol:   Nice one....nobody can build a quality product but DCE.  Thanks for the laugh.

Did I mention DCE?  I mentioned the components used, not the manufacturer.

Also, the positioning of the components on the motherboard can be an issue. The Teron board has plug interfaces at the end of where the PCI slots run, which means if you get a long card, you can't use the hard drives, for example. (before you get into it, I have a pro-studio AGP card here that is from the backplane to the front-brackets in my case, so it does actually occur)

Come on now.  Is it the use of a long PCI card or AGP card that would be a problem.  Everyone with a full length PCI or AGP card raise your hand?  Anyway, it's not like the IDE connectors stick up higher than the actual PCI connector (which cards don't usually go below anyay).  Ok, move the PCI card to another slot then... how about that.  

And if it's an AGP card?

Of course, the Teron does support that funky new winmodem port.....

Bzzzzzzzzzzzzz.. Wrong!  You are not keeping up with the news Downix.  No AMR slot on the PX boards.  (which btw are not only for modems)

I went by the picture of the CX board, as that's the only picture on the website.  I assumed that the slot positionings would remain the same.  If there is a change in the availibility of the AMR, that is my mistake made on lack of data to the contrary.

And if you noticed, I gave that as a plus, not a minus.  The lack of such a slot has hurt my Athlon here, due to lack of ports for a newer modem.

There are some other segments of the Pegasos as well, such as built-in sound, that lend it to being a nicer board.

But it can come down to personal needs and tastes.  I happen to like smaller systems.  I happen to like the locking-plugs used on the Pegasos.  I happen to enjoy the ease of access to the various parts of the system.  And, frankly, I like how the Pegasos looks vs the Teron boards.  Might sound silly, but asthetics do count.

[downix]

@MarkTime

Interesting points, and ones I'm not entirely deaf to either.

As for the G3 being fine, I actually like the G3 more than the G4.  I see more potential for performance in the G3's architecture.  So I am biased here.  However, G4 boards will be out soon enough, and then we can start comparing these beasts head-to-head-to-head.

[downix]

@ Strobe

There have been a few explinations for this.

The most obvious, however, is the low-overhead for Firewire.  To set up a firewire transfer, you just set the source address and target address and transfer size, and it does the job for you.

PCI, by comparison, is high-overhead.  It uses packet-sized, bus arbitration, and most of all, needs to use the CPU for all of this in most systems.  This means that approximately 20-40% of the clock pulses actually can't be used for the bus transfer.  They are used for the electronic equivelent of bookeeping.

When Firewire says 400 or 800Mbps, they mean you get exactly that for performance, due to the low overhead.  PCI's thuroughput, by comparison, comes in at a grizzly 166Mbps in many northbridge combos.  (do the math here, 33Mhz x 32-bits == 1056 Mbps potential for all devices on the bus combined.  A PCI bus allows up to 4 devices, so 1056/4 == 264 Mbps per PCI slot.  Now, some northbridges do better arbitration, allowing a PCI bus to send more bandwidth to a slot if there are fewer devices then 4, but many don't, so we'll stick to the 264Mbps here, to show a worst-case scenario.  Add in the bus overhead, let's say it only has 20%, so 264Mbps x .8 == 211.2  which is barely faster than the slowest Firewire availible, 200Mbps)