SCOPE XITE-1: new DSP hardware from Sonic Core!

The Sonic Core XITE hardware platform for Scope

Moderators: valis, garyb

User avatar
FrancisHarmany
Posts: 1078
Joined: Sun Jun 02, 2002 4:00 pm
Location: Haarmania

Post by FrancisHarmany »

So can I run both the Scope & Xite on one system ?

Just ADAT from Xite <> SCOPE or something ?
hubird

Post by hubird »

if SFP is able to 'read' different PCI-e slots altogether - just like it does read PCI slots now - this could be possible indeed:-)
Last edited by hubird on Thu Mar 13, 2008 9:29 am, edited 2 times in total.
User avatar
FrancisHarmany
Posts: 1078
Joined: Sun Jun 02, 2002 4:00 pm
Location: Haarmania

Post by FrancisHarmany »

hubird wrote:if SFP is able to 'read' different PCI-e slots altogether - just like it does now - this could be possible indeed:-)
Ok so this way people could have more outs on their system ? Then we can buy up old cards and have more IO.......

So if I route adat-in to some output it wouldnt hurt my pci bandwidth either right ?!
hubird

Post by hubird »

I'm just guessing and hoping... the combination of PCI and PCI-e in one comp could lead to unexpected troubles i.e.
on OSX there's no chance at all of course.
User avatar
alfonso
Posts: 2224
Joined: Sun Mar 25, 2001 4:00 pm
Location: Fregene.
Contact:

Post by alfonso »

FrancisHarmany wrote:So can I run both the Scope & Xite on one system ?

Just ADAT from Xite <> SCOPE or something ?
I'm not directly informed, but I'd say that a single Scope software can only use cards that are interconnected in a single environment....like if you have 2 PCI cards without stdm connection the system is broken.

If you have, instead a PC with it's own Scope cards and a laptop with Xcite-1, then you can connect all the adat, spidif and midi ports to benefit of 2 environments, like it is now with two Scope machines.
ScofieldKid
Posts: 307
Joined: Thu Nov 13, 2003 4:00 pm
Location: Oregon
Contact:

Post by ScofieldKid »

Just catching up on this thread... sorry if I'm repeating.

Just guessing, but I suspect we will have much better overhead in this design.

For example, if we are going from a design that was built with 32-bits and 33MHz at the heart of it.... and now we are in a place where the internals could easily be 64-bit and 200MHz, we should be in a better place. Since there are PCI-X to PCIE bridges, we may have just gotten a leap from old-style PCI to full-on PCI-X.

Again, this is all speculation, but I wouldn't be surprised if they optimized when they were designing this to take advantage of the faster speed capabilities of the new DSP's.

See also: http://www.analog.com/communities/audio ... mWare.html

NOTE: I'm sure someone mentioned it before, but note that PCIE is full-duplex, and serial. This will be a much more capable communication channel than the old-school PCI we have been wrestling with.

[ Ah, I see you guys already ran 100+ MasterVerbs.... As Emily Littella would say, "Nevermind"! ;-) ]
User avatar
spacef
Posts: 3230
Joined: Sun Jun 17, 2001 4:00 pm
Contact:

Post by spacef »

i am trying to put a dual core in my pentium II mother board... tss, something not worlking :x
User avatar
BingoTheClowno
Posts: 1722
Joined: Wed Nov 12, 2003 4:00 pm
Location: Chicago
Contact:

Post by BingoTheClowno »

ScofieldKid wrote:
NOTE: I'm sure someone mentioned it before, but note that PCIE is full-duplex, and serial. This will be a much more capable communication channel than the old-school PCI we have been wrestling with.
Well, the PCI is 32 bit wide working at 133MB/s, I don't think that speed was the issue.
User avatar
spacef
Posts: 3230
Joined: Sun Jun 17, 2001 4:00 pm
Contact:

Post by spacef »

Well, the PCI is 32 bit wide working at 133MB/s, I don't think that speed was the issue.
the old pci is shared with other peripherals and was a major cause of bottleneck (look for "southbridge" and evolution of pci / pci-e or benchmarks). the pci-e improvements has to be combined with improvements in how the motherboard manages other internal and external peripherals. I am not a specialist, but going to pci-e seems that it will deliver more than just the difference between pci and pci-e due to the fact that many peripheral, that used to use the pci bus, don't use it anymore, according to readings about x86 and new motherboard architecture (in which i am not a specialist)...
In this respect, I am SO glad they didn't go firewire, so many of my MAC friends fried them all (FW direct mobo connections) and so easily....
User avatar
BingoTheClowno
Posts: 1722
Joined: Wed Nov 12, 2003 4:00 pm
Location: Chicago
Contact:

Post by BingoTheClowno »

I'm not an expert either, but according to this article, PCIe uses the point to point SWITCH topology.

Image

What I get from this is that you will be able to use only ONE PCIe card at the time. In other words you won't be able to stack multiple cards as was the case with the PCI cards. Some one please confirm this.
So my next question would be, how many DSP chips are in the Xcite box?
voidar
Posts: 1264
Joined: Sun Aug 18, 2002 4:00 pm
Location: Norway

Post by voidar »

No problem, as the xcite-box is expandable.
Shayne White
Posts: 1454
Joined: Tue Dec 11, 2001 4:00 pm
Location: California
Contact:

Post by Shayne White »

BingoTheClowno wrote:I'm not an expert either, but according to this article, PCIe uses the point to point SWITCH topology.

Image

What I get from this is that you will be able to use only ONE PCIe card at the time. In other words you won't be able to stack multiple cards as was the case with the PCI cards. Some one please confirm this.
So my next question would be, how many DSP chips are in the Xcite box?
I'm pretty sure that's not the case -- otherwise, motherboards would only include one PCI-E slot!

PCI-E cards use separate "lanes" (channels) to the chipset, whereas PCI had one connection to the chipset that all the cards shared. That's the big difference between parallel and serial.

As for Scope, that fact that the Xite has an "XTDM" connection means you will be able to use multiple boxes together, I'm thinking.

Shayne
Melodious Synth Radio
http://www.melodious-synth.com

Melodious synth music by Binary Sea
http://www.binary-sea.com
voidar
Posts: 1264
Joined: Sun Aug 18, 2002 4:00 pm
Location: Norway

Post by voidar »

Think of it as PCI-latency and IRQ.

Only one PCI-e device gets attention at a given place in time. The Xcite-1 is a scalable PCI-e device.
User avatar
BingoTheClowno
Posts: 1722
Joined: Wed Nov 12, 2003 4:00 pm
Location: Chicago
Contact:

Post by BingoTheClowno »


Image


In order to transmit PCIe packets, which are composed of multiple bytes, a one-lane link must break down each packet into a series of bytes, and then transmit the bytes in rapid succession. The device on the receiving end must collect all of the bytes and then reassemble them into a complete packet. This disassembly and reassembly happens must happen rapidly enough to where it's transparent to the next layer up in the stack. This means that it requires some processing power on each end of the link. The upside, though, is that because each lane is only one byte wide, very few pins are needed to transmit the data. You might say that this serial transmission scheme is a way of turning processing power into bandwidth; this is in contrast to the old PCI parallel approach, which turns bus width (and hence pin counts) into bandwidth. It so happens that thanks to Moore's Curves, processing power is cheaper than bus width, hence PCIe's tradeoff makes a lot of sense.
http://arstechnica.com/articles/paedia/ ... pcie.ars/5
User avatar
spacef
Posts: 3230
Joined: Sun Jun 17, 2001 4:00 pm
Contact:

Post by spacef »

http://www.tomshardware.com/2007/03/27/ ... page2.html
PCI Express is a bi-directional point-to-point link, meaning that it offers the same bandwidth in both directions, and does not need to share its bandwidth with other devices, as is the case with parallel PCI.
x1 PCI Express mode provides a bandwidth of 256 MB/s upstream and downstream.
(that's 2*256 Mb/s (full duplex))
ScofieldKid
Posts: 307
Joined: Thu Nov 13, 2003 4:00 pm
Location: Oregon
Contact:

Post by ScofieldKid »

At a minimum, I think we are 4-times the bandwidth in the new setup. Which is an awesome thing.

But the point is moot. The testers have already proven that we have a massive improvement. So it's off to the races. I can forsee doing a lot of massive projects on this thing.
User avatar
Shroomz~>
Posts: 5669
Joined: Wed Feb 23, 2005 4:00 pm
Location: The Blue Shadows

Post by Shroomz~> »

This sounds amazing. Thanks for the messe feedback guys!

Btw, I too am a fan of a Z-link version = 16 channels @24/96. :cool:
User avatar
Mary Mungo
Posts: 99
Joined: Sat Feb 03, 2007 9:16 am
Location: Rural Frivolity

Post by Mary Mungo »

Can anybody buy me one, please :D
User avatar
firubbi
Posts: 1156
Joined: Wed May 21, 2003 4:00 pm

Post by firubbi »

today is 20th april....whats the current update.
only 10 days left :)
Shayne White
Posts: 1454
Joined: Tue Dec 11, 2001 4:00 pm
Location: California
Contact:

Post by Shayne White »

You mean 40 days left. SC has until 11:59pm on May 31st before they miss their deadline. I sure hope it IS beginning of May rather than late May. On the other hand, we want them to work out all the bugs before release. :D
Melodious Synth Radio
http://www.melodious-synth.com

Melodious synth music by Binary Sea
http://www.binary-sea.com
Post Reply