Topic: PCI-E 7 slot expansion board (Read 7750 times)
price = highly unreasonable..
http://www.thedebugstore.com/acatalog/SKU-040-02-PCIe-Splitter.html
Price:£121.00 / €148.83 (Excluding VAT at 20%)
I talkedto them about gpus on that thing and they say itll work 
Quote
Hello,
Thank you for your interest in our product.
This splitter has 3 x1PCI express adapters for plugging add-in boards. If your GPU has x16 PCI express interface you need to use additional passive adapter for converting x16 PCI express to x1 PCI express.
If you have any questions please don’t hesitate to contact us.
Best regards.
Support Team
AMFELTEC Corp.
From: h [mailto:.com]
Sent: Thursday, August 11, 2011 8:18 AM
To: [email protected]
Subject: GPGPU / CUDA / Stream? SKU-040-01
Hi,
i was wondering if this PCIe splitter (SKU-040-01) can handle some GPUs for folding and Bitcoin mining?
regards
Thank you for your interest in our product.
This splitter has 3 x1PCI express adapters for plugging add-in boards. If your GPU has x16 PCI express interface you need to use additional passive adapter for converting x16 PCI express to x1 PCI express.
If you have any questions please don’t hesitate to contact us.
Best regards.
Support Team
AMFELTEC Corp.
From: h [mailto:.com]
Sent: Thursday, August 11, 2011 8:18 AM
To: [email protected]
Subject: GPGPU / CUDA / Stream? SKU-040-01
Hi,
i was wondering if this PCIe splitter (SKU-040-01) can handle some GPUs for folding and Bitcoin mining?
regards
My understanding is that 2D acceleration was made possible with support from AMD. My point was that adding OpenCL to open source drivers would require AMD assistance (assistance they have so far been unwilling to provide) not that it would be impossible.
Now that I can agree with.The work required to add OpenCL support is pretty massive, and AFAIK there's pretty much 0 done on a open source OpenCL -> VLIWx shader asm compiler, several other large parts of the puzzle are also completely missing (runtime, runtime/opengl/driver integration/...).
But saying that the open source driver can't use shaders is wrong.
Now, for something like a dedicated miner for 5/6xxx you'd "only" have to handcraft a kernel for each arch in VLIWx ASM (fully documented in public AMD docs btw) and hack the radeon driver so it allocates input/output regions and loads/runs your shader program on command. Iirc there's still a open 200BTC bounty for a miner working like that. but it's a *lot* of work for very little benefit.
I think I understand. So while it wouldn't be possible to implement OpenCL (which is rather expansive API) hack together low level support just for mining and implement that in a fork of OpenSource drivers? The code wouldn't be openCL compatible but would function at a lower level.
Has anyone done a proof of concept? Something simpler than mining maybe something as simple as loading an value, increment it in a loop and then read the output?
I never thought about doing it that way. Hopefully someday AMD provides support for OpenCL in OpenSource drivers as that is a pretty ugly "hack".
My understanding is that 2D acceleration was made possible with support from AMD. My point was that adding OpenCL to open source drivers would require AMD assistance (assistance they have so far been unwilling to provide) not that it would be impossible.
Now that I can agree with.The work required to add OpenCL support is pretty massive, and AFAIK there's pretty much 0 done on a open source OpenCL -> VLIWx shader asm compiler, several other large parts of the puzzle are also completely missing (runtime, runtime/opengl/driver integration/...).
But saying that the open source driver can't use shaders is wrong.
Now, for something like a dedicated miner for 5/6xxx you'd "only" have to handcraft a kernel for each arch in VLIWx ASM (fully documented in public AMD docs btw) and hack the radeon driver so it allocates input/output regions and loads/runs your shader program on command. Iirc there's still a open 200BTC bounty for a miner working like that. but it's a *lot* of work for very little benefit.
Without AMD support it is unlikely any Open Source driver that takes advantage of shaders will ever exist.
Err, what?Take a look at the Xorg radeon driver and how it does several parts of 2D acceleration on 5xxx and 69xx
My understanding is that 2D acceleration was made possible with support from AMD. My point was that adding OpenCL to open source drivers would require AMD assistance (assistance they have so far been unwilling to provide) not that it would be impossible.
Without AMD support it is unlikely any Open Source driver that takes advantage of shaders will ever exist.
Err, what?Take a look at the Xorg radeon driver and how it does several parts of 2D acceleration on 5xxx and 69xx
Yeah, I pointed that out above but D&T is under the impression that Intel and open source ATI drivers don't support OpenCL when in fact they do ( somewhat ).
If someone could let us use the damn open source drivers that would be magical. No need to have X server running anymore !
Artz didn't say open sources drivers support OpenCL.
You claim is incorrect.
No Open Source driver supports OpenCL despite a lot of asking by the community.
Intel has no graphical driver which support OpenCL (they only support OpenCL on CPU).
Quote
If someone could let us use the damn open source drivers
Who is stopping you? Without AMD support it is unlikely any Open Source driver that takes advantage of shaders will ever exist.
Err, what?Take a look at the Xorg radeon driver and how it does several parts of 2D acceleration on 5xxx and 69xx
Yeah, I pointed that out above but D&T is under the impression that Intel and open source ATI drivers don't support OpenCL when in fact they do ( somewhat ).
If someone could let us use the damn open source drivers that would be magical. No need to have X server running anymore !
Without AMD support it is unlikely any Open Source driver that takes advantage of shaders will ever exist.
Err, what?Take a look at the Xorg radeon driver and how it does several parts of 2D acceleration on 5xxx and 69xx
If AMD ever changed their drivers to allow 16 (or better unlimited) GPU per system well then it might make more sense.
How much worse are the open source ATI drivers for linux than the proprietary ones for mining? It seems that it should be possible to modify them to remove a limit, but if they perform badly then there may be no point.Yeah open source drivers for mining would really be HEAVEN for me. Screw AMD and their crappy drivers and bugs.
Somebody with the skill ( like ArtForz ) needs to try and sort this out for us.
Amd doesn't provide sufficient data on internal workings of chip to write OpenSource OpenCL drivers. Many have asked AMD claims they can't for industry trade secret and competitive advantage reasons.
Without AMD support it is unlikely any Open Source driver that takes advantage of shaders will ever exist.
If AMD ever changed their drivers to allow 16 (or better unlimited) GPU per system well then it might make more sense.
How much worse are the open source ATI drivers for linux than the proprietary ones for mining? It seems that it should be possible to modify them to remove a limit, but if they perform badly then there may be no point.There are no open sources drivers which are OpenCL capable.
If AMD ever changed their drivers to allow 16 (or better unlimited) GPU per system well then it might make more sense.
How much worse are the open source ATI drivers for linux than the proprietary ones for mining? It seems that it should be possible to modify them to remove a limit, but if they perform badly then there may be no point.Yeah open source drivers for mining would really be HEAVEN for me. Screw AMD and their crappy drivers and bugs.
Somebody with the skill ( like ArtForz ) needs to try and sort this out for us.
Look here to start :
http://dri.freedesktop.org/wiki/GalliumCompute
http://www.x.org/wiki/RadeonFeature
If AMD ever changed their drivers to allow 16 (or better unlimited) GPU per system well then it might make more sense.
How much worse are the open source ATI drivers for linux than the proprietary ones for mining? It seems that it should be possible to modify them to remove a limit, but if they perform badly then there may be no point.There are no open sources drivers which are OpenCL capable.
If AMD ever changed their drivers to allow 16 (or better unlimited) GPU per system well then it might make more sense.
How much worse are the open source ATI drivers for linux than the proprietary ones for mining? It seems that it should be possible to modify them to remove a limit, but if they perform badly then there may be no point. I really don't know what to do guys, LOL.
Should I buy this ?
I think this is quite good where prices of 2*5870 << price of 5970 ?
Should I buy this ?
I think this is quite good where prices of 2*5870 << price of 5970 ?
Are you sure.
Price out:
"normal" rig w/ 3x5870 (no extenders necessary)
"normal" rig w/ 4x5870 (using extenders & entry level board)
"normal" rig w/ 6x5870 (using extenders & 6x PCIe board)
"extender rig" w/ 8x5870 ( low end board & 1x extender board)
then calculate entire system cost per GH of each system. I think you will find that it isn't as good as you think.
The problem is AMD drivers. Currently it is easy to get 4 graphics cards on 1 board. Getting 6 is more complicated (and expensive) but certainly possible.
So while this is a 7 slot design you aren't gaining 7 GPU. You are gaining 2 to 4 more GPU. $200 for 2 to 4 more slots = $50 to $100 per slot. No rig I have ever built costs $50 to $100 per slot (excluding GPU). If AMD ever changed their drivers to allow 16 (or better unlimited) GPU per system well then it might make more sense.
I really don't know what to do guys, LOL.
Should I buy this ?
I think this is quite good where prices of 2*5870 << price of 5970 ?
Should I buy this ?
I think this is quite good where prices of 2*5870 << price of 5970 ?
I'll summarise some points in that page as best as I can.
1. Prototype board.
2. It runs only 1x so it's not suitable for gaming.
3. Only tested in Linux. But should work in windows, as there is no software to install.
4. (If I read this correctly) There is separate 12V connector to power the cards connected to this daughter board. (I think you can see it in one of the pictures)
5. Chinese will take over the world. (j/k. My own words)
That's all.
Well, let's see...
$60 via board + $10 ram + $220 backplane for 7 cards. $41.43/slot
$50 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 8 cards. $40.00/slot
$80 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 10 cards. $35/slot (and lots of fun with 8 card driver limit)
vs.
$80 am3 board + $30 sempron + $10 ram + 2 $10 extenders for 4 cards. $35/slot.
Meh.
This. That is why I have no interest in PCI-E expansion boards. $60 via board + $10 ram + $220 backplane for 7 cards. $41.43/slot
$50 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 8 cards. $40.00/slot
$80 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 10 cards. $35/slot (and lots of fun with 8 card driver limit)
vs.
$80 am3 board + $30 sempron + $10 ram + 2 $10 extenders for 4 cards. $35/slot.
Meh.
Well, let's see...
$60 via board + $10 ram + $220 backplane for 7 cards. $41.43/slot
$50 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 8 cards. $40.00/slot
$80 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 10 cards. $35/slot (and lots of fun with 8 card driver limit)
vs.
$80 am3 board + $30 sempron + $10 ram + 2 $10 extenders for 4 cards. $35/slot.
Meh.
$60 via board + $10 ram + $220 backplane for 7 cards. $41.43/slot
$50 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 8 cards. $40.00/slot
$80 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 10 cards. $35/slot (and lots of fun with 8 card driver limit)
vs.
$80 am3 board + $30 sempron + $10 ram + 2 $10 extenders for 4 cards. $35/slot.
Meh.
Shouldn't the vs comparison use 7 cards?
i.e.,
$80 am3 board + $30 sempron + $10 ram + 5 $10 extenders for 7 cards. ~$24/slot
Which I was hoping would make the standard version look worse. *slaps forehead*
edit; I suppose the other benefit would be the ability to use an even cheaper mobo with 1x 16x slot.
Btw, where do you find a AMx board w/ 7 PCIe slots for $80?
And yes, I left out quite a few things. But start adding stuff like PSUs, materials/time for custom racks, PDUs, ventilation, ... and the whole thing gets overly complex. and the +$/slot ends up about the same.
Well, let's see...
$60 via board + $10 ram + $220 backplane for 7 cards. $41.43/slot
$50 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 8 cards. $40.00/slot
$80 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 10 cards. $35/slot (and lots of fun with 8 card driver limit)
vs.
$80 am3 board + $30 sempron + $10 ram + 2 $10 extenders for 4 cards. $35/slot.
Meh.
$60 via board + $10 ram + $220 backplane for 7 cards. $41.43/slot
$50 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 8 cards. $40.00/slot
$80 am3 board + $30 sempron + $10 ram + $10 extender + $220 backplane for 10 cards. $35/slot (and lots of fun with 8 card driver limit)
vs.
$80 am3 board + $30 sempron + $10 ram + 2 $10 extenders for 4 cards. $35/slot.
Meh.
Shouldn't the vs comparison use 7 cards?
i.e.,
$80 am3 board + $30 sempron + $10 ram + 5 $10 extenders for 7 cards. ~$24/slot
Which I was hoping would make the standard version look worse. *slaps forehead*
edit; I suppose the other benefit would be the ability to use an even cheaper mobo with 1x 16x slot.
Jump to: