Algorithmically placed FPGA miner: 255MH/s/chip, supports all known boards - page 16.

eldentyrell

donator

Activity: 980

Merit: 1004

felonious vagrancy, personified

Quote from: chrisp on July 25, 2012, 07:32:53 AM

Yep. And I'm currently working on using the onboard FTDI chip for exactly this process. The code is improving and I hope to have it stable in around a week, but we will see!

Thank you! I assume you are the same Chris who emailed me; I've sent you back a reply. Let me know if you need anything else. Also, for immediate responses to questions -- particularly software-related ones -- ask jfsebastien on #tml-support at irc.tricone-mining.com.

eldentyrell

donator

Activity: 980

Merit: 1004

felonious vagrancy, personified

Quote from: ?? on ??

This user is currently ignored.

Dear Entropy-usockpuppet,

Please take your trolling to another thread.

Thanks,

- eldentyrell

eldentyrell

donator

Activity: 980

Merit: 1004

felonious vagrancy, personified

Quote from: rph on July 26, 2012, 11:04:06 PM

ET, I could use some bitstream tech support when you get a chance.. C22 no worky..

Sorry rph! Your email got lost in the shuffle. I have replied.

ShadesOfMarble

donator

Activity: 543

Merit: 500

Does anyone have instructions on how to get TML working with CM1 via an USB Blaster?

rph

full member

Activity: 176

Merit: 100

ET, I could use some bitstream tech support when you get a chance.. C22 no worky..

-rph

BR0KK

hero member

Activity: 784

Merit: 500

i think he meant average of all boards (speaking all boards out there mining for different persons).

wildemagic

member

Activity: 112

Merit: 10

Quote from: eldentyrell on July 26, 2012, 03:57:05 AM

Yes, you did miss something: the overvolting.

My bitstream gets a hell of a lot more than 255MH/s at 1.41V. But the convention is to post non-overvolted numbers to make comparison easy; otherwise things get silly very quickly. I was easily able to get 283MH/s by overvolting, but I don't post those numbers in the topic title or use them for comparison (or anything else other than amusement with a big warning up front not to take them seriously).

Ergo, the ztex bitstream does not get anywhere near 248MH/s.

Ok, my apologies, I knew there was some middle ground I was missing.

kind regards

BR0KK

hero member

Activity: 784

Merit: 500

yes you are right about that

eldentyrell

donator

Activity: 980

Merit: 1004

felonious vagrancy, personified

Quote from: BR0KK on July 26, 2012, 04:02:08 AM

My highest clock on Ztex 1.15y are 228MHZ aka 228MH/s with default board and news bitsream.

Very few of the boards get 228MH/s; the average is 210MH/s.

I have several boards that get more than 255MH/s, but again, I don't post those numbers -- I post the average.

BR0KK

hero member

Activity: 784

Merit: 500

Quote

JTAG is a daisy-chain protocol.

Imagine having six keyboards, each with a built-in USB hub -- you only need one USB port on your computer.

Ah didn't know that, thought it was on device per jtag only

My highest clock on Ztex 1.15y are 228MHZ aka 228MH/s with default board and news bitsream.

eldentyrell

donator

Activity: 980

Merit: 1004

felonious vagrancy, personified

Quote from: wildemagic on July 26, 2012, 03:45:18 AM

Quote from: kakobrekla on June 29, 2012, 08:37:23 AM

This will be my last update or at least for a while, I have achieved maximum clock 248mhz at 1.41V, which is the ztex bitstream limit.
Still waiting for new releases of ET's TML release - maybe we will see some interesting numbers there.

Am I missing something?

Yes, you did miss something: the overvolting.

My bitstream gets a hell of a lot more than 255MH/s at 1.41V. But the convention is to post non-overvolted numbers to make comparison easy; otherwise things get silly very quickly. I was easily able to get 283MH/s by overvolting, but I don't post those numbers in the topic title or use them for comparison (or anything else other than amusement with a big warning up front not to take them seriously).

Ergo, the ztex bitstream does not get anywhere near 248MH/s.

eldentyrell

donator

Activity: 980

Merit: 1004

felonious vagrancy, personified

Quote from: BR0KK on July 26, 2012, 03:48:14 AM

Quote

By the way, I use one JTAG cable for every six Spartan chips -- works out to $1.50/chip.

How do you connect 6 Spartans to one Jtag cable?

JTAG is a daisy-chain protocol.

Imagine having six keyboards, each with a built-in USB hub -- you only need one USB port on your computer.

BR0KK

hero member

Activity: 784

Merit: 500

Quote

Well, since 9 USD = 7.5 EUR my answer would be "no".

On the other hand, if you're unsatisfied with your ztex board, you probably want to talk to ztex about that.

By the way, I use one JTAG cable for every six Spartan chips -- works out to $1.50/chip.

How do you connect 6 Spartans to one Jtag cable?

wildemagic

member

Activity: 112

Merit: 10

Quote from: eldentyrell on July 26, 2012, 03:40:20 AM

Quote from: wildemagic on July 26, 2012, 03:30:47 AM

Doesnt the ztex bitstream go up to 248mh/s now?

No, the ztex bitstream gets 210MH/s.

From kakobrekla's thread https://bitcointalksearch.org/topic/m.997994

Quote from: kakobrekla on June 29, 2012, 08:37:23 AM

This will be my last update or at least for a while, I have achieved maximum clock 248mhz at 1.41V, which is the ztex bitstream limit.
Still waiting for new releases of ET's TML release - maybe we will see some interesting numbers there.

Am I missing something?

kind regards

eldentyrell

donator

Activity: 980

Merit: 1004

felonious vagrancy, personified

Quote from: punin on July 26, 2012, 03:27:22 AM

So now I need to invest over 10€ per board (or more) to actually get it running?

Well, since 9 USD = 7.5 EUR my answer would be "no".

On the other hand, if you're unsatisfied with your ztex board, you probably want to talk to ztex about that.

By the way, I use one JTAG cable for every six Spartan chips -- works out to $1.50/chip.

eldentyrell

donator

Activity: 980

Merit: 1004

felonious vagrancy, personified

Quote from: wildemagic on July 26, 2012, 03:30:47 AM

Doesnt the ztex bitstream go up to 248mh/s now?

No, the ztex bitstream gets 210MH/s.

wildemagic

member

Activity: 112

Merit: 10

Quote from: punin on July 26, 2012, 03:27:22 AM

I already spent 100€ on fans, stacking pillars and underside heat sinks for my Ztex 1.15x cluster as required to achieve notable benefit from TML. So now I need to invest over 10€ per board (or more) to actually get it running? It'll take a loooong time (if ever) to pay back with the extra ~25 mhash/s per board that I will get. Half the time if I go through the hassle of SMD rework of the voltage controller resistors & ruin my warranty for extra ~25 mhash/s.

Doesnt the ztex bitstream go up to 248mh/s now? or are you already using that?

Sorry to slightly hijack Elden...

kind regards

punin

hero member

Activity: 560

Merit: 500

Has anyone actually managed to run TML on Ztex 1.15y (quad) board?

I already spent 100€ on fans, stacking pillars and underside heat sinks for my Ztex 1.15x cluster as required to achieve notable benefit from TML. So now I need to invest over 10€ per board (or more) to actually get it running? It'll take a loooong time (if ever) to pay back with the extra ~25 mhash/s per board that I will get. Half the time if I go through the hassle of SMD rework of the voltage controller resistors & ruin my warranty for extra ~25 mhash/s.

BR0KK

hero member

Activity: 784

Merit: 500

Yes they have one (theres a second one i think too named CON 4)

eldentyrell

donator

Activity: 980

Merit: 1004

felonious vagrancy, personified

Quote from: BR0KK on July 26, 2012, 02:56:23 AM

on the ztex singles the jtag pots needs to be soldered on at first.

The 1.15y (four-chip) boards have it -- it's the black thing between the upper two chips in this image:

On the earlier boards I guess he forgot to solder it on. You can order it from him; he calls it "the debug kit".

Topic: Algorithmically placed FPGA miner: 255MH/s/chip, supports all known boards - page 16. (Read 119444 times)