Oh, one other thing...
I had the opcode space to increase the 512KB address range to 1024KB (from 19 to 20 bits).
The silicon will still only have 512KB of RAM, but this extra bits gives some room for future memory increase in denser technologies.
Oh, one other thing...
I had the opcode space to increase the 512KB address range to 1024KB (from 19 to 20 bits).
The silicon will still only have 512KB of RAM, but this extra bits gives some room for future memory increase in denser technologies.
Sounds good, maybe that new space can trap or hw-map to off-chip memory for eXecute In Place designs ?
I just checked Altera's site to find out if they now support the Cyclone V -A9 with Quartus II Web Edition, and they don't, according to a document from last year (2014).
I just checked Altera's site to find out if they now support the Cyclone V -A9 with Quartus II Web Edition, and they don't, according to a document from last year (2014).
The CV A9 data sheet says "Quartus II" with no mention of the free Quartus II Web Edition. If Web Edition doesn't support it, that will be $3999 (last I knew) for a Quartus II Subscription Edition. It's odd they'd make a board for $149 with a $228 FPGA (Digi-Key) and leave someone to discover they'd need to spend another $4k to use it. Altera's website is about useless to determine what the actual case is. I guess someone needs to download the latest Web Edition, install it, and see if it lets them select an -A9 device. I'll download it now, just to get updated and find out.
...Well, it won't let me log in, even after it sent me a new temporary password - twice! This madness has gotten to be typical with about everything.
...Okay. It sent me a third, unsolicited, email and that password change worked. I'll download V15 now.
I just checked Altera's site to find out if they now support the Cyclone V -A9 with Quartus II Web Edition, and they don't, according to a document from last year (2014).
The CV A9 data sheet says "Quartus II" with no mention of the free Quartus II Web Edition. If Web Edition doesn't support it, that will be $3999 (last I knew) for a Quartus II Subscription Edition. It's odd they'd make a board for $149 with a $228 FPGA (Digi-Key) and leave someone to discover they'd need to spend another $4k to use it. Altera's website is about useless to determine what the actual case is. I guess someone needs to download the latest Web Edition, install it, and see if it lets them select an -A9 device. I'll download it now, just to get updated and find out
I just checked Altera's site to find out if they now support the Cyclone V -A9 with Quartus II Web Edition, and they don't, according to a document from last year (2014).
The CV A9 data sheet says "Quartus II" with no mention of the free Quartus II Web Edition. If Web Edition doesn't support it, that will be $3999 (last I knew) for a Quartus II Subscription Edition. It's odd they'd make a board for $149 with a $228 FPGA (Digi-Key) and leave someone to discover they'd need to spend another $4k to use it. Altera's website is about useless to determine what the actual case is. I guess someone needs to download the latest Web Edition, install it, and see if it lets them select an -A9 device. I'll download it now, just to get updated and find out.
Chip it's a really recent change and there's conflicting information out there, but there are other forums where users have got other things running. OzPropDev has done a test compile for that target successfully, read later in that same thread. You need version 15.0
Thanks for the updated info. If looks like you're meeting the original design targets (16 cogs, 512kB hub, analog), which is fantastic, but making provision for the future too.
The CV A9 data sheet says "Quartus II" with no mention of the free Quartus II Web Edition. If Web Edition doesn't support it, that will be $3999 (last I knew) for a Quartus II Subscription Edition. It's odd they'd make a board for $149 with a $228 FPGA (Digi-Key) and leave someone to discover they'd need to spend another $4k to use it. Altera's website is about useless to determine what the actual case is. I guess someone needs to download the latest Web Edition, install it, and see if it lets them select an -A9 device. I'll download it now, just to get updated and find out.
As well as comments in this forum, Google groups have this
["If I recall correctly, it was mentioned on the sdrstick group that the free Quartus 15.0 would support the CV A9 starting around May or June."]
& this post June 10th 2015["My CV-A9 was delivered this morning and I have managed to get the leds flashing, so I can confirm that free version of Quartus 15 does support it."]
Wow! We need to get an -A9 on our Prop 1-2-3 board, then. We're getting a good deal on the -A7 at just under $100. Uh, but maybe not that good of a deal if BE Micro can sell an -A9 board for $149. I think they must be getting that chip for what we are getting an -A7 for, or way less. Even so, those cost metrics don't make any sense for a U.S. company.
I just talked to Daniel who did our Prop 1-2-3 board and we are going to see if there is pin compatibility between the -A7 and -A9. There may be. Next, we've got to get a good price from Altera, ...er Intel.
Chasing wind.
Getting an A9 onto the 1-2-3 would be brilliant
The jump in price from A7 to A9 is nowhere near as bad as you might expect. About an additional 27% according to digikey prices.
In discussions with Altera (Intel?) shaking a finger at the low price of the BeMicroCV A9 might help you cut a good deal.
Wow! We need to get an -A9 on our Prop 1-2-3 board, then. We're getting a good deal on the -A7 at just under $100. Uh, but maybe not that good of a deal if BE Micro can sell an -A9 board for $149. I think they must be getting that chip for what we are getting an -A7 for, or way less. Even so, those cost metrics don't make any sense for a U.S. company.
I just talked to Daniel who did our Prop 1-2-3 board and we are going to see if there is pin compatibility between the -A7 and -A9. There may be. Next, we've got to get a good price from Altera, ...er Intel.
Chasing wind.
Not only that, it may make it easier to allow some of us to use the BEMicro CV A9 as a cheaper alternative to the 1-2-3 board (without the smart pin stuff, I'm sure).
The web edition of Quartus II 15.0 only goes up to the Cyclone V A6, I just checked with the device selection. I think I'm using the current version.
I can't see many people buying the 1-2-3 board if they have to buy the full version of Quartus II.
The web edition of Quartus II 15.0 only goes up to the Cyclone V A6, I think I'm using the current version. I just checked with the device selection.
I can't see many people buying the 1-2-3 board if they have to buy the full version of Quartus II.
I rather wish Chip had not picked today to post here. The explosion of new posts that it engendered has made it all the more difficult to roll this forum back to 28 June, which is what really needs to happen.
OTOH, would the consequent posts really be missed?
Seems to need the 15.0.1.150 update too ?["Altera TeacherI have Quartus 15 Web Edition with update (can be downloaded from Altera website) and I can generate A9 bitstream. Do you want a screenshot?"]
I rather wish Chip had not picked today to post here. The explosion of new posts that it engendered has made it all the more difficult to roll this forum back to 28 June, which is what really needs to happen.
OTOH, would the consequent posts really be missed?
-Phil
We need to roll it back to before it started looking like Hello Kitty.
I poured over the pinouts for the -A7 and -A9 devices in the F484 package and the ONLY difference is that sixteen I/O pins on the -A7 become no-connects on the -A9. Here they are:
G20
E22H20D22C19B22C18A22F19E20F18F20A18A20A17A19
They are all on Bank 7A, which we run at 3.3V. I know where I can lose at least 10, or so. It will require a new PCB layout, but hopefully no new solder stencil or pick-and-place program.
koehler said:
Rather than rebut, make it personal.
Happily, its not a trend here.
Evanh:
Pretty simple really. You've been rebuked ten times over on every point individually already. You've become a stuck record.
Damn!, this new forum software leaves a bit to be desired. I'm squished into a two line edit window!
Rebuked 10 times, who, where when?
If you like being a mushroom, thats your call.
More than a few people seem to 'crying' for some sort of update, even if you don't.
At a minimum, most of us now for the present have an idea of why things seem to have been stalled for a while.
You're welcome.
Aaanndd, thats a heck of an update from Chip.
Looks like while he has been off in the weeds with the export munge issue, most of the features discussed previously are still looking good for future completion and even HuvExec is still expected.
Going with -A9 won't affect the initial 1-2-3 FPGA deliveries, will it? My understanding is that there are a bunch of boards already made, and they are just waiting for the P2 FPGA image before they will be sold and delivered. I hope we're not talking about delaying this so that the -A9 can be used.
And here's the problem with updates. While they are interesting, they lead to forum discussion with Chip, and new ideas (some good) get thrown into the mix, causing a turnoff the freeway while a new direction is explored. Where will that new direction lead? - to getting to the destination quicker, probably not, as one's mind winds down from one task, spins up to the new task, explores the new, then has to wind down/spin up to the original task.
Having led development programs, one of my most difficult jobs was keeping creative scientists and engineers on task. They would always find new things that would "make the project better", but usually threatened to derail the main project schedule. When I had sufficient people, I could sometimes assign a junior person to run down the new idea to minimize (not eliminate) the delay of the senior developer. But with a one person project or where the person with the new idea was the critical path, delays built up.
As Chip wrote in his update: "It takes days for me to spool back up where I left off on the Verilog, afterwards."
I poured over the pinouts for the -A7 and -A9 devices in the F484 package and the ONLY difference is that sixteen I/O pins on the -A7 become no-connects on the -A9. Here they are:
G20
E22H20D22C19B22C18A22F19E20F18F20A18A20A17A19
They are all on Bank 7A, which we run at 3.3V. I know where I can lose at least 10, or so. It will require a new PCB layout, but hopefully no new solder stencil or pick-and-place program.
If there is serious consideration of putting an A9 on the 1-2-3 board, could we get a complete I/O listing? I'd like to see how it matches up to the BEMicro CV A9.
Going with -A9 won't affect the initial 1-2-3 FPGA deliveries, will it? My understanding is that there are a bunch of boards already made, and they are just waiting for the P2 FPGA image before they will be sold and delivered. I hope we're not talking about delaying this so that the -A9 can be used.
I'm guessing the A9 version would be a Rev B, since Chip indicated it would be a new PCB layout.
Maybe Parallax could lend out the A7 version to get the P2 tested (the most important use case at the moment), then build an A9 version ( Rev B ) to actually sell. The A7 versions could still be used as part of Parallax's educational offerings.
Comments
I had the opcode space to increase the 512KB address range to 1024KB (from 19 to 20 bits).
The silicon will still only have 512KB of RAM, but this extra bits gives some room for future memory increase in denser technologies.
That's a very nice feature!
I had the opcode space to increase the 512KB address range to 1024KB (from 19 to 20 bits).
The silicon will still only have 512KB of RAM, but this extra bits gives some room for future memory increase in denser technologies.
Sounds good, maybe that new space can trap or hw-map to off-chip memory for eXecute In Place designs ?
In the thread called BeMicro CV A9 - 301.000K LE this posthttp://forums.parallax.com/discussion/comment/1333218/#Comment_1333218
mentions newest Quartus 15.0 webpack now supports A9 ?That CV A9 board is ~$149, so it is quite a good price point.
In the thread called BeMicro CV A9 - 301.000K LE this posthttp://forums.parallax.com/discussion/comment/1333218/#Comment_1333218
mentions newest Quartus 15.0 webpack now supports A9 ?That CV A9 board is ~$149, so it is quite a good price point.
The CV A9 data sheet says "Quartus II" with no mention of the free Quartus II Web Edition. If Web Edition doesn't support it, that will be $3999 (last I knew) for a Quartus II Subscription Edition. It's odd they'd make a board for $149 with a $228 FPGA (Digi-Key) and leave someone to discover they'd need to spend another $4k to use it. Altera's website is about useless to determine what the actual case is. I guess someone needs to download the latest Web Edition, install it, and see if it lets them select an -A9 device. I'll download it now, just to get updated and find out.
...Okay. It sent me a third, unsolicited, email and that password change worked. I'll download V15 now.
In the thread called BeMicro CV A9 - 301.000K LE this posthttp://forums.parallax.com/discussion/comment/1333218/#Comment_1333218
mentions newest Quartus 15.0 webpack now supports A9 ?That CV A9 board is ~$149, so it is quite a good price point.
The CV A9 data sheet says "Quartus II" with no mention of the free Quartus II Web Edition. If Web Edition doesn't support it, that will be $3999 (last I knew) for a Quartus II Subscription Edition. It's odd they'd make a board for $149 with a $228 FPGA (Digi-Key) and leave someone to discover they'd need to spend another $4k to use it. Altera's website is about useless to determine what the actual case is. I guess someone needs to download the latest Web Edition, install it, and see if it lets them select an -A9 device. I'll download it now, just to get updated and find out
I just checked Altera's site to find out if they now support the Cyclone V -A9 with Quartus II Web Edition, and they don't, according to a document from last year (2014).
In the thread called BeMicro CV A9 - 301.000K LE this posthttp://forums.parallax.com/discussion/comment/1333218/#Comment_1333218
mentions newest Quartus 15.0 webpack now supports A9 ?That CV A9 board is ~$149, so it is quite a good price point.
The CV A9 data sheet says "Quartus II" with no mention of the free Quartus II Web Edition. If Web Edition doesn't support it, that will be $3999 (last I knew) for a Quartus II Subscription Edition. It's odd they'd make a board for $149 with a $228 FPGA (Digi-Key) and leave someone to discover they'd need to spend another $4k to use it. Altera's website is about useless to determine what the actual case is. I guess someone needs to download the latest Web Edition, install it, and see if it lets them select an -A9 device. I'll download it now, just to get updated and find out.
Chip it's a really recent change and there's conflicting information out there, but there are other forums where users have got other things running. OzPropDev has done a test compile for that target successfully, read later in that same thread. You need version 15.0
Thanks for the updated info. If looks like you're meeting the original design targets (16 cogs, 512kB hub, analog), which is fantastic, but making provision for the future too.
The CV A9 data sheet says "Quartus II" with no mention of the free Quartus II Web Edition. If Web Edition doesn't support it, that will be $3999 (last I knew) for a Quartus II Subscription Edition. It's odd they'd make a board for $149 with a $228 FPGA (Digi-Key) and leave someone to discover they'd need to spend another $4k to use it. Altera's website is about useless to determine what the actual case is. I guess someone needs to download the latest Web Edition, install it, and see if it lets them select an -A9 device. I'll download it now, just to get updated and find out.
As well as comments in this forum, Google groups have this
["If I recall correctly, it was mentioned on the sdrstick group that the free Quartus 15.0 would support the CV A9 starting around May or June."]
& this post June 10th 2015["My CV-A9 was delivered this morning and I have managed to get the leds flashing, so I can confirm that free version of Quartus 15 does support it."]
I just talked to Daniel who did our Prop 1-2-3 board and we are going to see if there is pin compatibility between the -A7 and -A9. There may be. Next, we've got to get a good price from Altera, ...er Intel.
Chasing wind.
The jump in price from A7 to A9 is nowhere near as bad as you might expect. About an additional 27% according to digikey prices.
In discussions with Altera (Intel?) shaking a finger at the low price of the BeMicroCV A9 might help you cut a good deal.
I just talked to Daniel who did our Prop 1-2-3 board and we are going to see if there is pin compatibility between the -A7 and -A9. There may be. Next, we've got to get a good price from Altera, ...er Intel.
Chasing wind.
Not only that, it may make it easier to allow some of us to use the BEMicro CV A9 as a cheaper alternative to the 1-2-3 board (without the smart pin stuff, I'm sure).
I can't see many people buying the 1-2-3 board if they have to buy the full version of Quartus II.
I can't see many people buying the 1-2-3 board if they have to buy the full version of Quartus II.
It's a recent (few weeks old) change. There are certainly people successfully compiling for A9's, OzPropDev confirmed you can successfully compile P1V's for that A9 target in the P1V forum.
Here's an altera forums threadhttp://www.alteraforum.com/forum/showthread.php?t=48871&highlight=bemicro
OTOH, would the consequent posts really be missed?
-Phil
It's a recent (few weeks old) change. There are certainly people successfully compiling for A9's, OzPropDev confirmed you can successfully compile P1V's for that A9 target in the P1V forum.
Here's an altera forums threadhttp://www.alteraforum.com/forum/showthread.php?t=48871&highlight=bemicro
Seems to need the 15.0.1.150 update too ?["Altera TeacherI have Quartus 15 Web Edition with update (can be downloaded from Altera website) and I can generate A9 bitstream. Do you want a screenshot?"]
Have tried multiple A9 compiles, all works fine
Have tried multiple A9 compiles, all works fine
Is that with the 15.0.1 update applied ?
http://dl.altera.com/?edition=web#tabs-5
It's a 1.1 Gb file.
No. Update 1 not loaded yet. Had issues with Altera's site so is on ToDo list.
OTOH, would the consequent posts really be missed?
-Phil
We need to roll it back to before it started looking like Hello Kitty.
G20
E22H20D22C19B22C18A22F19E20F18F20A18A20A17A19
They are all on Bank 7A, which we run at 3.3V. I know where I can lose at least 10, or so. It will require a new PCB layout, but hopefully no new solder stencil or pick-and-place program.
There certainly has been a lot of gotchas in trying to tie all this together.
koehler said:
Rather than rebut, make it personal.
Happily, its not a trend here.
Evanh:
Pretty simple really. You've been rebuked ten times over on every point individually already. You've become a stuck record.
Damn!, this new forum software leaves a bit to be desired. I'm squished into a two line edit window!
Rebuked 10 times, who, where when?
If you like being a mushroom, thats your call.
More than a few people seem to 'crying' for some sort of update, even if you don't.
At a minimum, most of us now for the present have an idea of why things seem to have been stalled for a while.
You're welcome.
Aaanndd, thats a heck of an update from Chip.
Looks like while he has been off in the weeds with the export munge issue, most of the features discussed previously are still looking good for future completion and even HuvExec is still expected.
Have tried multiple A9 compiles, all works fine
Is that with the 15.0.1 update applied ?
I've got 13 with the 15.0.1 update.
Having led development programs, one of my most difficult jobs was keeping creative scientists and engineers on task. They would always find new things that would "make the project better", but usually threatened to derail the main project schedule. When I had sufficient people, I could sometimes assign a junior person to run down the new idea to minimize (not eliminate) the delay of the senior developer. But with a one person project or where the person with the new idea was the critical path, delays built up.
As Chip wrote in his update: "It takes days for me to spool back up where I left off on the Verilog, afterwards."
Tom
G20
E22H20D22C19B22C18A22F19E20F18F20A18A20A17A19
They are all on Bank 7A, which we run at 3.3V. I know where I can lose at least 10, or so. It will require a new PCB layout, but hopefully no new solder stencil or pick-and-place program.
If there is serious consideration of putting an A9 on the 1-2-3 board, could we get a complete I/O listing? I'd like to see how it matches up to the BEMicro CV A9.
I'm guessing the A9 version would be a Rev B, since Chip indicated it would be a new PCB layout.
Maybe Parallax could lend out the A7 version to get the P2 tested (the most important use case at the moment), then build an A9 version ( Rev B ) to actually sell. The A7 versions could still be used as part of Parallax's educational offerings.
We care, we know you are aware... and proactive about it. All the acrimony will taper off, eventually.