Tim Newsom wrote:
Bah! I meant to copy the list on that question.
Thanks for the answer though. Maybe someone else can also help
clarify? I thought fpga were basically PLDs and that they worked
exactly the same. I didn't know they lose config without power and
need to be reprogrammed.
Nowadays the line that separates FPGAs from CPLDs is blurry and even
different vendors have different sayings about it. Initially, PLDs were
devices, which could hold a rather limited amount of logic, yet they
kept their configuration after being power cycled. Then FPGAs came
along, which could hold a much larger amount of logic, but they lose
their configuration whenever power is lost. So FPGAs need to be
programmed each time the power is cycled (It's interesting to note that
a FPGA is actually based on SRAM and LUTs.)
Also, note that FPGAs can work in either 'passive' or 'active' mode.
When in 'passive', someone needs to externally initiate the programming
of the device, tipically a JTAG chain. When in 'active' mode, the FPGA
will try to fetch its own configuration from a (small) ROM connected to
it, this allows for easy configuration in standalone devices.
Then came along CPLDs which offered more density than the CPLDs, though
not as a much as a FPGA, but they kept configuration even without power.
FPGAs however, are the most popular devices today, they have the
greatest density and allow to hold complex designs such as video codecs,
DSP blocks and even whole processors.
(As of these days, I'm working on a design consisting of a FPGA holding
an entire processor along with 'custom' hardware in order to speed up a
voice codec algorithm).
To sum up:
* FPGA : the device with the greatest density, they tipically lose
configuration when power cycled (note that Xilinx offers OTP (one time
programmable) FPGAs, which keep their configuration, but that's a whole
different story....let's stick to the everyday jargon :-) )
* CPLDs : devices less dense than FPGAs, but they keep their
configuration even after losing power.
I hope this mail wasn't THAT much confusing... :-)
By the way, keep up the good work, I think the OpenMoko initiative is a
terrific idea, and if it turns out as half as good as the ideas I've
seen in this list, it will sure turn out to be a great product!
Regards, Leonardo Etcheverry
--Tim
On Fri, 8 Dec 2006 5:51, Ole Tange wrote:
As far as I understand it is like RAM: It looses state if it looses
power. So it will have to read its config from some storage to start
working.
From http://en.wikipedia.org/wiki/CPLD:
Non-volatile configuration memory. Unlike many FPGAs, an external
configuration ROM isn't required, and the CPLD can function
immediately on system start-up.
/Ole
On 12/7/06, Tim Newsom <[EMAIL PROTECTED]> wrote:
Ok... So how many times can you reprogram it before it wears out?
Like flash has a max number of times it can be written and eprom and
eeproms did... What's that number for FPGAs?
On Thu, 7 Dec 2006 15:40, Ole Tange wrote:
You cannot use them simultaneously, but you can change set in 10 ms.
/Ole
_______________________________________________
OpenMoko community mailing list
[email protected]
http://lists.openmoko.org/cgi-bin/mailman/listinfo/community
--Tim
_______________________________________________
OpenMoko community mailing list
[email protected]
http://lists.openmoko.org/cgi-bin/mailman/listinfo/community
_______________________________________________
OpenMoko community mailing list
[email protected]
http://lists.openmoko.org/cgi-bin/mailman/listinfo/community
