[Cryptech Tech] Alpha board strategy

Joachim Strömbergson joachim at secworks.se
Fri Feb 20 09:50:12 UTC 2015


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Aloha!

Павел Шатов wrote:
> On 19.02.2015 14:45, Joachim Strömbergson wrote:
>> The Cyclone-IV is manufactured in an older technology that is
>> slower than Cyclone-V. The biggest Cyclone-IV is as big as the
>> biggest Cyclone-V that we can build with free tools (A7). Thus
>> selecting Cyclone-IV over Cyclone-V would only remove the
>> possibility of adding more headroom by buying tools. Also, if you
>> look at pricing, Cyclone-IV is not really cheaper than Cyclone-V.

> That is exactly what I wanted to say. Largest Cyclone-IV is larger
> than largest "free" Cyclone-V. Maybe this will be enough and we won't
> have to buy development tools.

No, that is not correct. The largest Cyclone-IV decvice is the EP4CGX150
it provides 149 kLEs. The Cyclone A7 provides 149 kLEs. Cyclone-IV is
not interesting, it is an older process and there is no room to grow
even if we take the hit and pay for the license.


> Artix-7 is relatively new (as is Cyclone-V). As far as I remember,
> at the time it was launched, only two largest chips in the family
> were available (100T and 200T). I believe that smaller chips should
> now be available too. I've taken prices for my table from Digi-Key: 
> http://www.digikey.com/product-detail/en/XC7A100T-1FGG484C/122-1885-ND/3925804
>
> 
http://www.digikey.com/product-detail/en/XC7A200T-1FBG484C/122-1871-ND/3925796

Yes, there are now smaller chips available (35T and 50T) and with the
same package. For use cases that want something smaller and cheaper it
is a nice solution.

Comparing the prices for Cyclone-V and Artix-7 is interesting. It scales
fairly linearly with capacity even though there are different vendors
and families. Says something about how competitive their device designs
are and that ultimately it comes down to die area. And the bigger the
more kLEs/USD.

A7 single unit prices with 484-BGA. Cheapest 159 USD
http://shorl.com/prybrykokistigy

100 kLE: 137 USD (T100)
150 kLE: 159  USD (A7)
210 kLE: 186 USD (T200)

The Artix-7 T200 really looks very interesting. It would provide ample
headroom (more than 4x resources compared to the LX45 or C5) and can be
designed with free tools. You have almost convinced me to go for the
T200. Too bad the ISE is so lousy compared to Quartus.

The increased cost from A7 to T200 is not big (17%) and would remove the
need to provide space on the PCB for a secondary FPGA. Makes for a nicer
board.

A T150 that made the step in resources between T100 and T200 smaller
would have been nice, but who am I to complain?

- -- 
Med vänlig hälsning, Yours

Joachim Strömbergson - Alltid i harmonisk svängning.
========================================================================
 Joachim Strömbergson          Secworks AB          joachim at secworks.se
========================================================================
-----BEGIN PGP SIGNATURE-----
Version: GnuPG/MacGPG2 v2
Comment: GPGTools - http://gpgtools.org
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
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=9EIT
-----END PGP SIGNATURE-----


More information about the Tech mailing list