On 03/12/14 13:43, Rick C. Hodgin wrote:> David, > > Jesus loves you. And I love you. Seek Him. > > Best regards, > Rick C. Hodgin >Does that mean you agree with what I wrote? Or that you disagree? Or that you don't understand it? Or that you just don't want to think about it?
Which Altera to buy?
Started by ●December 2, 2014
Reply by ●December 3, 20142014-12-03
Reply by ●December 3, 20142014-12-03
David, You are always writing posts to guide me in my endeavors. It is a good sentiment, and I wanted to do likewise. Your skills and knowledge are a great asset. Use them wisely. Best regards, Rick C. Hodgin
Reply by ●December 3, 20142014-12-03
rickman wrote:> I can't seem to find it now, but someone recently posted a link to > price/LUT vs size data in graph form. It gets a bit crowded at the > bottom end, but appears to show there is no real price difference > between the two brands. The data does include a very small number of > other devices than X and A, but not enough to be useful. > > In fact it is interesting that the prices get very crowded at the low > end jamming up the graph. I suggested that he present the data with a > logarithmic Y axis or even in log-log form. Clearly competitive market > forces at work. >The "low end" is where I live! So, I don't know about those $18,000 FPGAs, or who the heck USES them. I know that in very small quantities, the Xilinx XC3S50A is down to $6.12! This is pretty amazing! The non-volatile version of the part (XC3S50AN) is $9.91. These prices have actually come DOWN sometime this year! I have no idea what an equivalent Altera part would cost, due to the apples vs. oranges of different internal architecture. Jon
Reply by ●December 3, 20142014-12-03
rickman wrote:> I can't seem to find it now, but someone recently posted a link to > price/LUT vs size data in graph form. It gets a bit crowded at the > bottom end, but appears to show there is no real price difference > between the two brands. The data does include a very small number of > other devices than X and A, but not enough to be useful. > > In fact it is interesting that the prices get very crowded at the low > end jamming up the graph. I suggested that he present the data with a > logarithmic Y axis or even in log-log form. Clearly competitive market > forces at work. >Spartan 3 is not included in the chart. In the log-log version (THANKS, Theo!) it is clear that the Spartan 6 does VERY well against most other types in the price/LUT. And, Virtex 5 is just about the WORST! You can't compare Spartan 3 vs. Spartan 6 due to the significant changes in the LUT capacity, unless you have some kind of correction factor to apply. I have NO IDEA how one would establish such a factor, though! Jon
Reply by ●December 3, 20142014-12-03
Jon Elson wrote:> rickman wrote: > > >> I can't seem to find it now, but someone recently posted a link to >> price/LUT vs size data in graph form. It gets a bit crowded at the >> bottom end, but appears to show there is no real price difference >> between the two brands. The data does include a very small number of >> other devices than X and A, but not enough to be useful. >> >> In fact it is interesting that the prices get very crowded at the low >> end jamming up the graph. I suggested that he present the data with a >> logarithmic Y axis or even in log-log form. Clearly competitive market >> forces at work. >> > Spartan 3 is not included in the chart. In the log-log version > (THANKS, Theo!) it is clear that the Spartan 6 does VERY well against > most other types in the price/LUT. > > And, Virtex 5 is just about the WORST! > > You can't compare Spartan 3 vs. Spartan 6 due to the significant changes > in the LUT capacity, unless you have some kind of correction factor > to apply. I have NO IDEA how one would establish such a factor, though! > > JonSometime recently we went out for quotes on Spartan 6 and Artix 7 parts looking for the best price for a part with 4 GTP transceivers. The Artix-7 won that race, but the parts with fewer than 100K LE are still hard to get. The smallest Artix 7 (and possibly cheapest Xilinx FPGA) is the XC7A15T, which is only supported by the latest Vivado release and looks to be 15-18 weeks out. -- Gabor
Reply by ●December 3, 20142014-12-03
Note that quite a few of the more expensive Altera eval boards ship with a "free" design tools DVD. The problem is, to actually use those design tools you need a license which is most definitely not free (possibly several times more expensive than the board itself) --------------------------------------- Posted through http://www.FPGARelated.com
Reply by ●December 3, 20142014-12-03
On Wednesday, December 3, 2014 2:01:17 PM UTC-6, Jon Elson wrote:> rickman wrote: > > > > I can't seem to find it now, but someone recently posted a link to > > price/LUT vs size data in graph form. It gets a bit crowded at the > > bottom end, but appears to show there is no real price difference > > between the two brands. The data does include a very small number of > > other devices than X and A, but not enough to be useful. > > > > In fact it is interesting that the prices get very crowded at the low > > end jamming up the graph. I suggested that he present the data with a > > logarithmic Y axis or even in log-log form. Clearly competitive market > > forces at work. > > > Spartan 3 is not included in the chart. In the log-log version > (THANKS, Theo!) it is clear that the Spartan 6 does VERY well against > most other types in the price/LUT. > > And, Virtex 5 is just about the WORST! > > You can't compare Spartan 3 vs. Spartan 6 due to the significant changes > in the LUT capacity, unless you have some kind of correction factor > to apply. I have NO IDEA how one would establish such a factor, though! > > JonRan a number of opcores.org processor designs on Spartan 3&6, Kintex-7, Cyclone 2&4 and Arria II. ("http://opencores.org/project,up_core_list,downloads" pull "family comparison; numbers below are from a more recent version) The LUT count averages are: S-3: 1.46 4LUTs to one ALUT S-6 & K-7: 1.05 6LUTs to one ALUT C-2 & C-4: 1.58 4LUTs to one ALUT So roughly 1.5 4LUTs per 6LUT or ALUT. Fmax generally scales with process node (altor32, atlas_2K-base, eco32, leros, m1_core, mblite, navre, next186, pic16c5x, pdp11-34verilob, risc5, t65 & tv80; typically 1K to 4K LUT designs)
Reply by ●December 4, 20142014-12-04
I ended up ordering this board:
http://www.terasic.com.tw/cgi-bin/page/archive.pl?Language=English&CategoryNo=167&No=830
I was informed in email today I should receive it early next week.
I'm going to spend some time getting communication working between the
FPGA and the host computer as a debugging feature of the Oppie-1 CPU.
I will modify my Oppie-1 Debugger to receive the debug data from the
live device and display it in execution as it goes, or step-by-step,
as it does presently in the simulation.
I will also be able to send/receive updates to memory locations, change
register values, reset, restart, a true remote debugging app.
Then ... on to Oppie-2.
-----
If anyone has advice on how to get communication running most easily
on this board, I would appreciate it. Thank you in advance.
Best regards,
Rick C. Hodgin
Reply by ●December 4, 20142014-12-04
On 12/4/2014 12:25 PM, Rick C. Hodgin wrote:> I ended up ordering this board: > http://www.terasic.com.tw/cgi-bin/page/archive.pl?Language=English&CategoryNo=167&No=830 > > I was informed in email today I should receive it early next week. > > I'm going to spend some time getting communication working between the > FPGA and the host computer as a debugging feature of the Oppie-1 CPU. > I will modify my Oppie-1 Debugger to receive the debug data from the > live device and display it in execution as it goes, or step-by-step, > as it does presently in the simulation. > > I will also be able to send/receive updates to memory locations, change > register values, reset, restart, a true remote debugging app. > > Then ... on to Oppie-2. > > ----- > If anyone has advice on how to get communication running most easily > on this board, I would appreciate it. Thank you in advance.Doesn't look like you have a lot of options on this board. UART serial to USB is the one comms choice. I thought you were going to use Ethernet. That seems to be another $250, wow, more than the FPGA board. I guess you can add one via the Arduino interface for next to nothing. -- Rick
Reply by ●December 4, 20142014-12-04
On Thursday, December 4, 2014 5:28:25 PM UTC-5, rickman wrote:> On 12/4/2014 12:25 PM, Rick C. Hodgin wrote: > > I ended up ordering this board: > > http://www.terasic.com.tw/cgi-bin/page/archive.pl?Language=English&CategoryNo=167&No=830 > > > > If anyone has advice on how to get communication running most easily > > on this board, I would appreciate it. Thank you in advance. > > Doesn't look like you have a lot of options on this board. UART serial > to USB is the one comms choice. I thought you were going to use > Ethernet.I plan to. I won't receive the PHY board I bought until the end of December though. If there's another option I'll use that in the time in-between, and possible after that. I have one of the Silicon Labs chips arriving this week, but I'll need to get a converter from its TQFP48 form factor to something usable by human beings. I may go ahead and do that anyway. The Silicon Labs API is very clean and straight-forward. http://www.silabs.com/products/interface/ethernetcontrollers/Pages/default.aspx> That seems to be another $250, wow, more than the FPGA board. > I guess you can add one via the Arduino interface for next to > nothing.Best regards, Rick C. Hodgin






