difference between XC5VLX50-1FFG676C and XC5VLX50-1FFG676CES

Hi alls,

I've a question which seems stupid, but I don't find answer anywhere.. ( maybe I don't search correctly ) What is the difference between XC5VLX50-1FFG676C and XC5VLX50-1FFG676CES ? I don't understand what does 'ES' mean at the end of the FPGA reference ..

So, please, if anyone knows..

Thanks by advance,

best regards, Michel.

Reply to
michel.talon
Loading thread data ...

I would bet that is stands for "engineering sample"

Reply to
Mike Lewis

Michel,

ES = "engineering sample"

This is what we call parts BEFORE they are qualified for production.

In some cases, ES parts are completely identical to the final production, and in other cases, they are not identical (they have known bugs or problems we call "errata").

CES = "commercial grade engineering sample" (it has a speed grade)

Go look for V5 errata at:

formatting link

to find the exact differences.

If you have questions about any of these errata, please contact your distributor or Xilinx FAE.

We sell ES parts to customers early in order to give them a head-start on their design and debugging (over their competition).

The risk of purchasing an ES part, is they may have bugs (known, as detailed in the errata, or as of yet unknown, as all testing may not be complete nor all test programs in place). If the parts have bugs, there is no warranty (they are non-returnable, non-refundable). The normal warranties do not apply to an ES part.

As an example, ES parts are never accepted for a "RMA" (returned merchandise authorization) or for F/A (failure analysis).

It is recommended that ES never be used for production product. If used for production product, the purchaser must accept all of the risks, as Xilinx has stated these parts are not intended nor recommended for production product.

This is pretty standard practice in the industry, and is not unique to FPGA technology either.

Austin

Reply to
austin

Look here:

The link was found by doing a search for XC5VLX50-1FFG676CES on the Xilinx web site.

G.

Reply to
ghelbig

Makes perfect sense.

You guys don't want to determine why ES parts fail? That seems really strange.

Reply to
Eric Smith

Eric,

F/A rarely shows us anything useful. If the part has been fried (electrical over-stress) that is easily determined with an ohmmeter.

If the part has a particle defect, or some other process defect, we already knew these parts are from the first ever lots of a new process, and yes, that happens quite often. Statistics on defects are kept with "defect monitor" vehicles, so we don't care to tear apart one known early part and see one defect we already have seen before in the monitors.

As long as the part is on the customer's board, we have a chance at determining the problem. Once removed, determination of root cause drops to less than 5% success rate (as most problems are with the user's design, followed by their signal integrity, followed by solder problems).

So, yes, we are interested in why production parts fail, but no, we are not interested in performing costly F/A on an ES part.

We might actually request return of an ES part because we feel we may learn something (like develop a better production test), but that is pretty rare.

Good question. Thanks for asking it so I can educate folks,

Aust> aust>> As an example, ES parts are never accepted for a "RMA" (returned

Reply to
austin

Thank you for your answers!

Now I understand ! :-)

Reply to
michel.talon

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.