Article - Extinction Level Event

[Rob Gaddi]

http://www.xess.com/blog/extinction-level-event/

The synopsis of the guy's argument is: Given Intel bought Altera, and
rumors that *comm is eyeing Xilinx, that's likely to shift the focus of
both tier 1 FPGA companies to datacenters and away from traditional
programmable logic.

We need a good Friday thread, it's been a while. What do y'all think?


--
Rob Gaddi, Highland Technology -- www.highlandtechnology.com
Email address domain is currently out of order. See above to fix.

 

[Rick C. Hodgin]

On Friday, May 27, 2016 at 1:01:13 PM UTC-4, Rob Gaddi wrote:

http://www.xess.com/blog/extinction-level-event/

The synopsis of the guy's argument is: Given Intel bought Altera, and
rumors that *comm is eyeing Xilinx, that's likely to shift the focus of
both tier 1 FPGA companies to datacenters and away from traditional
programmable logic.

We need a good Friday thread, it's been a while. What do y'all think?

People need FPGAs. You can't economically produce flexible, workable
products which are true hardware IC, and completely variable, without
them. You'd be back to the model we had in the 80s and 90s where you
must hire firms to design things for you, and that's an expense that
companies won't want ... so someone would follow-on with an FPGA, even
if it's slower because it has to work around patents. Someone will
maintain that industry because it's essential to both design, and use
in real products.

There are also several other FPGA manufacturers that make lesser
products that may immediately be viable alternatives, or are then targets
in the industry for immediate growth into larger, more capable products.

Best regards,
Rick C. Hodgin

 

[Tim Wescott]

On Fri, 27 May 2016 16:57:30 +0000, Rob Gaddi wrote:

http://www.xess.com/blog/extinction-level-event/

The synopsis of the guy's argument is: Given Intel bought Altera, and
rumors that *comm is eyeing Xilinx, that's likely to shift the focus of
both tier 1 FPGA companies to datacenters and away from traditional
programmable logic.

We need a good Friday thread, it's been a while. What do y'all think?

Someone will come and pick up the pieces, possibly at a profit.

I believe that there are east Asian FPGA companies out there -- with
luck, they'd see sales opportunities over here if the Big Two dropped out
of the race. With bad luck, America will just export raw materials to
countries who turn them into consumer goods, we'll all be dirt poor
except for a very few, and we will have lost the Revolutionary War 250
years after the fact, only not to England.

--
Tim Wescott
Control systems, embedded software and circuit design
I'm looking for work! See my website if you're interested
http://www.wescottdesign.com

 

[rickman]

On 5/27/2016 12:57 PM, Rob Gaddi wrote:

http://www.xess.com/blog/extinction-level-event/

The synopsis of the guy's argument is: Given Intel bought Altera, and
rumors that *comm is eyeing Xilinx, that's likely to shift the focus of
both tier 1 FPGA companies to datacenters and away from traditional
programmable logic.

We need a good Friday thread, it's been a while. What do y'all think?

I think the article overstates the dichotomy of supporting the
traditional FPGA market and the datacenter market. Will FPGAs need to
vary so much to support one market vs. the other? They are hugely
programmable. Neither X nor A have pushed much on alternate
architectures that might be a significant advantage in a particular
market. I think it will mostly be steady as she goes with most of the
changes in marketing rather than engineering.

--

Rick C

 

[Rob Gaddi]

rickman wrote:

On 5/27/2016 12:57 PM, Rob Gaddi wrote:
http://www.xess.com/blog/extinction-level-event/

The synopsis of the guy's argument is: Given Intel bought Altera, and
rumors that *comm is eyeing Xilinx, that's likely to shift the focus of
both tier 1 FPGA companies to datacenters and away from traditional
programmable logic.

We need a good Friday thread, it's been a while. What do y'all think?

I think the article overstates the dichotomy of supporting the
traditional FPGA market and the datacenter market. Will FPGAs need to
vary so much to support one market vs. the other? They are hugely
programmable. Neither X nor A have pushed much on alternate
architectures that might be a significant advantage in a particular
market. I think it will mostly be steady as she goes with most of the
changes in marketing rather than engineering.

Not sure if it's the case in functionality, but I'm definitely seeing
shift in market focus. It used to be that I could go to any
of my distributors and have FPGAs that were effectively guaranteed to be
stocking; ship two days later. Now no one stocks, I have to keep a
fairly large buffer of parts on-site locally, and all of the interest
seems to be in making bulk sales to large customers rather than fanning
out to small ones. To what degree that's a function of the shifting
vision of the FPGA vendors or not, that's opaque to me. But it's
definitely real.

--
Rob Gaddi, Highland Technology -- www.highlandtechnology.com
Email address domain is currently out of order. See above to fix.