Advertisement

Zibb

Loring WirbelAnalyst Loring Wirbel covers programmable logic from an application perspective, providing a sneak peek at the vertical applications that help drive FPGA complexity, performance, and density. The blog will feature videos allowing engineers to spotlight their latest designs, along with news of products and corporate trends at FPGA vendors and the developers of third-party tools for programmable logic.



   Advertisement

Profile

RSS Feed

  • Add this blog to your RSS newsreader!

Recent Posts

Recent Comments

Most Commented On

Archives

By Category

Blog

Monday, August 3, 2009

Gigabit to the Masses

Aug 3 2009 8:56AM | Permalink |Comments (2) |


Consider this a sequel of sorts to the last post on Virtex and HyperTransport. When Altera Corp. launched its Arria II GX development system July 27, one aspect of the introduction failed to gain much comment or attention, perhaps because it’s considered commonplace. These days, transceivers on mid-range FPGAs are assumed to support multi-gigabit-per-second speeds, and the battle for high-speed backplanes is no longer relegated to Virtex vs. Stratix. The Arria platform was specifically intended to meet PCI Express 2.0 specs.

Did I hear someone say “So what?” If you’re designing virtually any IT or embedded industrial system where PCIe plays a role, it should matter very much. MAC and physical-layer characteristics of a vendor’s transceivers will play an increasingly important role in all systems currently entering design, including those for which entry-level FPGAs are still being considered. Yes, there are plenty of designs out there still based on CAN or 10/100 Ethernet, and transceivers supporting 3.75-Gbit speeds may seem like overhill.

But within the next year or two, 1-Gbit speeds will be accepted as table stakes for virtually any design in any field. FPGA vendors are wise to bring support for higher speeds down to every product class. Meanwhile, the Virtex-Stratix war can move to 10 Gbits/sec, as OEMs assure me the long-awaited commoditization of 10-Gbit Ethernet is within 18 months of realization. Hey, wait, didn’t they say that five years ago?

 


Related entries in: Communication functions | Computers, boards, buses | FPGA Gurus | Programmable Logic | 


Reader Comments



at 8/3/2009 2:37:15 PM, desert rat said:
Well, we already have PCIe 3.0 silicon (at 8 gig, although it is PAM and not 8b10b). So, the FPGA guys are a couple of generations behind already.....with 3.75 Gig. One Gig ethernet is dead as a doornail. 2.5 Gig PCIe is dead too (both from a design-in standpoint). The apps that can live with 10/100 E or CAN are pretty boring niches with low profits and low volumes. They are the bottom-feeders of the industry and are hardly palatable, even with a lot of garlic and butter.



at 8/5/2009 8:48:18 AM, Loring said:
I agree with everything you say, except that krill feeds the food chain, and slowpoke industrial nets do too, to a certain extent. As for where PCIe is going in real-time acquisition, check out the brand new post!

Post a comment



Display Name

Change Image
Before submitting this form, please type the characters displayed above.
Note the letters are NOT case sensitive.


ADVERTISEMENT

©1997-2010 Reed Business Information, a division of Reed Elsevier Inc. All rights reserved.
Use of this Web site is subject to its Terms of Use | Privacy Policy