Breaking News
News & Analysis

Altera tips plans for 20, 14nm FPGAs in 2014

6/10/2013 12:00 PM EDT
13 comments
NO RATINGS
More Related Links
View Comments: Oldest First | Newest First | Threaded View
Page 1 / 2   >   >>
rick merritt
User Rank
Author
re: Altera tips plans for 20, 14nm FPGAs in 2014
rick merritt   6/10/2013 3:11:28 PM
NO RATINGS
My question to the EE community is whether the bump Altera will get with Intel 14nm for Stratix 10 is enough to compel you to switch to its high-end FPGAs?

chipmonk0
User Rank
Author
re: Altera tips plans for 20, 14nm FPGAs in 2014
chipmonk0   6/10/2013 3:40:13 PM
NO RATINGS
Would 3-D stacking of chips be really necessary for Altera at 14 nm from Intel - unless they decide to put SERDES on separate chips ? In that case who would stick / stack them together ? Intel / OSATs ?

Max The Magnificent
User Rank
Author
re: Altera tips plans for 20, 14nm FPGAs in 2014
Max The Magnificent   6/10/2013 4:49:24 PM
NO RATINGS
They're also talking about putting stuff like memory on separate die in the 3D stack/package

rick merritt
User Rank
Author
re: Altera tips plans for 20, 14nm FPGAs in 2014
rick merritt   6/10/2013 10:15:16 PM
NO RATINGS
Exactly, and those memory chips may not WANT 14nm costs.

Hyperdude
User Rank
Author
re: Altera tips plans for 20, 14nm FPGAs in 2014
Hyperdude   6/11/2013 4:06:21 AM
NO RATINGS
There was no processor mentioned for the 14nm part. Is that an SOC or plain vanilla FPGA? What happened to their current generation of products? It would be interesting to see how the 20nm part got its performance boost compared to the current gen SOC. I think their current gen ARM is ~1GHz and they claim 1.5GHz for the 20nm part. You have to do some custom layout to get 50% gain.

mranderson
User Rank
Author
re: Altera tips plans for 20, 14nm FPGAs in 2014
mranderson   6/11/2013 4:49:37 PM
NO RATINGS
I think Altera has put themselves into a really good position recently in terms of future hardware and their existing software. A lot of people are interested in these new parts despite the costs. It is much easier to design and implement a board with one fpga than to have to use multiple fpgas. Altera's main competitor is Xilinx, not these other companies. Altera using Intel fabs is a big deal.

chipmonk0
User Rank
Author
re: Altera tips plans for 20, 14nm FPGAs in 2014
chipmonk0   6/11/2013 4:59:20 PM
NO RATINGS
How much SRAM ( probably 14 nm won't hurt ) and DRAM ( 2x nm ? ) are they aiming for ? Any idea ?

Hermann C
User Rank
Author
re: Altera tips plans for 20, 14nm FPGAs in 2014
Hermann C   6/11/2013 5:19:37 PM
NO RATINGS
Altera talked about 2X performance gain or 70% lower power at same performance (vs 28nm). If true, sounds like they use Intel's uPC process for max performance boost. Won't be surprised finding Intel processor embedded later.

rick merritt
User Rank
Author
re: Altera tips plans for 20, 14nm FPGAs in 2014
rick merritt   6/12/2013 1:25:04 AM
NO RATINGS
They are not saying. Xilinx has been more vocal on this front http://confidential.eetimes.com/news-updates/4230167/Xilinx-Pioneers-Uncertain-Future-for-Chip-Stacking Although Huawei did talk about what it is doing with Altera http://www.eetimes.com/electronics-news/4401446/Huawei--Altera-mix-FPGA--memory-in-2-5-D-device

rick merritt
User Rank
Author
re: Altera tips plans for 20, 14nm FPGAs in 2014
rick merritt   6/12/2013 1:26:28 AM
NO RATINGS
Altera is not saying what processor it will embed at 14nm with Stratix 10. I am guessing it might be a 64-bit ARM.

Page 1 / 2   >   >>
Most Recent Comments
michigan0
 
SteveHarris0
 
realjjj
 
SteveHarris0
 
SteveHarris0
 
VicVat
 
Les_Slater
 
SSDWEM
 
witeken
Most Recent Messages
9/25/2016
4:48:30 PM
michigan0 Sang Kim First, 28nm bulk is in volume manufacturing for several years by the major semiconductor companies but not 28nm FDSOI today yet. Why not? Simply because unlike 28nm bulk the LDD(Lightly Doped Drain) to minimize hot carrier generation can't be implemented in 28nm FDSOI. Furthermore, hot carrier reliability becomes worse with scaling, That is the major reason why 28nm FDSOI is not manufacturable today and will not be. Second, how can you suppress the leakage currents from such ultra short 7nm due to the short channel effects? How thin SOI thickness is required to prevent punch-through of un-dopped 7nm FDSOI? Possibly less than 4nm. Depositing such an ultra thin film less then 4nm filum uniformly and reliably over 12" wafers at the manufacturing line is extremely difficult or not even manufacturable. If not manufacturable, the 7nm FDSOI debate is over!Third, what happens when hot carriers are generated near the drain at normal operation of 7nm FDSOI? Electrons go to the positively biased drain with no harm but where the holes to go? The holes can't go to the substrate because of the thin BOX layer. Some holes may become trapped at the BOX layer causing Vt shift. However, the vast majority of holes drift through the the un-dopped SOI channel toward the N+Source,...

Datasheets.com Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)
Like Us on Facebook
EE Times on Twitter
EE Times Twitter Feed