Sequans Communications (United Kingdom)

With AI model size doubling every four months, the size of AI compute clusters is growing rapidly to keep up. Traditional data center front-end networks rely on optical interconnect for data movement with Terabits per second of connectivity and have been doubling data rate every 4 years. As the models grow it is also expected that the back-end GPU networks expand beyond the confines of the rack. Optical interconnects will be needed at distances of tens of meters with 10Tb/s data transfer rates and doubling every 2 years will be expected. How should optical interconnects best achieve this bandwidth increase and what is the best performance axis to push next? In this panel discussion we pose the question -"Is it better to scale with more wavelengths or by adding more fibers?" Join us to see if any of the panelists can convince you they have the right approach.

<p>This deliverable presents the work executed in the iNGENIOUS project with regards to dissemination, communication, standardization, exploitation, and innovation.</p>

Purpose – This paper aims to explore how emotional intelligence and the emerging field of neuroscience can tell us much about how leaders can use their own emotions to positively influence people. The ability for leaders to positively influence others is a critical skill and yet it is often identified as a gap in leadership teams. Design/methodology/approach – The paper draws on the work of Goleman, Kline, Ramachandran and Schwartz, as well as the author’s own recently published book. Findings – An emotionally intelligent leader, rather than assuming that those who speak the loudest have the most valuable contribution will be more likely to understand and value different thinking and personality styles and will take active steps to create an environment where people feel safe to contribute their very best thinking. Practical implications – This paper provides useful, practical advice to those involved in leadership development or leaders themselves who want to better understand how they can improve their ability to influence others. Originality/value – The paper provides practical advice, grounded in research, that professional learning and development executives or leaders themselves can put in action.

This paper builds on Goldbach’s weak conjecture, showing that all primes to infinity are composed of 3 smaller primes, suggesting that the modern definition of a prime number may be incomplete, requiring revision. The results indicate that prime numbers should include 1 as a prime number and 2 as a composite number, adding a new dimension to the most fundamental of all integers.

In this paper, the truncated singular value decomposition expectation-maximization (TSVD-EM) based channel estimation on single transmit antenna transmission is extended to transmit diversity transmission. Particular attention is given to Long Term Evolution (LTE) systems. In order to obtain the received symbols for each transmit antenna, an approximation is proposed by using the soft symbols and channel estimates from the previous iteration. We prove that the mean square error (MSE) of TSVD-EM on transmit diversity transmission has almost the same constant value at high signal-to-noise ratio (SNR) as on single transmitter antenna case. Simulation results show that the proposed TSVD-EM channel estimation improves system performances remarkably and approaches the performance with perfect channel state information. Even without pilot boosting used in LTE systems, the TSVD-EM achieves the same performances as with pilot boosting.

Multiple-input multiple-output (MIMO) techniques have become an essential part of broadband wireless communications systems. For example, the recently developed IEEE 802.16e specifications for broadband mobile wireless access include three MIMO profiles employing 2times2 space-time codes (STCs) and two of those, namely, Alamouti's STC for transmit diversity, and the 2times2 spatial multiplexing (SM) scheme, are mandatory on the downlink of Mobile WiMAX systems. The first of these has full diversity but it is only half rate, while the second is full rate but suffers from diversity loss. In this paper, we develop a full-diversity STC of size 2times2, which represents an interesting trade-off between these two schemes. More specifically, the presented code is of rate 3/4, it has full diversity, and it leads to simple maximum-likelihood (ML) detection whose complexity grows only linearly with the size of the signal constellation. This makes it a suitable candidate for future evolutions of broadband wireless systems toward higher data rates and performance.