Is VPX the Right Choice? Challenges, Considerations, and a Common Vision
by Michael Bowling, on Nov 9, 2021 10:30:00 AM
VPX-aligned configurations have their advantages.
Their intrinsically rugged design provides effective shock and vibration resistance; their pin-socket style offers designers quick, simple, and easy customizability and serviceability; and their 3U and 6U form factors are suitable acquisitions for numerous 3U and 6U embedded and rackmount deployments.
Certainly, for customers deploying 3U and 6U high- performance, platform-based single-board computers (SBCs) frequently exposed to harsh conditions and emphasizing high MTBF, low MTTR, and vendor interoperability, VPX remains a viable technology for HPC solutions in many cases.
But is there a more cost-effective, more SWaP-optimized, and more easily implemented option for similar deployments, programs, and applications in the same battlespace and with similar performance requirements?
Furthermore, could the cost, SWaP, and implementation challenges associated with VPX be eliminated or mitigated by incorporating such a solution into an already VPX- deployed battlespace?
And, finally, how, in a tactical theater where HPC diversity is the norm, can both types of solutions co-exist to help militaries everywhere connect, coordinate, and deploy new and advanced personnel, assets, and equipment?
This whitepaper explores these queries, advances a suitable alternative for customers seeking many of the same benefits of VPX solutions but without the cost, SWaP, and implementation challenges, and discusses how we at Trenton Systems, through our deployment of small-form-factor (SFF) mission computers, can support the armed forces as well as VPX solutions providersin facilitating a diverse, collaborative technological battlespace that keeps servicemembers empowered, no matter what.
Although a fitting choice for many programs and applications, VPX solutions can fall short regarding cost, SWaP, and implementation.
And for customers specifically overseeing legacy military platforms, which represent most of the U.S. military‘s computational infrastructure, hyperfocused VPX alignment initiatives may also overlook or altogether ignore the resource and training challenges associated with VPX modernization.
We address VPX‘s cost, SWaP, and implementation challenges individually in the sections below.
Where It Hurts: Cost Challenges
An inherent challenge associated with VPX implementation is the customer‘s need to acquire non- commercial plugin cards compatible with a VPX layout.
For example, purchasing or asking your HPC manufacturer to provide commercial off-the-shelf GPUs sourced directly from NVIDIA tends to cost significantly less than acquiring custom-made GPUs utilizing NVIDIA technology from another.
Highly custom, these VPX-aligned graphics cards can be anywhere from two to three times more expensive than standard graphics cards sourced from commercial providers.
Not to mention, our example covers VPX-aligned GPUs only. Add to those costs the costs of multiple VPX-aligned expansion cards with capabilities that many defense customers need, and you start fo feel the budgetary pressure.
Customers using and sourcing more widely supported configurations have the option of incorporating more affordable, commercially sourced option cards while satisfying the same performance requirements, saving thousands on expansion capabilities in the process.
Not the Right Fit: SWaP Challenges
Another inherent challenge associated with VPX implementation is VPX‘s 3U and 6U height requirements.
Many military platforms simply can‘t support the height demands of 3U and 6U VPX chassis. Some customers may be limited in space by existing configurations, while, for others, 3U and 6U rack heights are simply not feasible because of their platforms‘ preordained space constraints.
Some military platforms also have stringent low-weight requirements or may already be approaching the point of the proverbial broken camel‘s back, making implementing 3U and 6U VPX chassis, which weigh, fully equipped, anywhere between 18 and 50 pounds, nonviable at best and hazardous to mission-critical platforms at worst.
There‘s also the cooling consideration.
VPX solutions generally use direct air cooling, conduction cooling, or liquid flowthrough with quick disconnects; however, natural convection cooling, also known as passive cooling, despite this method‘s simplicity, low cost, high MTBF, and ease of implementation - all clear benefits to the customer - is not ideal for high-performance 3U and 6U VPX configurations unless the system‘s ambient environment has low to moderate temperatures.
It‘s not that there isn‘t a place for the SWaP characteristics of 3U and 6U VPX configurations - there certainly is, and we discussed some of their inherent benefits at the beginning of this paper - but there exist more SWaP-optimized options for high-performance computing deployments than what VPX has to offer.
No Legacy Platform Left Behind: Implementation Challenges
The vast majority of military platforms in use today are legacy and unaligned with VPX. They use long-trusted, industry-leading standards and configurations that utilize affordable, commercially sourced components and provide the benefits that military customers with strict program requirements and budgetary constraints often seek: cost-effectiveness, reliability, as well as ease and rapidity of implementation.
Not only can implementing VPX-aligned configurations into legacy platforms be costly, but it can also be stressful, time-consuming, or impractical for legacy platform customers with SWaP, time, monetary, resource, and training constraints, and, in many cases, the benefits offered by VPX may be unnecessary to meet the needs of their program or application.
Thus, customers looking for a similar but more cost- effective, SWaP-optimized, or easily implemented computing solution will need to look elsewhere.
The TAC Family: An Alternative for Small & Legacy Military Platforms
Although VPX solutions are suitable for numerous military land and airborne platforms, they‘re not designed for small and legacy platforms with budgetary, SWaP, and implementation limitations, including small uncrewed aerial and ground vehicles.
The vast majority of these military platforms would be hard- pressed to support a VPX solution, and thus their leaders should turn to more feasible, practical options that meet their budgetary, SWaP, and implementation requirements for LRU obsolescence.
Customers in charge of small and legacy C6ISR platforms can look toward high-performance embedded mission computers utilizing more common, more SWaP-optimized form factors, such as COM Express Type 7.
One option is Trenton Systems‘ Tactical Advanced Computer (TAC) family, a line of SFF edge computers designed for military and aerospace platforms dealing with LRU upgrades and obsolescence.
Certified to military and industrial standards, SWaP-C optimized, fanless, sealed, weighing less than 5 pounds, and capable of surviving temperatures ranging from -51°C to +85°C, TACs are ultra-rugged, resilient, light, and designed specifically for tactical edge deployment.
They‘re also built using next-generation Intel CPUs, can support GPGPUs via MXM, and benefit from multiple data-at-rest security enhancements, including encrypted storage.
L3Harris Technologies recently acquired a multimillion- dollar provision of fully custom TACs for a classified United States Department of Defense (DoD) aircraft program. From the air, these designed-to-meet TACs will provide data communications support to ground forces, bridging the gap between land-based server operations and airborne, edge-deployed embedded computing.
Other key decision-makers across military primes like L3Harris as well as the U.S. Armed Forces continue to approach Trenton Systems with requests for SFF
mission computers like the TAC that can support, in harsh environments, distributed edge computing and signals processing for air, land, sea, and space C6ISR, electronic warfare (EW), situational awareness, and weapons targeting systems and applications.
Other key decision-makers across military primes like L3Harris as well as the U.S. Armed Forces continue to approach Trenton Systems with requests for SFF.
A Diverse, Collaborative Technological Battlespace with One Common Goal
For platform designers incorporating high-performance compute clusters in environments where the challenges listed in this paper are less of a concern, VPX embedded systems may be the ideal choice.
But for the legion of legacy military platform leaders in charge of certain small vehicles and systems, and for others to whom cost-effectiveness, practical SWaP, and ease and rapidity implementation are today‘s foremost goals, a commercially supported mission computer from the TAC family, or a similarly supported SFF solution, may be worth considering.
Regardless, Trenton Systems aligns with the spirit of VPX and SOSA and is committed to supporting and providing complementary high-compute solutions to defense and aerospace platforms with existing VPX infrastructure.
Although the deployment of VPX-aligned solutions won‘t always make practical sense regarding cost, SWaP, and implementation, these solutions have their place and distinct advantages, especially regarding vendor interoperability, ruggedness, serviceability, and the other benefits mentioned in the foreword of this paper.
At Trenton Systems, we envision a diverse, collaborative technological battlespace, a large-scale, interconnected defense and aerospace ecosystem of high-compute technologies and capabilities that leverage both VPX-aligned and more widely supported configurations.
Regardless of new and emerging technologies and changing trends, this ideal battlespace never loses sight of the most important challenge of all: keeping people and servicemembers everywhere safe from harm.