Aug 29, 2024
Bringing the Speed of Thought to Cyber Defense
Josh Patterson
Since joining Accenture Cybersecurity Tech Lab, I’ve been hyper-focused on balancing the cost of attacks and defense. Lowering the cost of accelerated defense has been the goal…and a long journey. I’m proud to say that we’re further along than ever due to the hard work of many open source developers, cyber practitioners, and Voltron Data engineers. Composable open standards and accelerator native compute engines like Theseus paired with modern defense is a massive step forward. But, this is a long mission, and we can’t do it alone…
Today, we’re thrilled to announce the formation of the Voltron Data Cyber Advisory Board. Joining the board are individuals who have led groundbreaking cybersecurity initiatives across industries. We’re dedicated to pioneering a differentiated, composable, accelerated analytics solution in collaboration with industry leaders who have a strong vision for the future of cybersecurity.
This is a strategic step forward in redefining data systems for cybersecurity.
AI + Cloud Have Trivialized the Cost of Attacks
Why now? Well, it has become apparent that the “cost of attack” is continually decreasing.
The proliferation of cloud infrastructure and resources has made attacks significantly easier and less expensive, drastically increasing the attack frequency. The explosion of AI exacerbates the problem, increasing the volume, variety, and veracity of attacks, ultimately making cyber-attacks more effective, frequent, and unique. AI + Cloud has made attacking effectively trivial. We can’t keep building higher walls — it’s cheaper to build a ladder.
Bridging Cybersecurity Thought and Composable Data Systems
At Voltron Data, we believe that cybersecurity is fundamentally a data analytics challenge. As cyber threats grow in complexity and scale, and systems become more widely distributed (massive attack surface), it is increasingly difficult to detect threats. The need for composability is becoming more evident. Building on open standards brings high-performance, scalable, cost-effective, and energy-efficient capabilities to cyber data systems – at a fraction of the engineering cost.
Composability isn’t just for fun, it’s essential to bridge all the existing knowledge and acumen developed over the years in the cyber defense industry. To make composable accelerated solutions a reality, we leverage our decades of expertise in open source standards and developing software specifically for hardware accelerators like NVIDIA GPUs. This allows us to design and build solutions that are not only state-of-the-art and powerful but also power-efficient.
Cybersecurity Needs Cost (and Energy) Efficient Solutions
A key theme in my experience is that cyber intelligence and threat detection technology must be fast – at all costs.
Sadly, I too often hear about organizations making compromises on which data they can retain and analyze to meet their cyber security obligations due to current technology limitations. This means organizations are not seeing the full scope of cyber security risks – and nothing is more risky than that. Defenders know what they want to analyze, they are just limited by the tools that they have!
In the AI age – running 100TB queries on Petabyte+ datasets is rapidly becoming standard (if you’re not there yet – you more than likely will be). Going back to my earlier point: your security technology must be fast at all costs. Working with infrastructure designed for a pre-AI age drives unnecessary costs, slows performance, and impacts the environment. Defenders are capped on what they can do today.
We are building analytics solutions that deliver innovation without compromise or loss of existing knowledge: go faster and go greener. This is not a pipe dream - this is happening in our customers’ data pipelines today. With our current architecture using NVIDIA A100 GPUs, large-scale data analytics consume about 80% less energy — a fifth of the energy. We’re working on pushing that to 95% energy reduction; only 5% of the energy used — a 20th of the energy.
Our solutions are proven to reduce energy and cost:
-
Streamline ETL Pipelines: Our GPU query engine, Theseus, can analyze petabytes of log data, machine data, and large amounts of tabular data generated on the fly - no caching, indexing, or warm-up. Performance is so fast that use cases like analysis of TBs of raw log data to real-time anomaly detection are now possible in seconds thanks to our GPU Query engine that’s 60x faster than CPU systems. This saves a significant amount of time and resources and allows defenders to start exploring new techniques for threat detection and AI augmentation. Read our benchmarking report to learn more.
-
Power Data Connectivity: Through our open source engineering and enterprise support services, we help organizations reduce data silos and increase data movement and reuse of code with open standards. This means less time and energy is spent moving data or rewriting rules. Our customers’ engineering teams are scaling from local to distributed using the programming languages and tools of their choice. Code in SQL, Python, R, Java, Go, and more, and seamlessly operate on the same data inside whatever engine you choose.
How can a composable architecture help your cyber strategy?
Our solutions will enable real-time and batch data processing to tackle critical cyber use cases like anomaly detection, log analysis, user behavior analytics, threat detection, and more.
If these are use cases that matter to you and your current system isn’t meeting your SLAs or is too expensive, reach out to us: voltrondata.com/contact.
Or, download our Cybersecurity Solution Brief to see how we can help you.