Today’s compute-intensive workloads—like AI, HPC, analytics, and more—hunger for performance. And often, that means trading off affordability for speed. But not anymore.
Now, there are two great choices for your compute-intensive and AI workloads on AWS: Amazon Elastic Compute Cloud (EC2) C7i and the new C7i-flex custom instances.
Powered by custom 4th Gen Intel® Xeon® Scalable processors—an architecture you already know and trust—these instances enable you to tap into Intel® Accelerator Engines across AWS’s expansive global footprint. They offer outstanding performance and price performance benefits, with C7i-flex instances delivering 19% improved price performance and C7i delivering 15% improved price performance over C6i instances.(1)
C7i-flex instances provide an extra cost benefit when running a majority of compute-intensive workloads that don’t always require continuous high CPU usage, whereas C7i instances give you full compute performance 100% of the time. More on the differences between C7i and C7i-flex here.
The Market—Accelerating Business Outcomes
Today’s high-performance computing and AI landscape is growing rapidly—with worldwide AI spending expected to nearly double from 2023 to 2027.(2) And a massive proportion of that growth is in AI training and inferencing.
As vendors, enterprises, and cloud providers struggle with absorbing the fast cadence of proprietary GPUs, many are looking for an alternative—ideally, one built on platforms that are flexible, open, and readily available to meet their performance and cost-efficiency needs for compute-intensive applications. And that’s exactly what Intel’s innovative accelerator engines enable—an open, highly-performant, cost-effective platform for today’s popular AI and analytics workloads.
For Developers: Open, Flexible, Cost-Efficient Performance That’s Ready for the Future
Intel’s software-first strategy is designed to accelerate both the application development process and runtime performance of your applications.
How? It starts with Intel® Software Development Products—a set of tools, libraries, and frameworks powered by the oneAPI open programming model, with performance optimizations for Intel’s hardware portfolio, which includes the Intel Accelerator Engines featured in C7i-flex and C7i instances.
As an open standard, oneAPI enables you to develop code that runs across architectures and vendors, so your code can run on the latest CPUs and other compute technologies today and tomorrow.
Hardware Acceleration Built-In
Because the new C7i-flex and C7i instances are powered by 4th Gen Intel Xeon processors, developers can use oneAPI to exploit the various Intel Accelerator Engines built into the CPU to boost the performance of modern, compute- and data-intensive workloads.
These accelerators include:(3)
- Intel® Advanced Matrix Extensions (Intel® AMX)—designed to accelerate common AI inferencing and training tasks.
- Intel® Advanced Vector Extensions 512 (Intel® AVX-512)—designed to accelerate vectorization and highly parallel computations common in HPC, AI, advanced simulations, and more.
- Intel® Data Streaming Accelerator (Intel® DSA)—designed to offload the most common data movement tasks that create overhead and penalize performance when moving data between CPUs, memory, and caches, as well as attached memory, storage, and network devices.
- Intel® In-Memory Analytics Accelerator (Intel® IAA)—designed to accelerate database and analytics workloads.
- Intel® QuickAssist Technology (Intel® QAT)—designed to offload encryption, decryption, compression, and other common data processing tasks.
Beyond Faster Results—Performance for Sustainability and TCO
C7i-flex and C7i instances can also reduce energy consumption and emissions, thanks to 4th Gen Xeon processors’ ability to complete compute-intensive workloads faster. The comparisons below show the benefits of 4th Gen Xeon processors versus prior-generation CPUs and are not related to specific cloud instances.
Compared to prior generation CPUs, 4th Gen Xeon processors can provide 10x AI inferencing and training performance,(4,5) 3x data analytics performance,(6) and 2x faster compression with 95% fewer cores.(7)
That performance improvement can provide a 2.9x average performance per watt efficiency improvement for select workloads when utilizing the built-in accelerators,8 and up to a 70-watt power savings per CPU in optimized power mode, with minimal performance loss.9 And, it can reduce the total cost of ownership by 52% to 66%.(10,11)
Conclusion
C7i and C7i-flex, the newest Amazon EC2 instance families featuring 4th Gen Intel Xeon Scalable processors, are two great choices for compute-intensive and AI workloads on AWS. By hosting your workloads on these new instances, you can take advantage of increased compute performance and flexible pricing efficiency.
- Performance and price performance claims are made by AWS and based on AWS’s internal testing.
- Source: https://www.intel.sg/content/dam/www/central-libraries/xa/en/documents/2024-04/idc-infobrief-asia-pacific-ai-maturity-study-2024-singapore.pdf
- Intel QAT, Intel IAA, and Intel DSA accelerators are only available on the C7i bare metal instances. Intel AMX is available on both C7i and C7i-flex.
- See [A17] at https://edc.intel.com/content/www/us/en/products/performance/benchmarks/4th-generation-intel-xeon-scalable-processors/. Results may vary.
- See [A16] at https://edc.intel.com/content/www/us/en/products/performance/benchmarks/4th-generation-intel-xeon-scalable-processors/. Results may vary.
- See [D1] at https://edc.intel.com/content/www/us/en/products/performance/benchmarks/4th-generation-intel-xeon-scalable-processors/. Results may vary.
- See [N16] at https://edc.intel.com/content/www/us/en/products/performance/benchmarks/4th-generation-intel-xeon-scalable-processors/. Results may vary.
- See [E1] at https://edc.intel.com/content/www/us/en/products/performance/benchmarks/4th-generation-intel-xeon-scalable-processors/. Results may vary.
- See [E6] at https://edc.intel.com/content/www/us/en/products/performance/benchmarks/4th-generation-intel-xeon-scalable-processors/. Results may vary.
- See [E8] at https://edc.intel.com/content/www/us/en/products/performance/benchmarks/4th-generation-intel-xeon-scalable-processors/. Results may vary.
- See [E9] at https://edc.intel.com/content/www/us/en/products/performance/benchmarks/4th-generation-intel-xeon-scalable-processors/. Results may vary.
Notices and Disclaimers
Performance varies by use, configuration, and other factors. Learn more on the Performance Index site.
Performance results are based on testing as of dates shown in configurations and may not reflect all publicly available updates. See backup for configuration details. No product or component can be absolutely secure.
Your costs and results may vary.
Intel technologies may require enabled hardware, software, or service activation.
© Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.
Josh Hilliker is a Senior Cloud Solutions Architect in Data Center Sales focused on advancing our Autonomous Data Center solutions, including telemetry and infrastructure modernization. Josh has held many past roles at Intel including the following: Director in SMG University, Platform Marketing in Business Client Platform Division, Intel GE-Care Innovations as the Director of Social Media, Online Presence and Google Analytics for consumer products, Product Engineer with the IT Land Capital and Acquisitions team, Operations manager for Intel e-Business Infrastructure & OEM Business Link, Architect / Community manager for Intel® vPro ™ Expert Center, Intel Manufacturing, and Director of Small Business initiatives in Intel Architecture Group.