It was a rumor and it was finally confirmed. Those who said Intel wasn’t going to release more Atom because it created the Alder Lake architecture with heterogeneous cores for it were wrong. And also by far, since it presented three different series centered on three almost opposite strategic sectors. These are the Atom C5000, Atom P5300 and Atom P5700.
As the title suggests, we are talking about three series of totally different processors in their concept, which logically share many things. The architecture of these processors is called Parker Ridge and what we do know is that they will arrive fabricated at 10nm with a very interesting feature set.
Intel Atom C5000, optimized for networking and storage
This will be the most basic series and will consist of two well-defined product lines. The C5100 series is for storage, while the C5300 series will be for networking, where they will both have 4 to 8 cores, Intel QAT Gen 2, and TDPs of up to 50 watts:
- Intel Atom C5100: supporting local physical storage device storage applications in the small and medium business (SMB) segment. They have a TDP of 42 to 50 watts, no built-in Ethernet, and will operate over a commercial temperature range.
- Intel Atom C5300: Focused for use in network applications, they have a TDP of 32 to 41 watts and include extended and commercial temperature options, allowing you to choose your processor SKU based on your deployment context.
Intel Atom P5300: the pinnacle of optimized systems
Here the bet is raised, since we are talking about neither more nor less than configurations of 8 to 24 cores with TDPs ranging from 55 watts to 85 watts, much more serious bets. They focus on high performance and secure network devices, POPs or similar critical infrastructure systems. In addition, they have QAT Gen 3, being the most advanced in this key aspect for these processors.
What is curious about them is that Intel did not show any processor, nor names, nomenclatures, frequencies… Nothing more than what was mentioned, so we will have to wait to see the models later .
Intel Atom P5700: for embedded switches
This is the average range of all new Atoms, but due to its tasks and the importance of these, it is classified as the highest range. Intended for embedded switches and 8 to 20 core configurations with power consumption from 48 watts to 75 watts, here we have a series that takes advantage of QAT Gen 3 to support IPsec Online. What Intel is looking for here is a series of processors for high-volume cryptography and wide real-time data paths.
They also didn’t showcase any models from this series, but they did show all the new features that the C5000, such as the P5300 and P5700, will include, which is also interesting considering what’s new.
Brand new Atom C5000 P5300 P5700 series, a different approach
Intel has forgotten the previous orientation of its underperforming Atom and will take advantage of the name and popularity of the series to, as we have seen, focus on new markets where other specific technologies are required that would not not useful in PCs and servers.
Therefore, all new technologies or versions thereof are included:
- Intel Dynamic Load Balancer (Intel DLB) handles millions of queuing operations per second to efficiently distribute network processing across all processor cores. Managing network queues in hardware frees up CPU cycles typically consumed by software queues.
- Intel Ethernet 800 Series Technology: on-board system-on-chip (SoC) delivers up to 100 Gbps of capacity with link options from 1 GbE to 100 GbE, as well as support for advanced traffic management features on all packet types .
- Intel Network Acceleration Complex (NAC) enables high-performance Ethernet I/O, switching with accelerated inline packet processing, and an enhanced scheduler for higher performance with security features such as authentication and encryption/decryption.
- Intel QAT accelerates symmetric and asymmetric encryption, as well as lossless compression in hardware to free up CPU resources that can be used for other tasks.
- Intel Quick Data Technology allows the chipset to copy data instead of the CPU to move data more efficiently through the server, delivering faster, more scalable, and more reliable performance.
As expected, there is no information on arrival time or prices.