At the Advancing AI 2024 event in California, AMD unveiled its latest lineup of server processors, the 5th generation EPYC, codenamed “Turin.” This new release represents a significant advancement in the CPU market for data centers and servers, featuring models with up to 192 cores based on AMD’s Zen 5 and Zen 5c architectures. These processors are compatible with the SP5 socket, first introduced with the 4th generation EPYC (Genoa).
AMD’s new EPYC lineup includes 27 models, offering a range of core counts, clock speeds, and power consumption options. These processors combine high core counts with clock speeds reaching up to 5 GHz, aiming to set new benchmarks for efficiency and performance in large-scale computing.
The processors feature 4 nm lithography for Zen 5 cores, designed for raw performance and arranged in up to 16 Core Complex Dies (CCDs). Meanwhile, 3 nm lithography is used for Zen 5c cores, optimized for energy efficiency and organized into 12 CCDs, delivering a balance of performance and power efficiency.
What is the new 5th generation AMD EPYC like?
The 5th generation AMD EPYC processors utilize a modular chipset-based architecture, a design approach already proven successful in AMD’s Ryzen components for the consumer market. One standout example is the EPYC 9005, a 192-core, 384-thread processor, which boasts an impressive 150 billion transistors, making it one of the most complex chips ever created.
This new generation delivers a 25% increase in IPC (Instructions Per Cycle) over its predecessor, significantly enhancing performance in demanding workloads such as machine learning, virtualization, and databases. Additionally, these processors support DDR5-6000 memory across 12 memory channels, offering substantial bandwidth. They also integrate PCIe 5.0 compatibility, ensuring seamless connections with high-performance GPUs and high-speed storage devices.
AMD has strongly emphasized energy efficiency with configurations ranging from 125W to 500W TDP. Thanks to the optimized architecture and advanced manufacturing processes, these processors offer improved energy efficiency, making them an ideal choice for **high-density servers** that need to balance performance and power consumption.
5th generation AMD EPYC Turin | ||||||
Cores | CCD (Zen5/Zen5c) | Clock Base/Boost (GHz) | Standard TDP | L3 cache (MB) | Price (1 KU) | |
9965 | 192 | “Zen5c” | 2.25 / 3.7 | 500W | 384 | $14,813 |
9845 | 160 | “Zen5c” | 2.1 / 3.7 | 390W | 320 | $13,564 |
9825 | 144 | “Zen5c” | 2.2 / 3.7 | 390W | 384 | $13,006 |
9755 9745 | 128 | “Zen5” “Zen5c” | 2.7/4.1 2.4/3.7 | 500W 400W | 512 256 | $12,984 $12,141 |
9655 9655P 9645 | 96 | “Zen5” “Zen5” “Zen5c” | 2.6 / 4.5 2.6 / 4.5 2.3 / 3.7 | 400W 400W 320W | 384 384 384 | $11,852 $10,811 $11,048 |
9565 | 72 | “Zen5” | 3.15/4.3 | 400W | 384 | $10,486 |
9575F 9555 9555P 9535 | 64 | “Zen5” “Zen5” “Zen5” “Zen5” | 3.3 / 5.0 3.2 / 4.4 3.2 / 4.4 2.4/4.3 | 400W 360W 360W 300W | 256 256 256 256 | $11,791 $9,826 $7,983 $8,992 |
9475F 9455 9455P | 48 | “Zen5” “Zen5” “Zen5” | 3.65 / 4.8 3.15/4.4 3.15/4.4 | 400W 300W 300W | 256 192 192 | $7,592 $5,412 $4,819 |
9365 | 36 | “Zen5” | 3.4/4.3 | 300W | 256 | $4,341 |
9375F 9355 9355P 9335 | 32 | “Zen5” “Zen5” “Zen5” “Zen5” | 3.8 / 4.8 3.55 / 4.4 3.55 / 4.4 3.0/4.4 | 320W 280W 280W 210W | 256 256 256 256 | $5,306 $3,694 $2,998 $3,178 |
9275F 9255 | 24 | “Zen5” “Zen5” | 4.1 / 4.8 3.25 / 4.3 | 320W 200W | 256 128 | $3,439 $2,495 |
9175F 9135 9115 | 16 | “Zen5” “Zen5” “Zen5” | 4.2 / 5.0 3.65 / 4.3 2.6/4.1 | 320W 200W 125W | 512 64 64 | $4,256 $1,214 US$726 |
9015 | 8 | “Zen5” | 3.6/4.1 | 125W | 64 | $527 |
AMD EPYC Turin x Intel Xeon 6
When comparing AMD’s EPYC Turin processors to Intel’s Xeon 6 series, it’s clear that AMD has taken the lead in several critical areas, especially in workloads requiring high parallelism. The EPYC Turin chips, with configurations that reach up to 192 cores, offer significantly higher core density than the Xeon 6, which tops out at 144 cores. This core advantage directly translates into superior performance for tasks such as data center processing, artificial intelligence, and scientific computing.
In early benchmarks, AMD demonstrated that EPYC Turin processors can deliver up to twice the performance of Intel’s Xeon 6, particularly HPC (High-Performance Computing) applications. This makes the EPYC chips particularly effective for highly parallelized workloads.
Additionally, the native DDR5 memory support and 12 memory channels in EPYC Turin processors allow for higher memory bandwidth, vital for big data and real-time analytics environments. Intel’s Xeon 6, while competent, lags in this aspect, offering fewer memory channels and lower overall bandwidth.
Moreover, cost-per-performance is another area where EPYC Turin shines. While Intel’s Xeon 6 series performs well in traditional workloads, AMD’s EPYC processors provide a more powerful solution for modern, scalable workloads.
They manage to maintain energy efficiency while still delivering top-tier performance, giving data centers and enterprise users a more attractive option for handling intensive computing tasks. This balance of performance, core density, and efficiency makes EPYC Turin a compelling choice over its Xeon competition.
What can you expect from the new 5th-generation AMD EPYC?
AMD’s new 5th-generation EPYC processors are set to disrupt the server market with their greater core density, cutting-edge technology support, and improved energy efficiency. These CPUs promise an outstanding cost-performance ratio, making them ideal for businesses aiming to boost performance without increasing power consumption.
A significant advantage is their compatibility with the SP5 socket used by the previous EPYC generation. Data centers can upgrade their systems without overhauling their entire infrastructure, leading to significant cost savings. This strategic move by AMD emphasizes its commitment to offering scalable solutions that can be implemented quickly and efficiently.
These advancements are expected to expand AMD’s presence in the server market, positioning the EPYC Turin processors as a powerful and energy-efficient alternative to Intel, especially for modern workloads that demand high processing power and architectural flexibility.