AMD has officially launched its EPYC Venice CPU family, spanning several chips, with performance leadership in Agentic AI with Zen 6 cores.
The Industry's First 2nm HPC CPU Is Here, Meet EPYC Venice Family, Packing Up To 256 "Zen 6" Cores & 203 Billion Transistors
One component that is really shaping up as the king of the Agentic AI era is the CPU. AMD is leveraging its brand new Zen 6 core architecture that will be used on its 6th Gen EPYC CPUs, codenamed Venice. AMD's EPYC Venice chips are the first HPC product to enter volume production on TSMC's 2nm process technology, and their readiness for Helios AI racks proves that AMD will always deliver the best of the best for its high-performance customers.
The TSMC 2nm process technology transitions from FinFET to Nanosheet transistors (GAA), and offers 10-15% higher performance at the same power, 25-30% lower power consumption at the same performance, and up to 15% higher transistor density. The EPYC Venice CPUs pack up to an insane 203 Billion Transistors while leveraging the 2nm process for the compute chiplet, and 6nm for the IO dies.
AMD's 6th Gen EPYC Venice CPUs were teased earlier this year at CES, offering up to 256 cores and 512 threads with eight massive compute dies and two even larger I/O dies. For this chip, AMD has promised over 80% performance & efficiency improvement with its EPYC Venice CPUs, with a >30% increase in thread density.
Once again, the requirement for high-performance CPUs has risen tenfold with Agentic AI workloads. And competitors are firing across all cylinders with CPUs that are optimized around agents. As such, AMD is introducing a full top-to-bottom CPU stack in its EPYC 9006 family, which includes chips for both SP7 and SP8 platforms. The full list of AMD EPYC 9006 family includes:
- AMD EPYC 9006 SP7: Built for high-density agent sandbox execution with up to 256 cores and 512 threads, it helps run more agents per watt, per dollar, and per rack. In high-frequency configurations, it also serves as an AI host node, with up to 5GHz frequency, PCIe® Gen 6, and impressive CPU-to-GPU improvement delivering the most agents per watt, per dollar, and per rack.
- AMD EPYC 9006 SP8: The processor is built for efficient, right-sized performance across critical enterprise workloads and agent sandbox execution where efficiency is top priority. With flexible 8- to 128-core options and low-power designs, it supports everything from edge sites and power-constrained racks to smaller clusters and general-purpose servers engineered to deliver leadership performance per system dollar.
- AMD EPYC 9006X SP7: The processor is built for HPC, technical computing, and data-intensive workloads that benefit from high cache and memory bandwidth. With some variants reaching frequencies up to 5.15GHz frequencies and 3x the L3 cache per core of AMD EPYC™ 9006 SP7, it helps accelerate simulation, modeling, large-scale analytics, retrieval, in-memory analytics, and other memory-sensitive work.
- AMD EPYC 9006 LP: Formerly codenamed “Verano,” the AMD EPYC 9006 LP processor is purpose-built as an AI host node CPU for next-generation rack-scale AI systems. The processor, leveraging LPDDR memory and up to 5GHz frequency, helps keep accelerators fed across dense, accelerator-rich racks, allowing system utilization to scale as deployments grow from single nodes to AI factories.
EPYC 9006 "SP7" Server CPU - Shipping Q4 2026
The first chip is the flagship with 256 "Zen 6" cores, 512 threads, a maximum bandwidth of 1.6 TB/s, and PCIe Gen6 support for up to 64 Gbps AIC speeds. The chip is designed to achieve very fast frequencies of up to 5.0 GHz with 96 operational cores.
| Name | # of CPU Cores | # of Threads | Max. Boost Clock | Base Clock | L3 Cache | Default CPU Power |
|---|---|---|---|---|---|---|
| AMD EPYC 9996 | 256 | 512 | Up to 4.1 GHz | 2.55 GHz | 1024 MB | 600W |
| AMD EPYC 9966 | 192 | 384 | Up to 4 GHz | 2.9 GHz | 768 MB | 600W |
| AMD EPYC 9846 | 168 | 336 | Up to 3.7 GHz | 2.85 GHz | 768 MB | 500W |
| AMD EPYC 9756 | 128 | 256 | Up to 4 GHz | 3.15 GHz | 512 MB | 500W |
| AMD EPYC 9G76 | 96 | 192 | Up to 4.8 GHz | 3.4 GHz | 384 MB | 500W |
| AMD EPYC 9656 | 96 | 192 | Up to 3.7 GHz | 3.05 GHz | 512 MB | 400W |
| AMD EPYC 9686F | 96 | 192 | Up to 5 GHz | 3.4 GHz | 384 MB | 500W |
| AMD EPYC 9556 | 64 | 128 | Up to 4.3 GHz | 2.75 GHz | 384 MB | 300W |
| AMD EPYC 9586F | 64 | 128 | Up to 5 GHz | 3.75 GHz | 384 MB | 500W |
EPYC 9006 "SP8" Server CPU - Shipping 1H 2027
The SP8 Venice "EPYC 9006" chips are designed for mainstream servers with 8 to 128 core configs. The chips will support 8 DDR5 channels with two DIMMs per Channel support, offering 128 PCIe Gen 6.0 lanes, and feature flexible deployments in 1P and 2P solutions.
| Name | # of CPU Cores | # of Threads | Max. Boost Clock | Base Clock | L3 Cache | Default CPU Power |
|---|---|---|---|---|---|---|
| AMD EPYC 9746 | 128 | 256 | Up to 4 GHz | 2.9 GHz | 512 MB | 400W |
| AMD EPYC 9736P | 128 | 256 | Up to 3.7 GHz | 2.7 GHz | 256 MB | 360W |
| AMD EPYC 9736 | 128 | 256 | Up to 3.7 GHz | 2.7 GHz | 256 MB | 360W |
| AMD EPYC 9676F | 96 | 192 | Up to 5 GHz | 3.1 GHz | 384 MB | 400W |
| AMD EPYC 9646P | 96 | 192 | Up to 3.7 GHz | 2.8 GHz | 256 MB | 300W |
| AMD EPYC 9646 | 96 | 192 | Up to 3.7 GHz | 2.8 GHz | 256 MB | 300W |
| AMD EPYC 9576F | 64 | 128 | Up to 5 GHz | 3.55 GHz | 384 MB | 400W |
| AMD EPYC 9536P | 64 | 128 | Up to 4 GHz | 3.25 GHz | 256 MB | 300W |
| AMD EPYC 9536 | 64 | 128 | Up to 4 GHz | 3.25 GHz | 256 MB | 300W |
| AMD EPYC 9526 | 64 | 128 | Up to 3.7 GHz | 2.75 GHz | 256 MB | 220W |
| AMD EPYC 9476F | 48 | 96 | Up to 5 GHz | 3.65 GHz | 192 MB | 330W |
| AMD EPYC 9456P | 48 | 96 | Up to 3.7 GHz | 3.2 GHz | 256 MB | 265W |
| AMD EPYC 9456 | 48 | 96 | Up to 3.7 GHz | 3.2 GHz | 256 MB | 265W |
| AMD EPYC 9376F | 32 | 64 | Up to 5 GHz | 3.8 GHz | 192 MB | 285W |
| AMD EPYC 9356P | 32 | 64 | Up to 4.5 GHz | 3.6 GHz | 192 MB | 250W |
| AMD EPYC 9356 | 32 | 64 | Up to 4.5 GHz | 3.6 GHz | 192 MB | 250W |
| AMD EPYC 9336 | 32 | 64 | Up to 3.7 GHz | 3.15 GHz | 128 MB | 195W |
| AMD EPYC 9276F | 24 | 48 | Up to 5 GHz | 3.8 GHz | 96 MB | 230W |
| AMD EPYC 9256 | 24 | 48 | Up to 4.5 GHz | 2.85 GHz | 96 MB | 190W |
| AMD EPYC 9176F | 16 | 32 | Up to 5 GHz | 3.9 GHz | 192 MB | 200W |
| AMD EPYC 9116 | 16 | 32 | Up to 4.5 GHz | 2.85 GHz | 48 MB | 160W |
| AMD EPYC 9016 | 8 | 16 | Up to 4.8 GHz | 3.05 GHz | 48 MB | 130W |
Just like GPUs, NVIDIA is AMD's primary competitor in this space, which is leveraging Vera CPUs based on its custom Arm IP to power the Vera Rubin NVL72 racks. AMD's Helios AI racks powered by Zen 6 EPYC chips not only offer more cores, but faster performance and improved single-core capabilities, which is important for sustained throughput.
In benchmarks, AMD has shown that not only do its 5th Gen Turin "Zen 5" CPUs lead against Vera, but the 6th Gen Venice "Zen 6" chips lead massively, driving more performance & offering better TCOs at a similar power scale.
For example, in Agentic CPU servers "sandboxes", AMD is claiming up to 2.8x agents per watt versus Arm-based CPUs such as NVIDIA's Vera, and up to 3.3x performance per watt in General-Purpose CPU servers.
AMD is also bringing up to 1.8x tokens/s, 1.7x agents/watt, and 1.4x performance/watt versus its 5th Gen EPYC "Turin" CPUs.
AMD EPYC CPU Families:
| Family Name | AMD EPYC Verano | AMD EPYC Venice-Dense | AMD EPYC Venice | AMD EPYC Turin-X | AMD EPYC Turin-Dense | AMD EPYC Turin | AMD EPYC Siena | AMD EPYC Bergamo | AMD EPYC Genoa-X | AMD EPYC Genoa | AMD EPYC Milan-X | AMD EPYC Milan | AMD EPYC Rome | AMD EPYC Naples |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Family Branding | EPYC 9007 | EPYC 9006 | EPYC 9006 | EPYC 9005 | EPYC 9005 | EPYC 9005 | EPYC 8004 | EPYC 9004 | EPYC 9004 | EPYC 9004 | EPYC 7004 | EPYC 7003 | EPYC 7002 | EPYC 7001 |
| Family Launch | 2027 | 2026 | 2026 | 2025 | 2025 | 2024 | 2023 | 2023 | 2023 | 2022 | 2022 | 2021 | 2019 | 2017 |
| CPU Architecture | Zen 6+ | Zen 6C | Zen 6 | Zen 5 | Zen 5C | Zen 5 | Zen 4 | Zen 4C | Zen 4 V-Cache | Zen 4 | Zen 3 | Zen 3 | Zen 2 | Zen 1 |
| Process Node | TBD | 2nm TSMC | 2nm TSMC | 4nm TSMC | 3nm TSMC | 4nm TSMC | 5nm TSMC | 4nm TSMC | 5nm TSMC | 5nm TSMC | 7nm TSMC | 7nm TSMC | 7nm TSMC | 14nm GloFo |
| Platform Name | SP7 | SP7 | SP7 | SP5 | SP5 | SP5 | SP6 | SP5 | SP5 | SP5 | SP3 | SP3 | SP3 | SP3 |
| Socket | TBD | TBD | TBD | LGA 6096 (SP5) | LGA 6096 (SP5) | LGA 6096 | LGA 4844 | LGA 6096 | LGA 6096 | LGA 6096 | LGA 4094 | LGA 4094 | LGA 4094 | LGA 4094 |
| Max Core Count | TBD | 256 | 96 | 128 | 192 | 128 | 64 | 128 | 96 | 96 | 64 | 64 | 64 | 32 |
| Max Thread Count | TBD | 512 | 192 | 256 | 384 | 256 | 128 | 256 | 192 | 192 | 128 | 128 | 128 | 64 |
| Max L3 Cache | TBD | 1024 MB? (128 MB /CCD) | 384 MB? (48 MB/CCD) | 1536 MB | 384 MB | 384 MB | 256 MB | 256 MB | 1152 MB | 384 MB | 768 MB | 256 MB | 256 MB | 64 MB |
| Chiplet Design | TBD | 8 CCD's (1 CCX per CCD) + 2 IOD? | 8 CCD's (1 CCX per CCD) + 2 IOD | 16 CCD's (1CCX per CCD) + 1 IOD | 12 CCD's (1CCX per CCD) + 1 IOD | 16 CCD's (1CCX per CCD) + 1 IOD | 8 CCD's (1CCX per CCD) + 1 IOD | 12 CCD's (1 CCX per CCD) + 1 IOD | 12 CCD's (1 CCX per CCD) + 1 IOD | 12 CCD's (1 CCX per CCD) + 1 IOD | 8 CCD's (1 CCX per CCD) + 1 IOD | 8 CCD's (1 CCX per CCD) + 1 IOD | 8 CCD's (2 CCX's per CCD) + 1 IOD | 4 CCD's (2 CCX's per CCD) |
| Memory Support | TBD | DDR5-8000+ | DDR5-8000+ | DDR5-6400 | DDR5-6400 | DDR5-6400 | DDR5-5200 | DDR5-5600 | DDR5-4800 | DDR5-4800 | DDR4-3200 | DDR4-3200 | DDR4-3200 | DDR4-2666 |
| Memory Channels | TBD | 16-Channel (SP7) | 16-Channel (SP7) | 12 Channel (SP5) | 12 Channel | 12 Channel | 6-Channel | 12 Channel | 12 Channel | 12 Channel | 8 Channel | 8 Channel | 8 Channel | 8 Channel |
| PCIe Gen Support | TBD | 128-192 PCIe Gen 6 | 128-192 PCIe Gen 6 | TBD | 128 PCIe Gen 5 | 128 PCIe Gen 5 | 96 Gen 5 | 128 Gen 5 | 128 Gen 5 | 128 Gen 5 | 128 Gen 4 | 128 Gen 4 | 128 Gen 4 | 64 Gen 3 |
| TDP (Max) | TBD | ~600W | ~600W | 500W (cTDP 600W) | 500W (cTDP 450-500W) | 400W (cDP 320-400W) | 70-225W | 320W (cTDP 400W) | 400W | 400W | 280W | 280W | 280W | 200W |
Follow Wccftech on Google to get more of our news coverage in your feeds.
