As part of AMD’s 2022 Financial Analyst Day today, the company is offering a short, high-level update on the upcoming Zen4 CPU architecture. This information has been published as part of the company’s larger Zen architecture roadmap and has been extended today to announce Zen 5 in 2024.
The biggest news here is that AMD is the first to unveil IPC’s expectations for a new architecture. Addressing some post-Computex questions about IPC expectations, AMD reveals that it expects Zen 4 to offer an 8-10% improvement in IPC over Zen 3. That’s why AMD is committed to responding to it without spending months after the product launches.
This is a significant part of AMD’s overall expected single-threaded performance improvement of over 15%. This was previously disclosed in Computex and is essentially unchanged. That said, AMD strongly emphasizes the “bigger” aspect of its performance estimates. At this point, AMD hasn’t locked down the final clock speed, so it can’t be overly specific, but as we saw in the Computex demo, it’s currently at a peak of 5.5GHz (or higher). The clock speed is in the Zen4 table.
AMD is also discussing a bit more about today’s power and efficiency expectations. At this point, AMD predicts that Zen 4 will perform more than 25% per watt than Zen 3 (based on a desktop 16C chip running CineBench). On the other hand, the overall performance improvement is over 35%, both with the per-thread architecture performance improvement and the higher TDP previously disclosed by AMD (especially useful for improving MT workload performance). There is no doubt that you are using it. And yes, these are terrible graphs.
Finally, AMD has confirmed that its processor lineup includes the Zen4 SKU with V-Cache. No specific SKU has been announced today, but AMD has repeatedly stated that V-Cache is not a one-time experiment for the company, and that some Zen4 chips will also adopt the die-stack L3 cache.
