PC Ekspert Forum - View Single Post - AMD Ryzen "ZEN5" (9000 CPU series) Info thread

The Exiled · 17.07.2024., 22:39

Citiraj:

AMD is planning to launch both the Ryzen 9000 desktop and Ryzen AI 300 notebook processors this Summer. Some of the first notebooks powered by the Ryzen AI 300 series should be announced by notebook OEMs as early as this week; while the Ryzen 9000 series desktop processors will be available to customers from July 31, 2024. The new Zen 5 microarchitecture builds on AMD's proven CPU core technology, and doesn't try anything fancy like its generational counterpart from Intel, the Lion Cove P-core used on Arrow Lake. Zen 5 still brings a double-digit percentage IPC gain over the previous generation, and introduces several efficiency improvements over Zen 4 thanks not just to its newer 4 nm process, but also a host of other innovations. AMD was able to increase clock speeds, and lower TDP across the desktop processor lineup, and still achieve good performance gains.

Citiraj:

AMD has also worked on the physical characteristics of the processor's dies' cooling performance, which see a 15% reduction in thermal resistance, which results in an impressive 7°C reduction in temperature at the same TDP. We asked them to clarify this, and they confirm that no changes have been made to the IHS or the TIM. The improvements are due to better management of the "hot spot", and a different placement of the thermal sensor. Granite Ridge is a chiplet based processor, just like Raphael. AMD has largely carried over the 6 nm client I/O die (cIOD) from the previous generation, although the substrate has many changes owing to the different dimensions of the 4 nm Zen 5 CCDs, compared to the 5 nm Zen 4 CCDs.

Citiraj:

The Zen 5 and Zen 5c cores will be implemented across 4 nm and 3 nm foundry nodes. Both the Zen 5 CCDs on Granite Ridge, and the Strix Point monolithic processor are built on 4 nm, however AMD is working on high-density chipsets for 5th Gen EPYC that will see AMD implement 3 nm. Back when AMD Zen made landfall in 2017, nobody expected the company to remain competitive with Intel for more than a couple of generations, but the company has consistently proven everybody wrong. It has stuck to its roadmap, only promised what it could deliver, and delivered on those promises. We've been given consistent double-digit IPC gains for each new Zen generation, and every two generations, AMD has transitioned to a new class of process nodes. In a way, AMD and TSMC have been able to imitate Intel's tick-tock product development cycle far better than Intel itself.

Citiraj:

AMD Granite Ridge and Strix Point Zen 5 die-sizes and transistor counts confirmed

Citiraj:

To begin with, "Strix Point" is a monolithic silicon, which is confirmed to be built on the TSMC N4P foundry node (4 nm). This is a slight upgrade over the N4 node that the company built its previous generation "Phoenix" and "Hawk Point" processors on. The "Strix Point" silicon measures 232.5 mm˛ in area, which is significantly larger than the 178 mm˛ of "Hawk Point" and "Phoenix." The added die area comes from there being 12 CPU cores instead of 8, and 16 iGPU compute units instead of 12; and a larger NPU. There are many other factors, such as the larger 24 MB CPU L3 cache; and the sizes of the "Zen 5" and "Zen 5c" cores themselves.

The "Granite Ridge" desktop processor is a chiplet-based processor, much like the Ryzen 7000 "Raphael." AMD is confirmed to be reusing the 6 nm client I/O die (cIOD) from "Raphael." This chip measures 122 mm˛, and packs 3.4 billion transistors. For reference, the cIOD of Ryzen 5000 "Vermeer" and Ryzen 3000 "Matisse" is built on the Global Foundries 12 nm node, measures a similar 125 mm˛, but with a much lower transistor count of 2.09 billion. The key contributor to the transistor count increase is the tiny iGPU that the Socket AM5 cIOD comes with. It may have just 1 workgroup processor (2 CU), but comes with the same display engines and media engines as the iGPU on APUs.

And now, onto the CPU complex dies (CCDs), the key area of silicon innovation for AMD desktop processors. The 8-core "Zen 5" CCD is codenamed "Eldora," and is built on the 4 nm foundry node. The HardwareLuxx.de report says that this the same N4P node as "Strix Point," but we've heard several other credible sources claiming that it is the more advanced N4X node, which favors high frequencies. The "Zen 5" CCD has a transistor count of 8.315 billion, which is a significant increase over the 6.5 billion of "Durango," the 8-core CCD based on "Zen 4," powering "Raphael."

What's most interesting is that this staggering 28% increase in transistor counts from the "Zen 4" Durango CCD and "Zen 5" Eldora CCD comes at a die-area decrease of 0.5%. That's right, the "Zen 5" CCD measures 70.6 mm˛, whereas the "Zen 4" CCD is 71 mm˛. The "Zen 4" CCD is built on TSMC N5 (5 nm), which goes to show the tremendous leap in transistor densities achieved from the switch to N4P (or N4X). A maxed out Ryzen 9 9950X processor hence has a total transistor count of 20.03 billion, while the single-CCD Ryzen 7 9700X has a transistor count of 11.715 billion. AMD Ryzen 9000 series goes on sale from July 31, 2024.

Izvor: TechPowerUp i HardwareLUXX

Citiraj:

AMD Ryzen 9 9950X flagship “Zen 5” 16-core CPU tested From 40W up to 230W, matches Core i9-14900KS at 160W in Cinebench

Citiraj:

AMD's Ryzen 9 9950X, the fastest Zen 5 Desktop CPU, has been tested in Cinebench & matches Intel's top 14900KS chip at just 160W. The latest benchmarks of the AMD Ryzen 9 9950X 16-Core "Zen 5" Desktop CPU have been published by Anandtech forum member, Igor_kavinski, who previously showcased the full performance scaling potential of the flagship in Blender at various TDPs. For the latest tests, Cinebench R23 was selected as the go-to benchmark and the results are quite amazing. Once again, Igor's results come from a source running an ES (Engineering Sample) version of the AMD Ryzen 9 9950X CPU which is rated at lower clock speeds but was manually tweaked (PBO+CO) & tested at various target PPTs.

AMD Ryzen 9 9950X "Zen 5" Desktop CPU manages to match the 7950X3D at 120W and is able to match the Core i9-14900KS with a 160W TDP. The result versus i9-14900KS is the standout given that the latter was tested at its extreme profile which can push over 300W of power, almost double that of the 160W PPT and this isn't even the retail chip. With a PPT of 200W, the 9950X sits 5% ahead and with a PPT of 230W, the Zen 5 chip sits 8% ahead (and its still going to be less power hungry than the i9). The multi-threaded performance of the AMD Ryzen 9 9950X CPU with its 16 Zen 5 cores is looking great and with the added efficiency and lower temps, it will turn out to be a leadership product for the high-performance desktop segment.

Izvor: Interneti