TheRYZENcaches, we already spilled the beans on that one but initially all we got was the number which is 20 MB (for an 8-core processor) for the L2 and L3 cache, and that matches what we have discussed numerous times.
The 8-core part with two four core CCU units aka Summit Ridge akaRYZENin the end get a L1 data cache size of 32KiB, a L1 instruction cache size of 64KiB and then a L2 cache size of 512KiBper core.
L1 8 x 32 Kb Data
L1 8 x 64 Kb Instruction
L2 8 x 512 Kbytes
L3 2 x 8 Mbytes L3
So that is 8 x 512 KB (L2) + 2 x 8 MB (L3) = 20 MB in total for caches, again one clustered Core Complex Unit holds four processor cores. The L2 and L3 caches are similar in size compared to the Intel Core i7-6900K. The processor has dual channel DDR4 support (up-to 3200MHz), AVX2,AES, FMA3, AMD-V SSE 4.1 and 4.2 instruction sets etc. The bus frequency is 100 MHz multiplied by whatever the processor fires off at it. Ryzen is built with a 14nm FinFET fab node, this greatly helps where AMD is with the performance and power consumption. For the transistor aficionados, the number is 4.90 Billion of them.
That is the money shot right there, eh?
AMDis implementing technology that will make sure that this processor runs applications efficiently and optimized.SenseMIis a set of sensing and adapting technologies, including an artificial network inside every “Ryzen” processor to anticipate future decisions,preloadinstructions, and choose the best path through the CPU.
AMD isalso introducing a new interconnect calledAMDInfinity Fabric. This is a new and fast way of connecting various parts within anSoC. Infinity Fabric is not just used in theRYZENprocessors, but also in future (Vega)AMDGPUsand (almost) all otherAMDchips in the near future.InfinityFabric allows for faster and better secure connections within a chip. The inter- and intra-chip connector will be standardized and used in many AMD products and, as such, AMD can easily communicate over that very same interconnect. SenseMItechnology then; it is based on five parts as shown in the slide below. Pure Power is a technology that allows theRyzenchips and other Ryzen variants to work as efficiently as possible.
According toAMD, theCPUsare equipped with hundreds of sensors that monitor temperature, voltage and power consumption of all chip segments. Through the Infinity Fabric connections these sensors are all connected to a central control unit, the Infinity System Management Unit. The same network of sensors constantly monitors how much room there is left for extra performance, chip parts can be independently adjusted from each other. Clock frequencies within a split second can be altered in steps of 25 MHz. To understand the euphemism that isSenseMI, everything inside the processor works heterogeneously and optimized. So you get precise boost frequencies matched to power management. What is a simple to grasp manner to understand is thatSenseMiwill do a lot of analyzing and smart prediction based upon the workload at hand.
If I recall correctly, it was back in the summer in a presentation that AMD already announced that the existing Zen branch predictor (the code that predicts the path where branches in code probably will be) was mentioned to have been greatly improved. AMD indicated that has improved thanks to AI-like algorithms and protocols. The new branch predictor has a self-learning system on board, making decisions based on good or bad branches and thus it can adapt and improve on performance based upon specific workloads and surrounding dynamics. Branch prediction is essential for the efficient operation of the processor. Based on applications and workloads it can detect what course or path is better to follow, making the result a faster one. It will analyze the workload and adapt and optimize based on that. The huge gain in RYZEN IPC (now rated at a 52% IPC increase over last –gen) performance is to be found, among other things, right here with much-improved branch prediction.
So decisions are driven by software code execution and anticipate on following decisions,pre-load instructions, choose the best path through the CPU. It’s pretty impressive when you think about it, a processor that optimizes workloads based on algorithms and predictions. AMD will be able to precisely bin and adjust clock frequencies in steps as small as 25 MHz. RYZEN will get that much smarter a prefetecher that anticipates the location of future data accesses by application code. It has a learning algorithm model and learns application data access patterns allowing it to prefetch vital data into local cache memory so it’s ready for immediate use. Efficiency at its best, and after seeing the first results, AMD might have struck gold here.
Back to SenseMI;SenseMIis a monitoring algorithm that monitors and adapts on many levels. On a more arbitrary level this will also mean thatSenseMI, with its many sensors, can recognize thermal characteristics and adapt to them, e.g. a better cooler will optimize the processor to go faster and perhaps clock the turbo bins upwards a notch. ARYZENprocessor with water-cooling will get better temperatures, right? Well, the technology will detect that thermal signature and as such it will allow higher boost frequencies with better cooling. Thermal tweaking control at the processor level.
And that is where we arrive at XFR; XFR alone uses like 100+ sensors (out of a 1000+ sensors per core complex CCX) inside the system and can boost the frequency above the specs available. So it’ll give you a perf boost based on proper cooling and variables like processor binning. With a processor that normally boosts to, say, 4.0 GHz you could see the processor reach a higher boost at 4.1 GHz, if the circumstances are right.
AMD Ryzen 5 1500X and 1600X review We put Ryzen 5 to the test as we benchmark the six-core Ryzen 5 1600X and four-core Ryzen 5 1500 processors. The 1600X we can tell you already is a bit of a beast, and that quad-core just oozes proper...
AMD Ryzen 7 Memory And Tweaking Analysis review In this article we'll take the new G.Skill Flare X 3200 MHz memory kit and have a closer look at the effect of memory frequencies on applications and games. Next to that we'll also address tweaking ...
AMD Ryzen 7 1700 Review We test the most affordable Ryzen 7 series processor, the 1700 CPU will cost you only 329 USD. This units oozes value as the performance is really good. The 8-core processor will be tested on an X370 ...
AMD Ryzen 7 1700X Review We tested the Flagship processor, now we review the Ryzen 7 1700X Processor that is a 100 bucks cheaper. The 8-core processor will be tested on an X370 motherboard. ...