During the final episode of PCWorld’s “The Full Nerd” webcast series, guests Robert Hallock (AMD’s chief marketing officer) and Frank Azor (chief gaming solutions architect) answered a series of questions. and further detailed the Ryzen 7000 desktop CPUs and features. see you on the AM5 platform.
Details of AMD Ryzen 7000 Desktop CPU and AM5 motherboard platform features
So a series of questions were asked and both Robert and Frank did an absolutely splendid job of answering the people at PCWorld. Gradually we have been getting more details about the AMD Ryzen 7000 CPU and the AM5 platform shown since the presentation of the Computex 2022 and well, let’s start with the latest information.
The AMD software update enhances RSR and increases DirectX 11 gaming [Update: Full Version Out]
Some of the things that were confirmed in the interview include:
- Ryzen 7000 is a packet power of 125 W TDP / 170 W
- The 5.5 GHz Ryzen 7000 demo was in stock (no overclocking)
- The Ryzen 7000 dual L2 cache is an IPC advantage
- Ryzen 7000 CPUs have 28 PCIe Gen 5 lanes (24 usable)
- 1: 1 Infinity Clock Clock (no frequencies mentioned)
- B650 motherboards will support overclocking (like the B550 series)
- The built-in RDNA 2 GPU supports both video encoding / decoding
- Integrated RDNA 2 GPU for commercial / diagnostic purposes
- Smart Access Storage Details (Highlights)
AMD Ryzen 7000 CPU Package Power: 125 W TDP for Ryzen 7000 CPU and 170 W Package Power for AM5 Socket
First, the TDP (heat dissipation power) figures for AMD Ryzen 7000 CPUs will be 125 W while the PPT (packet power monitoring) will be 170 W. AMD provided this information in a reply. whether the 170 W figure was an actual TDP for the next chips or an upper limit for the package.
According to AMD, this is an increase of about 28 W with respect to the power limit of the AM4 package (PPT), which was 142 W, while the CPUs had a TDP of 105 W. According to AMD, the board manufacturers base will now be able to deploy more premium power features to their motherboards that should allow for better overclocking opportunities for enthusiasts and overclockers.
So what we want to clarify is that this is a 170 watt socket power that with AMD, this specification is PPT (packet power) for us. This is not to say that each CPU will go up to 170 watts, but it is 30 (watts) higher than the power limit of the AM4 socket which was 142 (watts). And we did it primarily to improve the performance of multiple wires, as many of the core counting chips slowed down overall computing performance by relatively modest power.
The other point I want to make is that by increasing the minimum power required or the minimum specification, you also increase the power supply with each motherboard built to these specifications, so that you get more robust power characteristics across all boards. that we are quite excited. It should also be good for people who want to experiment with overclocking, people who appreciate premium dashboard designs.
Robert Hallock (AMD Technical Marketing Director)
In addition, Robert confirmed that the AMD Ryzen 7000 CPUs include a total of 28 Gen 5 PCIe lanes, confirming our previous leaks. The fact is that of these 28 lanes, only 24 are usable by the discrete GPU and M.2 storage devices, as the remaining four lanes are connected to the 600 series PCH. As for the chipsets themselves. , Robert also confirmed that the B650 motherboards will be fully compatible with overclocking like their B * 50 class predecessors. We have more information on this and the B650E chipsets here.
AMD B650 series motherboards will feature CPU and memory overclocking support: B650E for mid-range and high-end options with full support for the Gen 5, October release
Demonstration of the 5.5 GHz AMD Ryzen 7000 CPU was in stock specifications, no overclocking or premium components were used
As for the 5.5 GHz clock speed game demonstration, Robert assured that the frequencies were entirely within stock specifications. The motherboard used was an X670 reference design and the cooling was an ASETEK 280 mm AIO standard refrigerator. It is also obvious that there was no overclocking, as the clocks ranged from 5.1 to 5.5 GHz.
AMD showed some incredibly fast frequencies with the same sample of Ryzen 7000 CPU reaching 5.52 GHz, but we saw some variation in clock speeds that started at 5.1 GHz and reached the top speed of 5 .52 GHz that everyone is talking about. Interestingly, Robert states that in the demonstration of the respective game, they saw that most of the threads recorded up to 5.5 GHz (i.e. 32 threads for the prototype used). The 16-core Ryzen 7000 prototype came out in late April or early May, so AMD could still get more out of this chip if they want to or just let the overclockers do the work.
16-core AMD Ryzen 7000 pre-production sample “game” watches:
We used a 280mm water cooler (ASETEK), so nothing exotic is just a 140-loop dual loop that you can buy on Amazon or Newegg. It had an AMD reference motherboard, so one of our AM5 internal socket motherboards, a 16-core prototype piece built in late April or early May, we plugged it in and ran it. It wasn’t an overclocked part, it’s just the natural frequency of this particular prototype.
So in the game, we had most of the wires around 5.5 (GHz), it depends on the load of the game, it depends on the scene, of course, the clock speed fluctuates up and down, so between 5.2 and 5.5 is quite common to all. the threads that play this game. So I want to assure people that this was nothing exotic when it comes to cooling and nothing exotic when it comes to parts selection, setup or overclocking. That’s exactly what you see, you plug in a part of the Ryzen 7000 series and play the game, and that’s how often!
Robert Hallock (AMD Technical Marketing Director)
AMD Ryzen 7000 can achieve an infinite 1: 1 fabric clock, doubles L2 cache
Robert also talked about how doubling the L2 cache can help improve CPI-related core performance. L2 cache did not increase dramatically, as more L2 cache results in more visits, but it also means poor latency and array sizes become larger, which increases the cost. Thus, AMD found a ground where the L2 cache is increased enough to increase the performance advantages of Ryzen 7000 desktop CPUs.
Therefore, the L2 cache, we duplicated it in Zen 4, is an IPC advantage, it helps the kernel to contain more information internally, it generally improves performance, and therefore will be beneficial for multiple categories of workloads, whether whole as games or more. things like floating point such as scientific and creative workloads.
Robert Hallock (AMD Technical Marketing Director)
When asked if the new AM5 platform will run DDR5 memory in a 1: 1 ratio with Infinity Fabric, Robert replied “Yes”, but did not mention how often it would run. This is another topic of discussion as we approach the official launch of the Ryzen 7000 desktop CPUs on AM5 motherboards.
IGPU AMD Ryzen 7000 RDNA 2: video encoding / decoding, APUs are still happening for desktops!
As for what the new RDG 2 iGPUs for Ryzen 7000 desktop CPUs bring to the table, AMD claims that with a graphics integrated into its entire CPU portfolio, they can expand their business to the commercial segment and that has a lot to offer. meaning since a lot of consumers don’t need a discreet graphics card and want something as simple as plug-and-play.
For DIY builders, the iGPU RDNA 2 can provide troubleshooting and diagnostic capabilities where users with graphics cards can debug whether or not their graphics card is defective or for other purposes. The same can be said for users without graphics cards or those waiting for one who can’t turn on their PC until they get their discrete GPU.
Frank Azor took the topic to a more interesting place by stating that while AMD Ryzen 7000 CPUs will have “small graphics” cores compared to APU’s “larger graphics” cores, they will still house some of the technologies. Smart Eco made by laptops. While AMD’s RDNA 2 iGPUs can support power usage below 50 W in idle mode, Smart Shift on desktop CPUs with RDNA 2 iGPUs can switch from discrete graphics to built-in graphics for light workloads and offer a power of less than 5 W or even a mW power.
The other thing is that, unlike AMD’s Navi 24 GPUs, the Ryzen 7000 iGPU based on the same RDNA 2 core architecture (but in a 6 nm Rembrandt revision) will come with a VCN engine that supports both encoding such as AV1 video decoding.
We still think of the Ryzen 7000 series as a CPU. The graphics cores in this IO array are not many, the purpose of adding graphics is threefold. First, it expands these products a lot in the commercial market where they don’t buy anything discreet, they just want to activate it, have video encoding / decoding and illuminate some screens for office work and that’s what the GPU in IO . die will offer, so this is a great business opportunity for us alongside Ryzen PRO as we begin to migrate these components to this business.
The second is for diagnostic purposes, how do you know you have a bad graphics card? Well, you have to switch to another graphics card, but with the graphics core we have, you can do some troubleshooting in the third place, we were thinking about users who plan to buy a discrete graphic and it is still in traffic by mail. but all the other hardware has come first, so everything is sitting there, looking at a bunch of components and not having a GPU to set it all up. This would disappear with the Ryzen 7000 series.
We will still make APUs with big graphics, so APU “BIG GRAPHICS”, CPU “small graphics”. That would be our strategy in advance.
Robert Hallock (AMD Technical Marketing Director)
We are developing many technologies that make use of integrated graphics in many ways and there are things we can do with technologies like Smart Shift ECO where we can turn off discrete graphics and run the notebook off the iGPU and say you want this …