The first AMD chips based on the company’s new Zen 4 CPU architecture are set to hit the streets starting September 27th, 2022. The chip maker says its new Ryzen 7000 series desktop processors deliver up to 29% better single-core performance and even bigger gains in multi-core performance, while also offering better performance per-watt than the previous-gen Ryzen 5000 desktop chips.
Prices start at $299 for an AMD Ryzen 5 7600X processor with 6 cores and 12 threads and go up to $699 for a top of the line Ryzen 9 7950X chip with 16 cores and 32 threads.
Here’s a run-down of the first four Ryzen 7000 series desktop chips, all of which should be available starting September 27th:
|Model||Cores / Threads||Boost / Base Frequency||Total Cache||PCIe||TDP||SEP (USD)|
|Ryzen 9 7950X||16C / 32T||Up to 5.7 / 4.5 GHZ||80MB||Gen 5||170W||$699|
|Ryzen 9 7900X||12C / 24T||Up to 5.6 / 4.7 GHZ||76MB||Gen 5||170W||$549|
|Ryzen 7 7700X||8C / 16T||Up to 5.4 / 4.5 GHZ||40MB||Gen 5||105W||$399|
|Ryzen 5 7600X||6C / 12T||Up to 5.3 / 4.7 GHZ||38MB||Gen 5||105W||$299|
AMD says when compared with a previous-gen Ryzen 9 5950X processor, the new Ryzen 9 7950X chip delivers:
- Up to 29% better single-core performance
- Up to 45% “more compute for content creators in POV Ray
- Up to 15% “faster gaming performance in select titles”
- Up to 27 better performance-per watt
Across the board, the new chips feature Zen 4 CPU cores manufactured using TSMC’s 5nm process and a new 6nm I/O die with support for hardware-accelerated video encoding and decoding as well as support for other simple graphics.
AMD is also introducing a new AM5 socket with support for up to 24 PCIe 5.0 lanes and DDR5 memory. AMD says motherboards with Ryzen the 7000-compatible sockets will begin to arrive in September with prices starting at $125. The company says it will support the AM5 socket at least through 2025, which means that a motherboard you buy this year should support next-gen AMD chips released over the next few years as well.
And the company hinted at what’s next, with new 5nm and 4nm Zen 4 chips featuring 3D V-Cache and Zen 4c chips expected in the coming year or so, to be followed by 4nm and 3nm Zen 5 chips by 2024.
Wow thats a bit of a jump in TDP.
The Ryzen 5 5600X had a 65W TDP, the Ryzen 5 7600X is 105W.
The Ryzen 9 5900X had a 105W TDP, the Ryzen 9 7900X is 170W.
They’re saying these will have a 27% increase in performance-per-watt. The Ryzen 9 5900X had a passmark score of around 39,000. So that’s around 370 passmarks per watt.
With an increase of 27%, we should expect the Ryzen 9 7900X to have around 470 passmarks per watt. Given a TDP of 170W, it should be close to a Passmark score of 80,000.
This has to be wrong. Is my math wrong, or is CPU marketing just that misleading?
Obviously I chose a stupid benchmark method, but I feel like we’re not even in the right ballpark.
I agree… not apples-to-apples with the power envelop. These are Intel-style power envelopes. Not planning to upgrade my 5600X for a while… my previous AMD chip was a X2 3800+ (switched to Intel after that).
Yeah, your math is completely wrong. The TDP isn’t how much power the processor uses, it’s more of a rough guide as to which power supply and fan you need to install.
The actual power used is more difficult to calculate, and many people simply take the Total Power Used from the wall/power point. And you can even make a comparison between Idle use, versus Gaming/Stress Use, versus Maximum Power Draw.
AMD’s figures aren’t wrong, and sometimes they under-report the gains. They’re much much more honest than Intel. With that said, I’ll tell you how they achieve those impressive figures.
AMD clocks the 5600X to its maximum and records the specific power-draw, and the performance in a suite of benchmarks. Since this is at the max frequency, the voltage is high, and your at the far end of the efficiency curve. When the 7600X gets compared, it can easily achieve that performance level at a more regular frequency. Since it doesn’t have to overclock, it stays in the healthy range of the efficiency curve. And that’s how you get the figure of +74% gains.
Now, if you look at the 5950X which has a lot of performance, it takes a lot of power for the 7950X to actually match it. It’s also not in the healthy zone, but at least it isn’t as stretched out. That’s why the difference here is only +23% efficiency gained.
Now if we were to scale this down to laptops, you would be shocked. It’s probable the Zen3 system of 6800u would only be 5% efficiency gain. In fact, at this point the Zen3 chipset might even have higher efficiency. Since the new chips haven’t been as optimised yet from the driver side, and those AVX-256×2 and DDR5 memory do hog power.
AMD was playing it safe with Zen4 (heavily based on Zen3), and they were keeping their cards close for the Zen5 release (new architecture). But Intel’s 12th-gen released has forced them to overclock the Zen4 chips and reclaim the performance crown. I doubt Intel could do anything substantial with Raptor Lake, not unless they fix their Fab and get a more competitive node out.