AMD @ Computex 2022 Talks Up Ryzen 7000 Series, Announces Mendocino Budget Laptop APUs

[rclark]

Good to see as the competition benefits everyone. I also like the idea of on-board graphics too rather than have special 'g' chips for the purpose. My server boxes for example run headless and normally don't need a monitor, but there are times when you need one to update bios settings and such. That's where the GPU on the CPU come in handy.

That said, I don't see me jumping on the AM5 system right away if not for a long time -- unless current motherboard/cpu breaks of course. I have so much performance just 'sitting there' with my Ryzen 5000 series boxes (5900X and 5600X), I just don't see any need for upgrading. I can't help wonder if there are others out there like me that will just stick with the AM4 platform for the foreseeable future. All my systems run Linux of course. VMs run very well... Just can't see/justify upgrading.

[WannaBeOCer]

You say this like it's a bad thing?

That said, I've not seen a major uplift in speed for the code I actually care about on the 6300 Xeons vs. the 6200 Xeons... and AMD will continue to be popular in the datacentre if their core count keeps going up. We'll see what Sapphire Rapids brings to the table.

For the data center buy AMD epyc for hypervisors, storage nodes and GPU nodes. Buy Intel xeon for CPU compute and databases. It’s the reason why AMD followed Intel by adding their own ML instructions to Zen 4. While Intel is adding AMX to Sapphire Rapids.

[Paradigm Shifter]

AMD woke a sleeping giant and so far Intel hasn't been slowing down with the IPC improvements.

You say this like it's a bad thing?

That said, I've not seen a major uplift in speed for the code I actually care about on the 6300 Xeons vs. the 6200 Xeons... and AMD will continue to be popular in the datacentre if their core count keeps going up. We'll see what Sapphire Rapids brings to the table.

[WannaBeOCer]

I'm confident if they claim 15 % more single thread performance it will be right there, maybe 1 or 2 % off for independent reviews. Since the start of Zen AMD has been pretty honest with their performance claims. If Intel made that claim I would instantly bet its around 7 to 5 % in reality. Intels "benchmarks" must be a meme by this point.
But yeah, only time will tell.

It's not IPC but Cinebench performance. That's the reason they are marketing "single-thread uplift" not IPC like they were previously. Intel has been on target with their performance numbers. First with Sunny Cove cores with an 18% IPC increase over Skylake followed by Golden Cove cores with a 19% IPC improvement over Ice Lake. You can be bitter about the old sandbagging Intel that was focused on storage but their performance numbers have been accurate thanks to competition. AMD woke a sleeping giant and so far Intel hasn't been slowing down with the IPC improvements. Their accurate roadmap shows Raptor Lake will also have a major IPC improvement.

[Anux]

I also do not think they are deliberately underestimating performance. I think they are just being very conservative in their estimates and careful with their wording.

I'm confident if they claim 15 % more single thread performance it will be right there, maybe 1 or 2 % off for independent reviews. Since the start of Zen AMD has been pretty honest with their performance claims. If Intel made that claim I would instantly bet its around 7 to 5 % in reality. Intels "benchmarks" must be a meme by this point.
But yeah, only time will tell.

[theriddick]

Hmm, lots of uncertainty surrounding these 7000series CPU's. It wouldn't be so bad if DDR5 wasn't TWICE the price of DDR4 here.. (Australia)

[Vlad42]

A lot of people on other forums are melting down over this right now.

https://wccftech.com/amd-greater-tha...nother-jebait/

The fine print on the slides could lead you to believe that they are deliberately underestimating the IPC/performance of Zen 4. Doesn't seem like a good strategy to me, but we'll find out the truth eventually the right way: by reading third-party reviews instead of leaks and marketing claims.

First, the slides only say greater than 15% uplift in single threaded performance. When looking at the footnote, they compared an unspecified 16 core preproduction 7000 chip versus the 5950X. However, we have no idea what clock speed the preproduction 7000 chip ran at - they could have matched the clock speed to the 5950X. I which case, the greater than 15% uplift would be on a per clock basis. We really just do not know, but it is unlikely to be just 15% faster. Just 15% faster would have the 7000 series only match the 58003DX in games that benefited from the extra L3 cache.

As to why they might be downplaying performance? They could be trying to dampen expectations give some of the wild rumors claiming 40% IPC increases both with and without 3D v-cache included and significant clock speed increases on top of the IPC gains. If these rumors went unchecked, then whatever AMD released would have been met with disappointment (even if they blew Intel out of the water across the board). From AMD's perspective, it's always better to under-promise on early performance previews and over-deliver on performance at release. They did the exact same thing with the original Zen chip.

I also do not think they are deliberately underestimating performance. I think they are just being very conservative in their estimates and careful with their wording.

Not at launch. You probably have to wait until 2023 to see a 3D cache model. Hopefully they will put it on at least the 16-core as well as the 8-core this time.

We really don't know if there will be a 3D v-cache model at launch. If Zen 5 launches in 2023 as expected, then there would not be much time for a 3D v-cache refresh of Zen 4. AMD is unlikely to want to pull punches with Raptor Lake launching in a similar time frame, so they might launch one or two models.

That sounds overly optimistic for Raptor Lake's single-threaded gains. It will probably have a similar bump to this (IPC + clocks = 10-15%), and it will be adding another 8 small cores, but only in the Core i9. But as it stands, that's more than enough to defeat Zen 4. We'll see if that changes in a few months.

I'm honestly expecting the Alder Lake to Raptor Lake p-core improvement to be like the Ice Lake to Tiger Lake improvement but with far less clock speed gains. That is, more or less a net neutral on IPC and a sight bump in clock speed (200-300 MHz at most). I don't think Intel has ever released two major CPU IPC increasing architectures updates year over year - definitely not since Conroe launched.

[jaxa]

I'm curious about 16 cores + 3D cache performance.

I believe it would be identical to the 8-core in many cases, since each chiplet could only access one stack of the L3 cache. For example, both the 8-core 5800X3D and 64-core EPYC 7773X have 96 MiB per 8-core CCX, so 96 MiB is the maximum amount a single core can use.

[shmerl]

Not at launch. You probably have to wait until 2023 to see a 3D cache model. Hopefully they will put it on at least the 16-core as well as the 8-core this time.

I'm curious about 16 cores + 3D cache performance.

[jaxa]

I am wondering if there is any plans for a beefier APUs, where single compute chiplet is replaced with GPU chiplet?
Limiting CPU to 8C/16T, but still enough.

Raphael (and by extension, Dragon Range for laptops) will have weak graphics located in the I/O die, not a primary chiplet. They don't seem to have such plans yet, although maybe they will make a version for server/HPC eventually.

Intel on the other hand does seem to be going that route, for example by putting a Battlemage graphics tile into Meteor Lake. Presumably this tile would be shared with discrete GPUs that have 1-4 of them.

Hm, 15% single core of which 12% is accountable to the frequency jump (4,9 to 5,5 GHz) leaving 3% per clock improvments. That is either played low or really bad, it would make zen 4 lose to zen 3D in some cases. Having 170W TDP limit also isn't flying with me.
Didn't they do anything besides porting zen 3 to 5nm and doubling the L2?

A lot of people on other forums are melting down over this right now.

https://wccftech.com/amd-greater-than-15-single-threaded-performance-for-ryzen-7000-cpus-could-be-another-jebait/

The fine print on the slides could lead you to believe that they are deliberately underestimating the IPC/performance of Zen 4. Doesn't seem like a good strategy to me, but we'll find out the truth eventually the right way: by reading third-party reviews instead of leaks and marketing claims.

Will new CPUs have 3D cache?

Not at launch. You probably have to wait until 2023 to see a 3D cache model. Hopefully they will put it on at least the 16-core as well as the 8-core this time.

Looks like the Zen architecture is finally hitting its limitations. Intel’s Raptor Lake is going to be released with another 15% single threaded IPC increase over Alder Lake along with a 40% increase in multi-core performance. It’s insane how Intel was on 14nm for 6 years and AMD was only able to beat them for one generation due to TSMC’s 7nm process.

That sounds overly optimistic for Raptor Lake's single-threaded gains. It will probably have a similar bump to this (IPC + clocks = 10-15%), and it will be adding another 8 small cores, but only in the Core i9. But as it stands, that's more than enough to defeat Zen 4. We'll see if that changes in a few months.