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

[Paradigm Shifter]

I'm looking at getting the 5700G, since it's not a bad increase to get two more cores, and at 65W it will handily beat Phenom II 945 X4 at 110W of power. So a big speed increase and a big power drop as well. Can't lose! So 32gb RAM, new MB and CPU looks to be $800, since I have the rest of the stuff I need.

Just don't do what I did - for some reason, when I was looking at upgrading, I completely blanked on the fact that the 5700G is PCI-E Gen.3, not Gen.4.

I mean, I just repurposed it, but... yeah, felt like an idiot.

The 65W chips are nice, though - a lot of performance for a lot less power draw.

[l8gravely]

Ryzen 5 5600 gets almost all of the performance of the 5600X but should have better price/performance. Firesale pricing on high-core count models may be attractive if you can actually use all of that performance. For example, the 5950X has been obtainable for under $550 in recent months, less than what people paid for a 5900X at launch.

If there is a 5600G at $200 and 5700G at $250, just stretch your budget and go for the 5700G.

I'm looking at getting the 5700G, since it's not a bad increase to get two more cores, and at 65W it will handily beat Phenom II 945 X4 at 110W of power. So a big speed increase and a big power drop as well. Can't lose! So 32gb RAM, new MB and CPU looks to be $800, since I have the rest of the stuff I need.

[microcode]

Not if you want to put the GPU on laptop for example. Rurrently all AMD APUs are laptop design chips, retrofitted on AM4 socket, without any I/O die, but a monolithic.

GPU dies are generally bigger than CPU dies even in laptops; or in the case of single-die SoC, the GPU is typically taking more area than the CPU, even for low-end integrated graphics.

[jaxa]

From the looks of this, even though I want to wait for Zen4, it would make more sense for me to upgrade to a recent Zen3 processor/MB/DDR4/SSD combo to replace my ancient AMD Phenom II X4 DDR2 16gb main server system. It's hitting almost 14 years old now and still running strong in alot of ways. It's just really starting to show it's age and I'm sure *anything* would be an improvement now, especially with 32 or 64gb of RAM so I can run more stuff as VMs.

So what do people think is the sweet spot of Zen3 in terms of price/perfomance/power these days? The 8 core chips look really good, but only up until a point. Buying an extra 200mhz for $100-$150 more doesn't appeal when it makes more sense to get more memory.

Ryzen 5 5600 gets almost all of the performance of the 5600X but should have better price/performance. Firesale pricing on high-core count models may be attractive if you can actually use all of that performance. For example, the 5950X has been obtainable for under $550 in recent months, less than what people paid for a 5900X at launch.

If there is a 5600G at $200 and 5700G at $250, just stretch your budget and go for the 5700G.

[ddriver]

So what do people think is the sweet spot of Zen3 in terms of price/perfomance/power these days?

The 6 core is IMO the sufficient performance / high value sweet spot. It will still be a huge update to a phenom. The 12 core if you need the performance. Get an apu if you don't have a gpu.

[l8gravely]

From the looks of this, even though I want to wait for Zen4, it would make more sense for me to upgrade to a recent Zen3 processor/MB/DDR4/SSD combo to replace my ancient AMD Phenom II X4 DDR2 16gb main server system. It's hitting almost 14 years old now and still running strong in alot of ways. It's just really starting to show it's age and I'm sure *anything* would be an improvement now, especially with 32 or 64gb of RAM so I can run more stuff as VMs.

So what do people think is the sweet spot of Zen3 in terms of price/perfomance/power these days? The 8 core chips look really good, but only up until a point. Buying an extra 200mhz for $100-$150 more doesn't appeal when it makes more sense to get more memory.

Lots of fun specing out a new system for sure...

[Anux]

To my understanding Zen 4c is not an efficiency core in the sense of energy efficient but area efficient. Server CPUs with Zen 4 are going to have 96 cores and the 4c versions will have 128 cores. But 4c includes SMT and probably won't be mixed like in intel big.little.
If we ever see them in desktops? Hardly, but maybe in the next consoles.

[ddriver]

I guess you're referring to Zen 4C as "Zen 3+". If they want to take advantage of what Intel's doing, we could see a mix of chiplets such as Zen 5 + Zen 4C.

Your analysis of the performance misses that Zen 4 will compete directly with 13th gen Raptor Lake which will bring its own single-threaded increase, and Intel is planning to rapidly release 14th gen Meteor Lake after that. There was only an ~8 month gap between Rocket Lake and Alder Lake.

Yep, it is closer to 3 than to 4 imo. I'd rather have only one type really. The difference is not as pronounced as with intel's p and e cores, 4c will be quite strong, and double the cores will more than often outpace half the zen 4 cores. No point of inviting scheduling complexity unless amd introduces something really low power.

Again - this is basically intel's best case, and not necessarily something the new revision might improve upon further, as it clearly has other deficits which need to be addressed first. Remember that 12 gen struggles against 1.5 years older zen 3 computationally. I think that ultimately, amd will have stronger overall gains, where it really matters for hpc. 50 additional FPS if you already have 500 make no difference. Getting compute done quickly and on a good power budget becomes increasingly more important as systems scale up in power footprint. 12 gen took ridiculous power to edge out zen 3 in latency sensitive tasks and still failed to eclipse old amd chips that use half the power in compute.

13 gen might manage to get a slight edge in ST over zen 4, but I do expect the latter to reign supreme for productivity.

127223.png
Intel scores well in short burst and small memory footprint tasks, but when it comes to chugging out production data, 12 gen is nowhere nearly as fast as everyone pretends to be. 13 gen will need a good 10% gain to just edge zen 3, and zen 4 will probably be around 20+% faster. So a few pyrrhic victories for intel 13 gen.

[jaxa]

I wonder if they are gonna find some nifty enterprise use for the igpu as well, servers could use some baseline gpu support as well, plus it can accelerate certain workloads on soc level.

Next step - stack 16 gb hbm on top of the io die, huge soc level wide fast memory pool without extra footprint - still fits am5. That plus the 64 gigs of L3 on the compute dies. Should be thermally feasible too. And they get to claim the world's first 16 gb cache cpu, even if not sram, there are still major performance and power efficiency gains, and you get to spill over to system ram, and significantly alleviate the strain on the narrower ddr5 controller.

Pretty soon they will also have "zen 3+" 16 core chiplet at 5nm for am5, which I guess will do pretty well against zen 4 in most use cases. So technically, amd could do as much as 32 zen3+ or 16 zen cores plus more of a mid-ranger igpu to am5. But again - only with hbm. And it will be a bit constrained at 170 watts, they will have to balance the power focus the same way they do on 17 watt power contained mobile apus.

I'm hearing that we will see separate ML, neural, and other accelerators rather than the iGPU being repurposed. For example the Tensilica Vision Q6 and C5 DSPs in Rembrandt, but even more of that with Zen 5 and later.

Big HBM L4 cache on the I/O die (which gets less hot) would be nice to see, even if only for halo products like a 5950X/7950X equivalent. One way Intel has made its more expensive CPUs look better is by enabling more L3 cache per core as you go up the stack. The same could be done with L4 cache. Buy a 6-core, you get 2 GB of HBM, buy a 16-core, you get 8-16 GB.

I guess you're referring to Zen 4C as "Zen 3+". If they want to take advantage of what Intel's doing, we could see a mix of chiplets such as Zen 5 + Zen 4C.

Your analysis of the performance misses that Zen 4 will compete directly with 13th gen Raptor Lake which will bring its own single-threaded increase, and Intel is planning to rapidly release 14th gen Meteor Lake after that. There was only an ~8 month gap between Rocket Lake and Alder Lake.

[Anux]

It's not IPC but Cinebench performance. That's the reason they are marketing "single-thread uplift" not IPC like they were previously.

You're right, i totally missed that. And seeing 5800X3D being slower in Cinebench therefore having no gains from cache it leaves room for other code to be substantially faster.

Intel has been on target with their performance numbers.

Yeah if you didn't miss any of their footnotes and knew about the stuff they "forgot to mention" like comparing mobile chips with double the power consumption that will get heavily throttled in real world notebooks. Or compairing it to a 4 year old chip while having already 4 iterations between it.
Always trying to misslead their customers.

I also like the idea of on-board graphics too rather than have special 'g' chips for the purpose.

Yes, having a small GPU on die is really helpfull it just needs enough power to drive 2 4k Monitors + a little head room and provide video acceleration. But the G chips are also great, I'm using 2400G and its amazing how many indie or older games you can play on it. Having a powerful graphics card on board opens many doors in the silent PC, SFF and HTPC area.

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.

If you're not extremly rich you shouldn't buy a new plattform at release, especially not in the current market situation. The new boards will hover around 200 €, DDR5 is double the cost of DDR4 with no performance gains and the new CPUs won't be on the cheap side either.
I always buy one generation older or atleast wait till the sucsessor is anounced. The performance difference isn't that big between generations and you can easily get a higher grade model of the older generation to make up for that. Meanwhile prices are lower and you get better linux support.
Having no pressure to upgrade I'm waiting for 5700G to drop to around 200 € or get a decent used deal. That might last me for another 7 to 8 years and saves me from buying 32 GB DDR5 and an expensive mini-ITX board while at least doubling the MT performance to the 2400G.

I also note AMD is leaning heavily into Zen 3 and even Zen 2 still for low end products. 8-core design and good yields are just killing AMD's low-end desktop offerings. I guess that AMD's response to 'Where is A620?' would be 'Why would we launch that when there are no products for it?', but I think that lack of products is a negative development. With increased cadence in development cycle, they will probably have Zen 3 as low end when Zen 5 is out, and the customer just isn't well served by that.

I think they make use of the older designs, to have more capacity for their newer designs. If everything they sell is comming out of 5 and 6 nm fabs only the output would be much less. You also don't buy a 50 € CPU to pair it with 200 € boards and 100 € RAM.