The point about Power 10 currently

[MauryG5]

Friends of the Power community, I think the time has come to take stock of the Power 10 situation because in my opinion at the moment there is a bit too much impatience combined with general confusion. So currently, if you have read everything correctly, it is clear that Power 10 was launched by IBM but only and exclusively for the 1080 series servers and we all know that they are the high-end ones. Furthermore, it has been repeatedly observed that the processors currently in use by these servers, obviously are the high-end ones, ie with a minimum of 15 cores (16 but one is turned off if necessary) and a maximum of 30 cores. Now we all know well that these numbers, also linked to the range to which these servers belong, clearly makes us understand that these are CPUs clearly of the type to understand, direct successors of MONZA and LAGRANGE of Power 9 and therefore high-end. It is perfectly obvious that currently, even if the firmware were completely free, Raptor would not be able to produce a P10 based motherboard on the spot because the costs would obviously be prohibitive. So the reasons that today cannot lead to a production of P10-based motherboards are both the current lack of free firmware but also the lack of medium and low-range solutions, equivalent to our current Sforza P9. So I would say that instead of continuing to ask Raptor when and if it will produce P10 based cards, we should concentrate on waiting for the release of the mid and low range solutions in the meantime, which I don't think will arrive shortly and later, once they start. to exit, then also see if IBM will start as I think, to release the firmware sources completely open and ask Raptor what programs they will have at that point on P10. Currently I repeat, there is nothing feasible because even if the firmware were already free, the prices are prohibitive and there are no CPU offers equivalent to Sforza which is what we all use on our Power systems for desktop. and small home servers ... Let's be patient and see what will come out maybe already in 2022 by IBM on the medium and low range and then we will start to have everything clearer!

[ejfluhr]

https://www.itjungle.com/2021/11/08/inside-the-ibm-denali-power-e1080-system/

Pictures/info here show fancy "on the substrate" connectors for external cables, plus requirement for memory-buffer-based DIMMs.
It does seem that the expense of building around these features is pretty high, not something for an end-user-affordable system.

[MauryG5]

In fact, this is what I have been saying since the beginning, currently Power 10 cannot be used in any way in a home environment, even if it were all open source! They must first enrich the offer with the mid-range and low-range versions, then in that case there will certainly be a big difference!

[ejfluhr]

Well, I do not believe that anything about that POWER10 module precludes being used in a home environment, just that IBM would have to sell it on the OpenPOWER market and Raptor or someone would have to build a motherboard to support it.  I think this is the HotChips presentation?
   https://regmedia.co.uk/2020/08/17/ibm_power10_summary.pdf

Those external-cable connectors would be for the PowerAXON interfaces which let the processor scale up to lots of sockets.   The memory and PCI interfaces wouldn't use such cables.   Much like all the POWER9 processors have lots of socket-to-socket I/O that aren't used by the Raptor systems, a bare-bones POWER10 system could ignore all the cabled interfaces and just use the ones that connect thru the socket. 

Of course, the likely answer is that the cost of such a design is still prohibitive, since it presumably has to use the fancy buffered memory DIMMs:
   https://fuse.wikichip.org/news/2893/ibm-adds-power9-aio-pushes-for-an-open-memory-agnostic-interface/

It isn't clear to me how the spring mid-range and low-end announcements will be more amenable to a consumer-focused POWER10 system, but here's to hoping that the landscape changes come 2022!   I would love to buy a POWER10 system!!

[MauryG5]

Well if you think about it, having the possibility of building a motherboard around a mid-range or low-end processor, all the costs to be incurred are considerably reduced and consequently it becomes accessible to us desktop users.  Currently the P10 is unthinkable, both in terms of costs but also in terms of spaces and exaggerated dimensions of the CPUs, which obviously are high-end and not suitable for us desktop users ...

[xilinder]

I agree.
    IMO. We POWER9 desktop and small server users have only begun to utilize the CPU and the motherboard design.
We don't know what we can do to expand these machines and possible future of motherboard design to include more
functionality for either desktop or even commercial users.

    It is just too soon to consider POWER10 as a next motherboard CPU for other than commerical use if there is a market for
such a machine.

    Let us push POWER9 just as far as possible into the future.

[MauryG5]

Exactly friend Xlinder, the situation at the moment is this, we are in the full P9 period, he has reached a good level of maturity and therefore it is time to make the most of his skills.  For example, just look at Canonical which from this year will start with dedicated P9 support on its Ubuntu ... In the meantime, the most suitable P10 processors will arrive and you can then think about something new ... But it still takes time ... Patience and time to use to make good use of P9!

[ejfluhr]

Seems unlikely these processors will make it to consumer boxes given the firmware blobs for DDIMMs & PCIe, but one can always dream...
https://www.nextplatform.com/2022/07/12/can-ibm-get-back-into-hpc-with-power10/



Agreed that P9 is still more than plenty capable, likely lots of capability that hasn't yet been tapped.

[ClassicHasClass]

I'm holding out for P11 myself.