How much does a application crash - or worse - system crash factor into people whose workstation is their life?
This will be a difficult processor for almost anyone to justify. It's more of an academic achievement and less a realistic option. I think if they had at least provided ECC support instead of fusing it off, and getting at least one manufacturer to produce a "niche" ECC supported board, you could somewhat justify it.. but I just can't see the value here.
I'd imagine most people at this level of money are going to be wanting ECC if they are serious about protecting their source of income, and that means that all i9's are removed from the equation in favour of a Xeon W vs TR vs Epyc argument. Maybe the Xeon scalables as well, maybe not.
Yeah I would agree with that AkG. I was kind of jazzed up about TR when it was on the horizon but in the end the utilization of the cores isn't there, as you said, and isn't likely coming for a couple years if I was being opptomistic. If the underlying architecture, i.e. the motherboard/chipset/features, weren't changing so quickly too then maybe there would be an argument for attempting to future proofing some. But if the CPUs keep requiring us to change motherboards every few years (or less) then you're absolutely right; why plunk down that much on something only to feel the itch to upgrade again in two years anyway?
This is kind of me talking through my next move. Part of me really wanted to enter the HEDT world but the reality is the next few sets of CPUs that aren't technically HEDT in the normal sense will be more than enough for the majority of what I do. Moving to one that is 6-8 physical cores will be upgrade enough, especially considering the 4790K is still no slouch.
You guys are always so responsible! You never talk me into spending more money! lol
Interestingly though I do believe both of those consoles are 8 core systems (two quad-core modules)... So your statement might actually support us having 6 or 8 in the PC world. Hyperthreading aside of course.
While none of us can predict the future, there's a few things we can consider specifically that suggests a higher core count computer won't prove any benefit in the short term.
1) We have not seen significant improvement in PC ports or quick DX12 implementation due to AMD hardware being used in consoles. Developers are slow to implement new technologies, and are rarely given the extra time and resources to ensure features make it to PC versions.
2) Consoles are no longer lasting as long as they use to. It seems we are in a period where they are shifting to a "refresh --> next gen" cycle. In this shift, perhaps the addition of cores may have more to do with marketing than considering longevity of the platform. Depending on how you interpret this point, you could see it as a contradiction with 1), or further proof that developers won't necessarily care about the extra resources afforded to them - not yet anyway.
3) There are legacy and current games that do no like processors with high core counts. This will mean you will have to set processor affinity, or use alternative methods like the TR "Gaming" mode which halves the usable core count. I don't consider this a huge issue, in fact it may not affect you at all - but it's something that may come up for you, and you'll have to decide if this additional hassle is worth having a computer that may not ever have it's full core count used in games before seeing replacement.
From my own perspective, the major advantage of these rigs outside of the obvious workstation scenarios is consolidation of separate rigs into one. You could potentially setup a hypervisor with a VM specific to gaming, and VMs for other uses (HTPC, servers of various sorts, etc). Is it worth the trouble and loss of flexibility over seperate rigs? No :haha: