How you like your 2696 v4? i was thinking about this cpu too. Good to know it works in the dell 5810. It's an oem part so the compatibility is not as wide as the non oem regular 2699v4. The 2696 is cheaper than the 2699 by at least $40. But i think the 2696v4 will probably work in my asus Z10 MB.When I replaced a 160w 10-core E5-2687w v3 with the 150w 22-core E5-2696 v4 in my Dell 5810, I noticed temps went down substantially (80ish to mid 60s) on virtually any stress test or benchmark while performance was always better. Just a bit better in single core work but lots better most of the time. Always thought the v3 was thermally constrained in the workstation. Before I replaced it, I wondered if it could handle the 22 core heat. Turned out, it runs far cooler now, probably saving a bit on the electric bill. You'll pay more for the top-end v4s so you'd need to use them to max capacity to get your money's worth (except ...reasons.) At the lower end (TDP < 135w), the v4s are essentially free so it's a no brainer to get one rather than the similarly free v3s.
The 2696v4s are good fits for my Dell workstations - only a bit slower than the fastest 8-core 1680v4 on will turbo on 1-2 core workloads (4GHz.) Steppings allow the 2696v4 to run 1-2 core workloads at 3.7GHz, while it turbos all cores at 2.8. You can find v3s with 18 cores and similar speeds. They'll throttle far quicker in OEM workstations so you can't take advantage of the turbo speeds with the v3 to the extent you can with the v4.How you like your 2696 v4?
Most v4s will run 2400MHz memory while the v3s top out at 2133. For some odd reason, my Dells frequently revert to 1866 when 2133 sticks are installed. For bandwidth-constrained applications, those 2400MHz sticks are a 20%+ improvement.Another factor ("reason") with the high core-count parts is the large L3 cache. Can be a gane-changer for certain use-cases (for times when its role is a workstation w/single(or low)-thread load--and 55M >>> twice-27.5M. Capiche?
I consider it a rite of passage in the steps towards homelab.
One of the advantages of a big homelab is that, if you have nothing to do with it you can at least take out all the memory sticks and stack them around your desk and pretend you are some kind of rdimm Scarface.I consider it a rite of passage in the steps towards homelab.
To max out specs before later getting to the "what do i actualy use" and picking spec appart again.
Not long since i removed another tb ram from my lab since its just sitting there using power.
Just went the boring route of getting back what i paid for the ram + some profit.One of the advantages of a big homelab is that, if you have nothing to do with it you can at least take out all the memory sticks and stack them around your desk and pretend you are some kind of rdimm Scarface.
I wanted to compare it mini PC's after reading that so here's the same comparison with that Beelink's CPU. It isn't fair since you're getting the GPU, AVX-512 and everything with the Beelink but ECC, more lanes, and everything with the E5 V4. It's still overlooked that modern desktop processors are like 28 core E5 Broadwell. I've switched to mostly TMM at home and I'm saving ~$80/mo in electricity now.My 2 cents: this is an OK deal for these specific CPUs, but bad value anyway.
Check this comparison between the 2696 v4 and 2680 v4; the 14 core 2680 v4 is ~10% slower in single core (thanks to the 300MHz lower single core turbo) and ~27% slower in multicore (despite having only 64% of the cores of the 2696 v4). 2680 v4s are incredibly cheap at the moment; you can get them on US eBay from either US or Chinese sellers for only 20-30 USD each.
Unless you actually need the extra performance or you just want to have more cores for...reasons, then the SKUs at, or near, the top of the pile are almost always bad value, even on older platforms like his.