MP 1,1-5,1 Realistically, would I even notice an X5680 upgrade?

[eicca]

I'm currently rocking dual E5620s, which configuration scores around 450 single and 3000 multi on Geekbench. The dual X5680 setup benchmarks around 600 and 6000. That's 1.5 times more power for the simple tasks, and twice as much power for heavy simultaneous workloads.

Here's the thing. The hardest I make this computer work is usually in Logic, and I have yet to surpass 30% CPU usage in that case, so increasing my multicore performance is unnecessary for what I do.

Would I notice the boost in single core speed though? While this thing never flinches at multitasking, I do notice it slow down when doing stuff like resizing windows with lots of elements in them or while scrolling through album art in iTunes or while browsing complex web pages. It's really the only time I'm reminded that this computer is quite old. Would the dual X5680 upgrade smooth any of that out?

It's a cheap enough upgrade, it might tide me over until I can get an M1 machine or whatever the next gen M chip will be.

 

[tsialex]

I'm currently rocking dual E5620s, which configuration scores around 450 single and 3000 multi on Geekbench. The dual X5680 setup benchmarks around 600 and 6000. That's 1.5 times more power for the simple tasks, and twice as much power for heavy simultaneous workloads.

Here's the thing. The hardest I make this computer work is usually in Logic, and I have yet to surpass 30% CPU usage in that case, so increasing my multicore performance is unnecessary for what I do.

Would I notice the boost in single core speed though? While this thing never flinches at multitasking, I do notice it slow down when doing stuff like resizing windows with lots of elements in them or while scrolling through album art in iTunes or while browsing complex web pages. It's really the only time I'm reminded that this computer is quite old. Would the dual X5680 upgrade smooth any of that out?

It's a cheap enough upgrade, it might tide me over until I can get an M1 machine or whatever the next gen M chip will be.

If you don't need the multicore performance, go for a pair of quad core X5677 Xeons, it's the fastest single core Xeon that works with a MP5,1 and the turbo speeds are kept for much longer than with a 6-core X5690 or with a X5680. Btw, the X5677 pair will be a lot cheaper than a pair of X5680s.

A dual CPU X5677 Mac Pro is a lot snappier than the dual E5620 and for the small price, it's a extremely valuable upgrade. If your memory is not 1333MHz, almost all Apple memory from factory is already 1333MHz, change it too. It's one more small upgrade, but it's one that you will notice.

About really noticing the upgrade or not, I've learned years ago that you rarely feel these not so dramatic upgrades, but the downgrades are always extremely painful. So, use your upgraded Mac for some days/a week, then go back to the original processors for a day, you will notice it instantly.

 

[eicca]

Now that is some prime advice good sir. Those X5677s are dirt cheap. I can justify that quite easily. They'll still net me a 50-60% bump in multicore performance too!

I think I'm running 1066MHz RAM too, so I can justify upgrading that with all the money I'm saving on CPUs.

 

[w1z]

it's the fastest single core Xeon that works with a MP5,1 and the turbo speeds are kept for much longer than with a 6-core X5690 or with a X5680.

Do you have a reference?

From what I know, the w3690, x5677, x5690 can sustain 3.73GHz on 2 cores only and 3.6GHz on the remaining cores

 

[h9826790]

Do you have a reference?

From what I know, the w3690, x5677, x5690 can sustain 3.73GHz on 2 cores only and 3.6GHz on the remaining cores

No, 3.73 only avail "when two cores are active".

If the remaining still running at 3.6GHz, they are not inactive.

 

[w1z]

No, 3.73 only avail "when two cores are active".

If the remaining still running at 3.6GHz, they are not inactive.

View attachment 1697422

I didn’t get your explanation nor shared chart.

With HT disabled, a max of two cores can sustain 3.73GHz while the remaining cores run at a max of 3.6GHz. It’a a different thing all together with HT enabled.

I have tested this under Windows when attempting to squeeze the most single core performance setup for gaming.

 

[tsialex]

Do you have a reference?

From what I know, the w3690, x5677, x5690 can sustain 3.73GHz on 2 cores only and 3.6GHz on the remaining cores

There is a thread about our findings/tests back in the day somewhere here. My point with my previous post was that a X5677 with the exactly same die as a X5690, the greater T-case Max (80.4ºC vs 78.5ºC) and with the two cores disabled acting as heatsinks/heat spreaders, the TBT is not so constrained by the temperature and the frequency boost can be sustained over longer periods.

 

[h9826790]

I didn’t get your explanation nor shared chart.

With HT disabled, a max of two cores can sustain 3.73GHz while the remaining cores run at a max of 3.6GHz. It’a a different thing all together with HT enabled.

I have tested this under Windows when attempting to squeeze the most single core performance setup for gaming.

I never test that with HT off.

Can you share any test result which can indicate 2x 3.73 + 4x 3.6 running at the same time?

 

[eicca]

X5677s are in, along with 24GB of 1333mhz RAM. True to the single core benchmarks the computer feels pretty much exactly 50% faster in single-core things like scrolling YouTube or resizing complex windows. Would I have paid $150 and up for the X5680 or -90s to get this improvement? Oh heck no. But $35 for the X5677s feels totally worth it, and that money saved paid for the RAM.