Xeon == longer lifespan

C

chiu

Limp Gawd
Joined
Sep 14, 2011
Messages
344
Some people claim that Xeon processors are more durable than the i7 counterparts. I seriously doubt it.

Some also claim that Xeon processors are more reliable. I also doubt it.

Can someone have experience to support/against these claims?
 
Xeons go through a more advanced testing process in order to satisfy server OEM's but that's it. There are no details on the amount of rejects from Xeon testing vs consumer chips.

Same silicon ,same process can't see them being "more durable" but since the bulk of them run at lower clocks with strict thermal envelopes there's a lesson there.

Besides which, finding a dead non-OC'd CPU is still pretty rare no matter how old it is. Even people who have had dead stock CPU's (does happen) are often victims of bad motherboard power circuitry without realizing it.
 
Simplyfun said:
Xeons go through a more advanced testing process in order to satisfy server OEM's but that's it. There are no details on the amount of rejects from Xeon testing vs consumer chips.

Same silicon ,same process can't see them being "more durable" but since the bulk of them run at lower clocks with strict thermal envelopes there's a lesson there.

Besides which, finding a dead non-OC'd CPU is still pretty rare no matter how old it is. Even people who have had dead stock CPU's (does happen) are often victims of bad motherboard power circuitry without realizing it.
Click to expand...

Very true. It's not unheard of to see cpu's from 20+ years ago still in service, from the consumer market.

They downclock the xeon chips for stability, and attach a longer warranty. The die and silicon are the same.
 
doesn't matter if they are xeon or i-7 as long as they aren't abused or faulty (components) they all should run for many years passed the warranty.
 
Nah, Xeon's are typically a higher bin than their desktop equivalent (if a desktop equivalent exists), but that's about it. Same silicon, both will last forever.
 
Xeons are binned for lower power and lower voltages, so theoretically they will last longer. However, the difference in lifespan for the life cycle of processors is rarely ever relevant.
 
The benefit of the Xeon is their low power, low voltages, and generally lower base clocks. The reasoning is to keep the heat profile down. The less heat they produce the less likely they will be to wear down in a tightly fitted server with minimum airflow. If you were to try and subject a consumer chip to the same conditions as the Xeon chip in a server cluster environment, it may be prone to failing sooner, but there is no guarantee. Also, nowadays Xeon chips are geared more towards multi-functionality, providing more cores and threads to meet the needs of a server.
 
I am running both LGA 2011 xeon E5-1650 (pretty much the same as an i7-3930k) and i7-3930k Sandybridge. I also have both xeon E5-1650v2 and an i7-4930k Ivybridge
All those CPUs are unlocked hexcores and they overclock real well, they run pretty close to each other as far as voltage and temps go.
All of them have been holding up well for me running BOINC 24/7 100% load. I use AIO 280mm coolers.
I find the xeons cheaper than the i7 used on ebay.
Even my LGA 1366 xeon x5660 and w3680 hexcore overclocked very well and run cool.
Not one CPU i7 or xeon has died on me yet.
 
Xeons dont last longer. They are just slightly better binned.

Intel seems to operate with 25 year lifespans or longer.
 
It's all luck of the draw. I have a q6600 still running from launch at 3.6 and a i7 920 2nd gen still running at 4.0ghz. Proper cooling, good supporting components with good clean voltage leads to longer life.
 
I've never had a CPU fail in the 23 years I've been building PCs'.
 
Zepher said:
I've never had a CPU fail in the 23 years I've been building PCs'.
Click to expand...

Well I can't say that.

I probably fried 15 Manchester/Toledo/Windsor cores using cascades and mousepots back in the day chasing various world records. Fried dozens of RAM sticks too. Who knows how many Dothans bit the dust chasing the SuperPi record. They weren't all failures, but man this can be an expensive hobby.

In terms of regular use I had an Opteron in a quad set baff out once and one P4 type Xeon bite the dust. Nothing since.
 
Zepher said:
I've never had a CPU fail in the 23 years I've been building PCs'.
Click to expand...


You aren't trying hard enough then... lol

I've killed a few in the attempts for my megahertz(megahurts?? LOL). Never killed any built for duty purposes though.

I had a resister on a AMD Slot A 650 pop off and hit me square in the eyeball while trying to push to limits. These were the card-type CPU's, with heat plate on one side and a plastic shell on the other(shell removed of course to use gold-finger device and to mod the card).

Cooked a couple Pentium 233's, some Socket A Athlons, a Coppermine PIII, K6-2 III, a Pentium 4, and probably a few others. All of course due to trying to push their limits.
 
4330thgink said:
You aren't trying hard enough then... lol

I've killed a few in the attempts for my megahertz(megahurts?? LOL). Never killed any built for duty purposes though.

I had a resister on a AMD Slot A 650 pop off and hit me square in the eyeball while trying to push to limits. These were the card-type CPU's, with heat plate on one side and a plastic shell on the other(shell removed of course to use gold-finger device and to mod the card).

Cooked a couple Pentium 233's, some Socket A Athlons, a Coppermine PIII, K6-2 III, a Pentium 4, and probably a few others. All of course due to trying to push their limits.
Click to expand...
ones like these?
procs.jpg
 
chiu said:
Some people claim that Xeon processors are more durable than the i7 counterparts. I seriously doubt it.

Some also claim that Xeon processors are more reliable. I also doubt it.

Can someone have experience to support/against these claims?
Click to expand...

Only reason I can think of is that, at least in comparison to a regular consumer i7, a Xeon is a little more physically robust, with a metal TIM between the die and the IHS. It's also physically larger.

They're pretty much identical to Haswell-E/Broadwell-E i7s, though, physically.

As Zepher said, though, I can't remember ever actually killing a CPU of any type. Even back in the bare die Pentium 3/Athlon days, the worst that ever happened is that I had to reset the BIOS.
 
jmilcher said:
Very true. It's not unheard of to see cpu's from 20+ years ago still in service, from the consumer market.

They downclock the xeon chips for stability, and attach a longer warranty. The die and silicon are the same.
Click to expand...

The Xeon is downclocked so they consume less power and generate less heat. This is essential when placing them in higher density environments in large datacenters. Raw clock speeds are less important in those usage scenarios where these CPUs are being used in multi-threaded applications such as SQL servers or in virtual infrastructure where they need a higher core count. Power consumption and being able to cool tons of these things in blade enclosures or high density racks isn't a trivial matter. It's about catering to their customers needs, nothing more, nothing less. Stability has nothing to do with it.
 
RazorWind said:
Only reason I can think of is that, at least in comparison to a regular consumer i7, a Xeon is a little more physically robust, with a metal TIM between the die and the IHS. It's also physically larger.

They're pretty much identical to Haswell-E/Broadwell-E i7s, though, physically.

As Zepher said, though, I can't remember ever actually killing a CPU of any type. Even back in the bare die Pentium 3/Athlon days, the worst that ever happened is that I had to reset the BIOS.
Click to expand...

I don't know what you are talking about. LGA2011 Xeons are the same size as their Core i7 counterparts. All HEDT processors have soldered IHS's and not a shitty TIM.

Master_shake_ said:
cpus fail?
Click to expand...

Yes they do, although not very often. I've seen probably half a dozen Intel CPU's fail over the years. I've seen probably a dozen bad AMD CPUs. Keep in mind that I've been dealing with computer hardware professionally for 20 years. I've built and serviced thousands of PC's in my time as a hardware technician in service centers and working in the IT profession. Failures are extremely rare, but they do in fact happen.
 
A lot of Xeon workstation and server kit doesn't allow for much or any overclocking so the chips will run totally stock for most if not all of their operational life. Plus the times they are switched on and off could be far lower than domestic machines.
 
You must log in or register to reply here.
Back
Top