Do Xeon CPUs have a high or even the highest binning?

Macusercom

Reputable
Jul 9, 2015
7
0
4,510
I just tried to undervolt my Intel Xeon E3-1231v3 processor and could get it to run stable at -120mV offset.

I just wondered if I'm just lucky or if Xeons are generally higher binned than other CPUs. Moreover, I'm wondering if i7 CPUs have the same binning or if Xeons have a higher than i7 or the other way round.
 
It depends on the chips. Xeons likely go through a slightly different fab process since some have igpu's while others like the 1231v3 don't. Chips are binned within a 'family', ie a 4690k is higher binned than the 4690 since it's designed to be pushed harder than stock. An i7 4790k is binned higher than a 4790 and comes with not only an unlocked multiplier but higher stock speeds. Since xeons don't have unlocked multipliers, the higher binned within the same family will have higher clock speeds.

If you look at the xeon 1200 v3 family, for the chips released lately (q2 of 2014) in order from fastest (highest binned) to lowest -

1286v3 3.7ghz igpu
1281v3 3.7ghz no igpu
1276v3 3.6ghz igpu
1271v3 3.6ghz no igpu
1246v3 3.5ghz igpu
1241v3 3.5ghz no igpu
1231v3 3.4ghz no igpu
1226v3 3.3ghz no igpu

These are strictly the 2014 (most recent haswell) e3 v3 xeons, there are others in between from 2013. There's only 1 xeon, the 1226 binned lower than the 1231v3 meaning it didn't make the cut for the 5 models above it. It also doesn't include the low power L versions. The highest speed xeon here in the list is 3.7ghz, the 4790k starts at 4ghz so I would say the i7 is higher binned. What else goes into the xeon fab process different from the i7 I'm not entirely sure, I'm thinking they're different lines. The highest quality wafers go to the fastest chips. Binning them sorts them based on their test performance in the lab, so the A grade wafers go in the 1286 and on down and down until they reach the 1231 and the chips that didn't meet 1231v3 requirements were placed finally in 1226v3.
 




That isnt exactly how the binning process works. a lower clocked xeon could be binned higher than a i7.
 
"The best chips are generally binned as higher-end parts, being sold as not only the fastest parts with their full caches enabled, but also the low-voltage and ultra low-voltage models. Note: Based on market demand, these highest-end chips can also be sold as lesser chip parts." - "Chips which do not perform as well as the best chips are often sold for lower clock speed models"
http://www.geek.com/chips/from-sand-to-hand-how-a-cpu-is-made-832492/

"In order to undergo binning, manufactured products require testing. Finished products enter a machine[3][4] that can test hundreds of pieces at a time, taking only a few hours to complete. Each piece can be tested to determine its highest stable clock frequency and accompanying voltage and temperature while running"
https://en.wikipedia.org/wiki/Product_binning

"Another part of it, though, is that Intel fabricates millions of dies for use in all manner of computing platforms. When you deal with those sorts of numbers, the manufactured components are going to have a wide spectrum of performance tolerance, at which point (to increase yield per wafer / ROI) off-spec components are binned into higher/lower spec products. If Intel has a bin filled with underperforming dies that "didn't make the cut" for a high-end product, they can increase yield-per-wafer by shipping those in lower-end products that don't require the same level of performance.

Intel's spec for K-SKU processors demands higher thermal, frequency, and voltage tolerance than—for instance—their power-saving S-line. In general, this means that consumers purchasing K-SKU products will have an objectively "better" die housed under their IHS, being that it's been tested and qualified for more abusive computing environments. Intel knows their K-SKU devices will be marketed to overclockers, and thus selectively uses higher-quality components for them."
http://www.gamersnexus.net/guides/1140-silicon-die-bin-out-process-explained

Binning doesn't always mean higher vs lower quality. Binning is 'sorting'. It can be based on any number of criteria. Stability, raw speed, min power consumption etc. I would imagine cpu's go through several phases of production, where so long as it's haswell for instance, an i3/i5/i7 may go through similar phases. Then at the ht phase, i3's, i7's and xeons go through a common process and i5's skip that step. For igpu's, i5's, i7's and some xeons will go through that step and other xeons won't.

As far as one product being more 'durable' than another, some of it is marketing. I can take the same identical part as a consumer grade and call it 'enterprise' by slapping an extended warranty on it and charging more. Doesn't necessarily mean it's different. Perfect example, gigabyte motherboards. The ud5h and ud3h both have a 'black edition'. The black edition costs more money than the non black edition. They are 'certified' that they go through 168hrs of testing (exactly 7 days) prior to being packaged and shipped. They also come with a 5yr warranty vs 3yr warranty.

Aside from the color schemes, the products are identical. Just because one was tested and passed doesn't mean a non tested one won't pass, that assumption can't be made. So someone can say yes, but they're extending the warranty of usefulness so they must be confident in that part.

They don't have to be, warranty costs are factored into every product. Like auto insurance, everyone pays, few will actually need to draw on it and it can actually be a revenue stream for the company. In addition they charge more. Most people aren't returning the mainstream model, much less the bk edition - for those that do, no skin of gigabyte's nose, they've collected enough extra to just pack up and ship out another board.

In the case of xeons, they're touted as being durable or more durable - than what? How many people, many of which use their cpus constantly, leave them powered on, running torrents, serving files, running a plethora of @home type software that's constantly computing - have had a plain old desktop chip just wear out? In a sense it's marketing to sell at a higher price with no proven sense of need or urgency. And if a xeon did die, they've already collected so much extra profit from all their chips intel won't miss even a $2000 high end xeon. Simply slap a stamp on it. Not to say chips can't fail, anything can fail. But it's already a rarely seen phenomena in the 'boring ol retail' market much less the workstation arena. I also highly doubt these workstation chips are server grade, not many are going to put a 1231v3 in a heavy duty server. That's what the $2-3k 8-12+ core xeons are for.
 
If you'd read what I had to say instead of hurting yourself trying to be clever, you would have noted where I said "Chips are binned within a 'family'... If you look at the xeon 1200 v3 family, for the chips released lately (q2 of 2014) in order from fastest (highest binned) to lowest" from my first post you would have saved yourself the energy RobCrezz. Where among the xeons listed the 1231v3 is a budget xeon. Always has been, always will be. Next to last as a matter of fact.

Being based on haswell architecture with shared similarities to the i7 as in same cpu core tech, same size cache, hyper threading and so on it would make sense the two are produced on the same or similar lines. If you'd read (or even understood the concept of 'binning') it's sorting. It's not necessarily better/worse. One 'bin' may be lower voltage tolerances while the other 'bin' is core speed. Obviously the xeon 1231 would fall into a lower power bin than the 4790k while the 4790k would fall into its' bin due to higher speed tolerance than the 1231 can handle. Which is 'better' and by what metric are we determining 'better'? Different would be a more accurate statement.

Again, you're just rebreathing what I already 'copy/pasted' which said exactly that - that binning can be for lower voltage as well as clock speed. Second post, first sentence, second line I do believe. Nice read.