-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
What is VID?
- Thread starter LOUISSSSS
- Start date
Tempered81
Diamond Member
i think that means Voltage identification. Isnt that what your actual voltage is after your vdroop? so if you set bios vcore to 1.28, maybe 1.2125 is your actual?
i may be wrong about this, wait for the next answer~ 🙂
i may be wrong about this, wait for the next answer~ 🙂
Tempered81
Diamond Member
yeah i'm wrong about that, just read where a guy has a g0 q6600 at 3.8ghz 1.395 volts on the core, but he says 1.2125v VID. (cant have THAT much vdroop) its a specification i guess.
EarthwormJim
Diamond Member
It's the default voltage for your specific processor. It can vary from processor to processor, even of the same series.
In the old days when an IDM binned their chips for clockspeed based on the shmoo plot (voltage versus clockspeed) the IDM would simply use a "one size fits all" voltage for a given product SKU.
This made things like TDP per SKU more managable in addition to cutting down on the tester time. But the tradeoff is it was on the high-end of the safe side, many chips simply did not need that "one size fits all" voltage to operate at the rated GHz, thus they ran hotter than necessary.
What Intel fid with VID and their 65nm Core architecture was throw in the extra test time necessary to determine the next best approximate voltage (something less than the maximum allowed) for the chip to function at its clockspeed.
This basically means each chips individual shmoo plot is determined and the chip is then told "you will operate at multiplier X and at stock voltage Y".
Thus some Q6600 have VIDs of 1.2500V whereas others have a VID of 1.3250V...meaning the chip with a VID of 1.2500V had a shmoo plot for which operating frequency of 2.4GHz was safely acheived with a mere 1.2500 Vcore while the other chip's shmoo plot indicated it required 1.3250V in order to reliably function at 2.4GHz.
By the way if a chip's shmoo plot indicates it needs >1.3250V to reliably operate at 2.4GHz then it won't be binned as a 2.4GHz chip...they will go down the shmoo plot to say 2.2GHz and see if it operates at that speed with an acceptable Vcore.
(more or less, the numbers here may differ from Intel's exact procedures, but the gist of the process is what I have described here)
MadScientist
Platinum Member
Graysky has a similar thread on the Bleedingedge forum. Link In the last post of the thread he states:
" I'm not totally sure the integrity of the data collected is that high. What I mean by that is I have read several reports of different reported VIDs for the same chip on different boards. I have also read about the VID changing based on the speedstep state and other factors.
I started this thread hoping to see some sort of correlation between VID magnitude and vcore @ a given o/c level. I have received mixed reports on this front as well. I think the bottom line is there isn't a correlation between VID and overclockibility."
The VID on my Q6600 on an Abit IP35-E board is 1.2375V. I need 1.3575vcore to be stable at 3.600 ghz
There is a correlation as the VID actally comes from the chips shmoo plot (Vcore vs. GHz), but the inherent problem with "polling the masses" for such data is that all your results are convoluted with everyone's unique hardware setup.
Different chipsets, different ram, different cases, different case cooling, different ambiant temps, different HSF's, different thermal pastes.
In fact had you got anything remotely resembling a stastistically valid trend from a poll then I personally would be completely skeptical.
The second reason you will only "at best" establish a loose correlation between VID and the likelihood of reaching a specific overclocked speedbin is that every chip will have its own shmoo plot...they are all similiar but they are not identical.
(This is also the inherent problem with review sites that tout the overclockability of a single chip sample as being indicative of the overclockability of all chips with the same SKU.)
The change in voltage to enable a change in clockspeed (i.e. the slope of the shmoo plot) is unique for each CPU depending on the process variations that occurred in the fab over time. Like snowflakes, no two chips are eaxactly identical but they can be very very similiar.
So am I right in thinking that all four of these chips are identical but the E2200 is binned higher because it has been tested @ 2200 MHz (with higher base voltages) and passed those tests ?
Intel Pentium Dual-Core Desktop Processor E2140: VID = 0.85V ? 1.5V
Intel Pentium Dual-Core Desktop Processor E2160: VID = 0.85V ? 1.5V
Intel Pentium Dual-Core Desktop Processor E2180: VID = 0.85V ? 1.5V
but this new bad boy has higher VID ...
Intel Pentium Dual-Core Desktop Processor E2200: VID = 1.162V-1.312V
-
Intel Pentium Dual-Core Desktop Processor E2140: VID = 0.85V ? 1.5V
Intel Pentium Dual-Core Desktop Processor E2160: VID = 0.85V ? 1.5V
Intel Pentium Dual-Core Desktop Processor E2180: VID = 0.85V ? 1.5V
but this new bad boy has higher VID ...
Intel Pentium Dual-Core Desktop Processor E2200: VID = 1.162V-1.312V
-
Tempered81
Diamond Member
is there a program that will read the VID, like everest or cpu-z?
MadScientist
Platinum Member
Tempered81
Diamond Member
where do i see vid on that, i have coretemp open now
nm, its one of only 4 or 5 outputs, and i see it
now. 🙂
nm, its one of only 4 or 5 outputs, and i see it
now. 🙂
VID is not the sole binning characteristic looked at during test. There will literally be hundreds if not thousands of metrics that are stressed and tested to ensure the chip has excellent likelihood of functioning 10 years from now.
Too high of a VID is just one reason of hundreds of reasons a chip migh get binned to a lower clockspeed.
The rather large range of acceptable VID's per your reference to Intel's SKU specs just shows that Intel gives itself lots of latitude for determining the VID for these chips.
Mobile chips for laptops and the like will of course be much more aggressively binned for lower VID at a given clockspeed. Whatever pays the bills.
This may sound really stupid but I've just got to ask about this binning process. Intel (or whoever make them for Intel) must churn out millions/billions of CPU chips in a year. How does this speed testing and binning of CPU chips take place. I can't believe that each individual CPU chip is tested for speed, volts, etc then binned accordingly, that would take forever. So how do they do it. TIA
Yes they are, and yes it does take a long time.
"Test" is a rather expensive portion of manufacturing an IC because you basically have to invest 10's if not 100's of millions of dollars in capex in order to buy enough "testers" to run in parallel to handle moving your volume of product.
There is an entire community (economy if you will) working on making "test" cheaper and more efficient. Intel is a heavy investor in the test community.
Be aware though this is not unique to Intel. Every chip manufacturer must test every single chip (they may do it at wafer level though, but still every chip is tested before being cut from the wafer).
Test time correlates to chip complexity (number of functions that need to be verified as functional) and thankfully so to does market price (generally) so the more expensive chips that take longer to validate also fetch higher prices so it all evens out.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 25K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 24K
-
-
