| View previous topic :: View next topic |
| Author |
Message |
crazybubba64
Joined: 03 Jul 2018
Posts: 1371
Location: WI, USA
|
 Posted: Tue Jul 02, 2019 10:46 am Post subject: Noteworthy ES chips with unlocked multipliers |
 |
|
Every now and again someone asks "Are ES chips unlocked?", and for the most part, the answer is no.
However, there are plenty of obviously unlocked ES chips, such as the many QS press samples for various extreme editions and k-series parts, but what about the other oddballs?
I've found a few ES chips that are unlocked that aren't marked as such on CPU-World (or just simply don't exist on the site):
QBDS (CPUID) - Seems to be a theoretical E5-1640? fully unlocked.
Q1TK (according to HWInfo64) (CPUID) - seems to be a desktop lynnfield chip masquerading as a clarksfield in the PGA988 package running at 1.33-1.6GHz natively, but is fully unlocked (albeit very unstable)
Do you know of any other ES with unlocked multipliers?
_________________
My collection |
|
| Back to top |
|
 |
xsecret
Joined: 01 Feb 2004
Posts: 1846
Location: France
|
| Posted: Tue Jul 02, 2019 10:56 am Post subject: |
|
|
All post Northwood QS CPU are locked and almost all ES starting from SNB are also locked. That's for the old base multiplier.
For SNB and up, the concept of "base multiplier" doesn't exist anymore from a technical point of view. Ratios are dynamic and based on P-State. The only "lock" (for non-K/X CPUs) removes the ability to set the P-State Max multiplier beyond a pre-defined value.
_________________
ES-Only Collector : http://www.engineering-sample.com
Universal Chip Analyzer (UCA) : https://x86.fr/uca / http://www.cpu-world.com/forum/viewtopic.php?t=34349 |
|
| Back to top |
|
|
crazybubba64
Joined: 03 Jul 2018
Posts: 1371
Location: WI, USA
|
| Posted: Tue Jul 02, 2019 11:09 am Post subject: |
|
|
| xsecret wrote: | All post Northwood QS CPU are locked and almost all ES starting from SNB are also locked. That's for the old base multiplier.
For SNB and up, the concept of "base multiplier" doesn't exist anymore from a technical point of view. Ratios are dynamic and based on P-State. The only "lock" (for non-K/X CPUs) removes the ability to set the P-State Max multiplier beyond a pre-defined value. |
"almost" all being the key phrase here.
And yeah this includes "turbo unlocked"/unlocked p-state chips like the ones shown above.
_________________
My collection |
|
| Back to top |
|
|
xsecret
Joined: 01 Feb 2004
Posts: 1846
Location: France
|
| Posted: Tue Jul 02, 2019 11:24 am Post subject: |
|
|
Starting with SNB and P-State control, QS are locked or unlocked according to the commercial counterpart (K/X or non-K/X).
For ES, the same QSpec can be locked or unlocked.
1/ Some QSPEC came factory unlocked (usually ES1) but they can be software-locked later.
2/ Most QSPEC (usually ES2) came factory locked but they can be unlocked (or re-locked) later by software tricks. The unlock procedure is persistent to reboot/power off. Same for cache slices activation. So you can find the same ES2 QSPEC in locked or unlocked state.
_________________
ES-Only Collector : http://www.engineering-sample.com
Universal Chip Analyzer (UCA) : https://x86.fr/uca / http://www.cpu-world.com/forum/viewtopic.php?t=34349 |
|
| Back to top |
|
|
frag_
Joined: 17 Nov 2008
Posts: 4015
Location: Estonia
|
| Posted: Tue Jul 02, 2019 12:49 pm Post subject: Re: Noteworthy ES chips with unlocked multipliers |
|
|
| crazybubba64 wrote: |
QBDS (CPUID) - Seems to be a theoretical E5-1640? fully unlocked.
|
All v1/v2 E5-16xx are unlocked so it makes sense. |
|
| Back to top |
|
|
KK_Fin
Joined: 01 Jul 2013
Posts: 835
Location: Finland
|
| Posted: Wed Jul 03, 2019 9:16 am Post subject: Re: Noteworthy ES chips with unlocked multipliers |
|
|
| frag_ wrote: | | crazybubba64 wrote: |
QBDS (CPUID) - Seems to be a theoretical E5-1640? fully unlocked.
|
All v1/v2 E5-16xx are unlocked so it makes sense. |
AFAIK, E5-1603, E5-1607, E5-1620 and E5-1607 v2 are locked. Probably E5-1620 v2 also. Others (with 6 cores) are unlocked (like QBDS).
_________________
1242 x86 CPUs in my collection (week 2 2026). |
|
| Back to top |
|
|
crazybubba64
Joined: 03 Jul 2018
Posts: 1371
Location: WI, USA
|
| Posted: Wed Jul 03, 2019 9:40 am Post subject: Re: Noteworthy ES chips with unlocked multipliers |
|
|
| KK_Fin wrote: | | frag_ wrote: | | crazybubba64 wrote: |
QBDS (CPUID) - Seems to be a theoretical E5-1640? fully unlocked.
|
All v1/v2 E5-16xx are unlocked so it makes sense. |
AFAIK, E5-1603, E5-1607, E5-1620 and E5-1607 v2 are locked. Probably E5-1620 v2 also. Others (with 6 cores) are unlocked (like QBDS). |
I need to confirm this, but I think the E5-1620V2 is unlocked, as I believe it is a repackaged i7-4820k. The E5-1620 is definitely locked, as the i7-3820 was as well.
_________________
My collection |
|
| Back to top |
|
|
KK_Fin
Joined: 01 Jul 2013
Posts: 835
Location: Finland
|
| Posted: Wed Jul 03, 2019 10:21 am Post subject: Re: Noteworthy ES chips with unlocked multipliers |
|
|
| crazybubba64 wrote: |
I need to confirm this, but I think the E5-1620V2 is unlocked, as I believe it is a repackaged i7-4820k. The E5-1620 is definitely locked, as the i7-3820 was as well. |
That's logical and makes sense, but I have read reports that 1620 v2 is locked. I don't have v2 CPU so I can't confirm. I have had 1620 "v1" and it was locked.
_________________
1242 x86 CPUs in my collection (week 2 2026). |
|
| Back to top |
|
|
frag_
Joined: 17 Nov 2008
Posts: 4015
Location: Estonia
|
|
| Back to top |
|
|
crazybubba64
Joined: 03 Jul 2018
Posts: 1371
Location: WI, USA
|
| Posted: Wed Jul 03, 2019 11:22 am Post subject: |
|
|
The mystery deepens...
Maybe it depends on how the board handles the chip?
I'll have to grab some of these from work next time I get the chance and confirm if they are locked or not.
_________________
My collection |
|
| Back to top |
|
|
xsecret
Joined: 01 Feb 2004
Posts: 1846
Location: France
|
|
| Back to top |
|
|
KK_Fin
Joined: 01 Jul 2013
Posts: 835
Location: Finland
|
| Posted: Wed Jul 03, 2019 2:23 pm Post subject: |
|
|
Good question and I don't know the answer. I couldn't change the multiplier over 38x (max turbo mult.) on my E5-1620 (SR0LC). If I put K-series i7 CPU on same motherboard, multiplier can be changed.
Maybe there both locked and unlocked chips? It would explain a lot about conflicting reports.
_________________
1242 x86 CPUs in my collection (week 2 2026). |
|
| Back to top |
|
|
lazyfortress
Joined: 16 Sep 2018
Posts: 279
Location: Maryland, USA
|
|
| Back to top |
|
|
wren4777
Joined: 13 Dec 2016
Posts: 571
Location: Litija, Slovenia
|
| Posted: Wed Jul 03, 2019 5:43 pm Post subject: |
|
|
1607 v2 is locked.
KK_Fin, what board/BIOS are you using your 1620 with? It is probably preventing you from overclocking since it is a Xeon. The 1620 is just a rebadged 3820, and should have a multiplier unlocked to around 42x, iirc.
_________________
My WTB List
My collection
Have a nice day!  |
|
| Back to top |
|
|
KK_Fin
Joined: 01 Jul 2013
Posts: 835
Location: Finland
|
| Posted: Fri Jul 05, 2019 6:27 am Post subject: |
|
|
| linuxfanatic wrote: | 1607 v2 is locked.
KK_Fin, what board/BIOS are you using your 1620 with? It is probably preventing you from overclocking since it is a Xeon. The 1620 is just a rebadged 3820, and should have a multiplier unlocked to around 42x, iirc. |
Board was Asus P9X79, BIOS version probably newest version, but I'm not sure since it was ~2 years ago. But now after reading your post, I might know what's going on. Non-K Sandy i5's and i7's have that "limited unlocked" feature which allow partial overclocking (4/5 bins above max turbo). And as you said 1620 is rebadged i7 3820, it is also partially unlocked and it explains links send by frag_.
I have to some day retest my chip, but so far I'd say that:
1603 locked
1607 locked
1620 partially unlocked
1650 unlocked
1660 unlocked
_________________
1242 x86 CPUs in my collection (week 2 2026). |
|
| Back to top |
|
|
|
![[ Hidden ]](https://cdn.cpu-world.com/forum/templates/CPUWorld/images/lang_english/icon_email.gif "[ Hidden ]"){kind=icon}