CPU shown as engineering sample?

Hi all,
I'm running an Athlon X3 445 Rana unlocked to a 4 core Athlon (showed an a Phemon B45 IIRC) and it has been running fine with no problems. I open system properties and suddenly find it says it's an engineering sample. So I check CPU-Z and that says the same thing, it's a 1.6ghz sample again. I haven't tried rebooting yet and my motherboard is an M3A790GXH.
Any explanation for this strangeness?
Thanks
P.s here is an image

Reply to thefrenchie

Where did you buy the cpu ?

That will help answer some unknowns.

Reply to notty22

The CPU was bought in november from ebuyer.com. It was working fine until it did that. It happened after I'd pushed it to 100% which I haven't really done until now.

Message quoted 1 times

Reply to thefrenchie

just a glitch or its a real Es twkr that didn't cut it as a full twkr

http://www.tomshardware.com/pictur [...] -twkr.html

Message edited by obsidian86 on 01-05-2011 at 11:54:14 PM

Reply to obsidian86

thefrenchie wrote :

The CPU was bought in november from ebuyer.com. It was working fine until it did that. It happened after I'd pushed it to 100% which I haven't really done until now.

Are you saying that the cpu used to report something different in software ?

Reply to notty22

I doubt it was a glitch because it was stably encoding video at 1.58v and the chip normally runs at 1.4v. Maybe I got lucky and maybe it means I could really push this chip hard with a major overclock. However, it could always be my funky motherboard, they are notoriously unusual as a whole. People have had a lot of problems with them.

Message quoted 1 times

Reply to thefrenchie

It used to report itself as a Phenom X4 B45 after I unlocked it, before that it was a Rana X3. It's odd.

Reply to thefrenchie

An engineering sample or ES can be a pretty big deal to places like AMD or Intel. They almost always have traceable serial numbers, as they are loaned out under very strict terms of use.

In the past Intel have been known to offer nice exchanges for the return of any ES. You may like to try contacting AMD to see if they can verify exactly what it is. If they confirm it IS indeed an ES, they will likely want it back, and id assume offer you any AMD chip of your choice in return.

As for Intel CPU's, they usually had 'ES' or 'Engineering sample' on the heat spreader... which kinda gives it away.

Message quoted 3 times

Reply to mrmez

Hmmmmm, I think I will email them later, however I just want to ask. Would a normal chip run stably at 1.58 volts without freezing or locking up? Because if not, then this chip must have made it's way to retail from a bin somehow.
I've just rebooted the PC now and it's gone back to the X3 445 that it is, the 4th core has relocked itself.

Reply to thefrenchie

*gasp* The mystery of the strange bi-polar CPU continues...

Reply to amdfangirl

mrmez wrote :

An engineering sample or ES can be a pretty big deal to places like AMD or Intel. They almost always have traceable serial numbers, as they are loaned out under very strict terms of use.

In the past Intel have been known to offer nice exchanges for the return of any ES. You may like to try contacting AMD to see if they can verify exactly what it is. If they confirm it IS indeed an ES, they will likely want it back, and id assume offer you any AMD chip of your choice in return.

As for Intel CPU's, they usually had 'ES' or 'Engineering sample' on the heat spreader... which kinda gives it away.

^+1

------------------------------ A+, Net+, MCDST, DSCE (Dell)
Reply to malmental

mrmez wrote :

An engineering sample or ES can be a pretty big deal to places like AMD or Intel. They almost always have traceable serial numbers, as they are loaned out under very strict terms of use.

In the past Intel have been known to offer nice exchanges for the return of any ES. You may like to try contacting AMD to see if they can verify exactly what it is. If they confirm it IS indeed an ES, they will likely want it back, and id assume offer you any AMD chip of your choice in return.

As for Intel CPU's, they usually had 'ES' or 'Engineering sample' on the heat spreader... which kinda gives it away.

Nah, this is urban legend stuff. Intel and AMD give out lots of ES parts and they dont track them once they are shipped out. We are talking thousands for every stepping they produce. They are marked as ES so they do not get sold or returned for warranty replacement.

They are given to the press, to PC and server manufacturers, and to any company they have a relationship with. The CPUs are marked with an ES on top which Toms will usually show in pictures - and reported with CPUz. ES parts may be for a stepping that is never released to the public because it has flaws, so its part of why you dont want to pay for them - but if the price is right - go for it.

In this case the modification seems to be confusing CPUz

Reply to vvhocare5

Hmmmm that's strange because I bought the part from a legit site and it looked retail when I built the PC. I'm emailing AMD now to see what they have to say about it.
Also, I just noticed that it says it's a TWKR. Is it worth seeing if I can sell it? I'm sure people would pay.

Reply to thefrenchie

thefrenchie wrote :

I doubt it was a glitch because it was stably encoding video at 1.58v and the chip normally runs at 1.4v. Maybe I got lucky and maybe it means I could really push this chip hard with a major overclock. However, it could always be my funky motherboard, they are notoriously unusual as a whole. People have had a lot of problems with them.

You don't want it to run over 1.5V. Have you set the voltage in the bios?

Reply to simon12

so it's running stable unlocked into a phenom ii quad and how far can you clock it.?
have you tried.?

------------------------------ A+, Net+, MCDST, DSCE (Dell)
Reply to malmental

It automatically set itself to 1.58v, now I rebooted again and it has gone back to a Phenom X4 B45 at 1.4v. It'll probably go back to an Engineering Sample after a reboot. It was an Engineering Sample the last 3 times I booted it. However, the B45 has gone from the name.
I've never tried overclocking it past 3.4ghz but it ran stably then at 1.425v.
I'm waiting on a reply from AMD too, if they want it back, I'll see if I can get it replaced with a real Phenom X4 or even an X6.

Reply to thefrenchie

for the sake of amusement, run the WEI with it locked and then unlocked, what's the differences if any REPORTED..

------------------------------ A+, Net+, MCDST, DSCE (Dell)
Reply to malmental

Well it clocks down to 1.6ghz with a max multiplier of x8 instead of x15.5 so it becomes noticeably slower than my ULV notebook.
I'd try doing it but I can never be sure what it'll reboot as. I'm leaving it running for now because this time it booted as a real Phenom.

Reply to thefrenchie

understood..

------------------------------ A+, Net+, MCDST, DSCE (Dell)
Reply to malmental

UPDATE: I had to check the BIOS anyway and I saw a mindblowing multiplier setting for my CPU. 6300mhz is the limit apparently. Sorry for the poor quality but I only had my phone's camera.

This really seems to be an ex TWKR CPU.
Any thoughts?

Reply to thefrenchie

interesting..

------------------------------ A+, Net+, MCDST, DSCE (Dell)
Reply to malmental

How far should I try and push it on the stock heatsink? It keeps around 35 under load on stock settings.

Reply to thefrenchie

Well, interestingly enough if it is a ES I wonder how AMD does it. I know one guy contacted Intel (this was back in 2007) about a QX6700 ES he got and Intel sent him a unreleased Core 2 Quad QX6850 for reporting it. Not sure if AMD would do the same since ES chips are meant to be with testers, not consumers.

As for the maximum frequency, thats only with the multiplier at 200MHz. If you can increase the BCLK, then it can go higher. so say 250MHz BCLK at 31x yould be 7750MHz. But you wont be able to hit that unless its cooled with LN2. I would say on air you can probably hit about 4-4.5GHz on air. Then if you do water cooling, you can probably hit 5GHz+.

Reply to jimmysmitty

Well I emailed AMD and they didn't want to know so I think it may well be worth giving them a ring and speaking to a real human being. It may be an ES that they assumed would be OK going out as an X3.
I tried overclocking to around 3.8ghz and it worked fine at just over stock voltage for the X3 which is 1.4v, I haven't tried it at 1.55v which the CPU states is the max when it's an ES.
Unfortunately, the CPU has a mind of it's own. It's hit or miss what it will boot as. I will try fiddling with ACC to see if there is a specific combination which works more reliably to have it boot as the ES.

Message quoted 1 times

Reply to thefrenchie

I'd say give them a ring. They might give you an actual TWKR 42.

------------------------------ Dying
Is an art, like everything else.
I do it exceptionally well.
-Slyvia Plath Lady Lazarus
Reply to amdfangirl

"Stock' voltage is around 1.325v --- 1.4v is the specified 'Max' voltage.

1.58v is too much (unless you are under DIce or LN2).

When you 'unlock' an AMD processor you bork the system readings, sensors and ID. AMD sold the CPU as an X3 445 and you have claimed:

Quote :

... I tried overclocking to around 3.8ghz and it worked fine at just over stock voltage for the X3

So what's the issue here, again?

Reply to Wisecracker

thefrenchie wrote :

Well I emailed AMD and they didn't want to know so I think it may well be worth giving them a ring and speaking to a real human being. It may be an ES that they assumed would be OK going out as an X3.
I tried overclocking to around 3.8ghz and it worked fine at just over stock voltage for the X3 which is 1.4v, I haven't tried it at 1.55v which the CPU states is the max when it's an ES.
Unfortunately, the CPU has a mind of it's own. It's hit or miss what it will boot as. I will try fiddling with ACC to see if there is a specific combination which works more reliably to have it boot as the ES.

don't bother contacting AMD it's obvious they do not care, just push that chip as hard as it will go..

------------------------------ A+, Net+, MCDST, DSCE (Dell)
Reply to malmental

I've got a Phenom X4 9850 that I bought just after the launch of the Phenom II 920/940, got it retail from newegg. It's paired with a Biostar 790GX based board(not sure exact model number and I'm not on the computer to check). It occasionally reads as an engineering sample, but when it does a simple reboot makes it show Phenom X4 9850 again. I attribute it to some sort of software glitch within my BIOS or something else unknown.

Reply to loneninja

Lots of engineering samples "escape" over the course of the development and sampling to our partners.

The fact that you bought it from a vendor would be of interest to AMD as those parts are supposed to be legally sold.

Someone having an engineering sample is not a real concern (it happens often) but somebody buying one is a big concern.

------------------------------ While I work for AMD, my posts are my own opinions.

http://blogs.amd.com/work/author/jfruehe/
Reply to jf-amd

^I would imagine that's the one case. Someone given one wouldn't seem to be a problem. But them selling it is where it gets flaky.

I would love to know how AMD handles it

Reply to jimmysmitty

I have had to deal with this before. Generally I'm willing to give someone the real thing, but only if they are only going to work with our investigations department.

If you tell me you want a new processor but don't want to get involved, it's not going to go to far.

Message quoted 2 times

------------------------------ While I work for AMD, my posts are my own opinions.

http://blogs.amd.com/work/author/jfruehe/
Reply to jf-amd

Gee, imma make a bet that engineering samples must sell for heaps!

I mean like how cool is it to have an ES chip?

Message quoted 1 times

------------------------------ Dying
Is an art, like everything else.
I do it exceptionally well.
-Slyvia Plath Lady Lazarus
Reply to amdfangirl

jf-amd wrote :

I have had to deal with this before. Generally I'm willing to give someone the real thing, but only if they are only going to work with our investigations department.

If you tell me you want a new processor but don't want to get involved, it's not going to go to far.

Working together?

Pffft

Your never going to win over the intel / nivida fanbois like this.

Message quoted 1 times

Reply to Randomacts

Randomacts wrote :

Working together?

Pffft

Your never going to win over the intel / nivida fanbois like this.

Then again, you can't win over fanbois.

They'll support whatever they want.

------------------------------ Dying
Is an art, like everything else.
I do it exceptionally well.
-Slyvia Plath Lady Lazarus
Reply to amdfangirl

jf-amd wrote :

I have had to deal with this before. Generally I'm willing to give someone the real thing, but only if they are only going to work with our investigations department.

If you tell me you want a new processor but don't want to get involved, it's not going to go to far.

If I'm honest, I'd rather have something that's a bit more reliable because it keeps resetting itself. I know the multi is unlocked so it isn't just a glitch but the CPU flits between saying it's an ES then saying it isn't.
I bought the CPU from a legit site, namely ebuyer.com who are a big company in the UK.
Seeing as you work for AMD, what would be the best way to get someone to look into the problem?

Reply to thefrenchie

^
now this is going to be good..

------------------------------ A+, Net+, MCDST, DSCE (Dell)
Reply to malmental

Engineering samples are potentially older steppings that could have issues in them - which might be why the processor is resetting itself.

Shoot me a PM with your name, address, phone and email along with the OPN (product part number) and the serial number.

I will get that over to investigations and they will contact you.

I can't say for sure (because I only deal with server parts) but generally if you provide us with all of the relevant info (including the receipt) we will replace it with the same part, but a final production part that won't have that problem.

------------------------------ While I work for AMD, my posts are my own opinions.

http://blogs.amd.com/work/author/jfruehe/
Reply to jf-amd

amdfangirl wrote :

Gee, imma make a bet that engineering samples must sell for heaps!

I mean like how cool is it to have an ES chip?

well, if it has bugs in it that were fixed with spins of the silicon it would be worth little

------------------------------ While I work for AMD, my posts are my own opinions.

http://blogs.amd.com/work/author/jfruehe/
Reply to jf-amd

^
please explain that..

------------------------------ A+, Net+, MCDST, DSCE (Dell)
Reply to malmental

Well, through manufacturing, most products go through development cycles.

Take Firefox for example, 4.0 is currently in beta. One of the reasons they release the beta is so extension developers can write code for the next-gen browser, allowing for extensions on launch day. Obviously, these aren't considered to be stable for cilent/consumer use, but good enough for testing/coding purposes.

In a similar way, AMD sends samples of CPUs to companies so they can integrate them into systems, try them out or to write optimised code for their next-generation processors. However, like Mozilla's Firefox, all the kinks of the new product haven't been worked out.

You see processor revisions from time to time, much in the same way Mozilla patches Firefox (abeit with lesser frequency). Newer AMD CPU steppings in the Phenom I fixed the infamous TLB (albeit with performance deficit) as well as the C3 stepping for the Phenom II and advancements on the 45nm node allowed them to make a Phenom that originally ran with a 125W TDP with the previous C2 revision at 95W with the C3.

With the ES chip, often they use steppings from before production, so in the Phenom II's case it might have been the precursor to C2.

ES chips are valueable because AMD, unlike Mozilla tends not to give out ES/beta products out to customers. They're unique and hard to find. Kinda like a trophy.

JF-AMD correct me if I'm wrong... I probably am.

Message edited by amdfangirl on 01-15-2011 at 09:38:46 PM

------------------------------ Dying
Is an art, like everything else.
I do it exceptionally well.
-Slyvia Plath Lady Lazarus
Reply to amdfangirl

So, here's the "birds and the bees" of how silicon is born.

There is a first rev, that is referred to as "early silicon." That is internal only and typically only a few dozen parts. These never leave our labs. The expectation is that there will only be a few functional cores per die, some of the cache won't work, and the clock speed is really low. This basically lets you test the functionality of the design.

The next rev is called "proto" because these are the prototype parts. Protos should be fully functional (but sometimes might not) and these go through our labs and then get handed out to a few select (ie. big) OEMs who happen to have really good programs for crunching silicon (sometimes a second and third set of eyes helps you find things.)

The next spin of silicon is EVT for early validation testing. This is now several hundred parts. These go through a few weeks of test and then after we have a certain amount of coverage and feel good about the health, they go out to all of our major OEM partners, board partners, etc.

The next release is DVT for design verification testing. DVTs are in the thousands and are generally considered "production quality." DVT should be exactly like final silicon. If there is any change that happens, it is typically only a few gates and definitely only a single layer at that point. A multi-layer change would require another round of DVT parts.

Then, after all of the tests are completed and the silicon passes, it is declared production silicon and we are able to sell it. We cannot sell any pre-production silicon, so there is a marking on the lid.

Between each round there could be a layer change (respin).

Most products don't change between EVT and DVT; Istanbul basically nailed the design on the first round of silicon, which is why it launched much earlier than expected.

So, getting DVT silicon is like getting a final production product. But getting EVT silicon is generally bad because there could be something that caused another silicon respin.

Software partners generally get DVT parts. Compiler partners typically get EVT and DVT parts, but the procedure will vary with some products, so it is not a hard rule.

------------------------------ While I work for AMD, my posts are my own opinions.

http://blogs.amd.com/work/author/jfruehe/
Reply to jf-amd

How are steppings made?

This is awfully fascinating stuff.

------------------------------ Dying
Is an art, like everything else.
I do it exceptionally well.
-Slyvia Plath Lady Lazarus
Reply to amdfangirl

I'm learning a lot. I never really realised what went into making a CPU.

Reply to thefrenchie

There are between 8 and 12 layers on most chips. A stepping is when the circuits change. A minor stepping is generally a single layer changing. A major stepping is multiple layers changing.

------------------------------ While I work for AMD, my posts are my own opinions.

http://blogs.amd.com/work/author/jfruehe/
Reply to jf-amd

^
I love all this detailed insider info jf-amd.
share your brain, let's slice her open...LOL

------------------------------ A+, Net+, MCDST, DSCE (Dell)
Reply to malmental

Everything I said is public. Please do not imply that I am sharing inside information.

Message quoted 1 times

------------------------------ While I work for AMD, my posts are my own opinions.

http://blogs.amd.com/work/author/jfruehe/
Reply to jf-amd

jf-amd wrote :

Everything I said is public. Please do not imply that I am sharing inside information.

i understand a I didn't mean anything to imply illegal activity... :ange:
just being grateful for the info you have shared and stated in a jokingly manner, or so I thought.
my apologies..

Message quoted 1 times

------------------------------ A+, Net+, MCDST, DSCE (Dell)
Reply to malmental

malmental wrote :

i understand a I didn't mean anything to imply illegal activity... :ange:
just being grateful for the info you have shared and stated in a jokingly manner, or so I thought.
my apologies..

He's trying to tell you that you're awesome, JF.

------------------------------ Dying
Is an art, like everything else.
I do it exceptionally well.
-Slyvia Plath Lady Lazarus
Reply to amdfangirl

Dadiggle wrote :

Amd is it true that there's about 60 percent of working dies on a single wafer?

We never talk about yields.

------------------------------ While I work for AMD, my posts are my own opinions.

http://blogs.amd.com/work/author/jfruehe/
Reply to jf-amd

Is that something to do with speculators and stock investors?

------------------------------ Dying
Is an art, like everything else.
I do it exceptionally well.
-Slyvia Plath Lady Lazarus
Reply to amdfangirl

Forum CPU & Components : CPUs CPU shown as engineering sample?

CPUs

Power Supplies, Cases & Mods

Sound Cards

Servers

NICs

Other Components

Akasa

Antec

Cooler Master

Cougar

DeepCool

Enermax

Please mind