so umm what code did this BSOD give?

---

In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope

A cache_manager BSOD is typically cause by the system running out of usable memory. How much memory do you have installed in your machine and what O/S are you running?

---

Traveling through space at ~67,000mph!

I am running 4 gigs on a modern i7 puter. Win7. Task manager shows memory usage is only about 25%.

There is a very long stop: error message, very cryptic, means nothing to me. Dell support said it is a software problem........why I am asking here. It only blue screens running boinc.

STOP: OXOOOOOO34(OXOOOOOOOOOOO50830,OXFFFFF8800334D687,OXFFFFF8800334CEEO,OXFFFFF800028E1C8D

---

May the blessings of the Lord overtake you and run you down

Overclocking?
PSU?

---

STOP error x34 is often caused by:

* Faulty memory
* Insufficient memory
* Outdated or BETA graphics drivers
* Corrupted Hard Disk NTFS data
* Failing Hard Disk

I would suggest running MemTest86 to make sure your RAM isn't faulty. Then try upgrading your graphics drivers. Then run a CHKDSK on your hard drive(s), including the full surface scan option to ensure that it is not failing.

This is a nearly new dell computer, their specs, no overclocking(I don't know how to o/c a dell)The psu seems to have stable voltages using HWmonitor pro. It is still under warranty so I will push it until something obvious fails.
I ran prime 95 for several hours and it failed a precision number test by a marginal amount. The second longer test was good, no failures.

---

May the blessings of the Lord overtake you and run you down

Sounds like you need to run it through a memtest check overnight. See if it fails, may be a case for bad ram. Seti pushes a computer pretty hard.

---

Traveling through space at ~67,000mph!

Don't forget the obligatory virus check with an up-to-date anti-virus checker.
I assume you run CPDN on that computer as well? If so, may I ask how many CPDN models at the same time it runs?

Check http://msdn.microsoft.com/en-us/library/ff557491%28v=vs.85%29.aspx for the more official low-down of this error.

---

One failure on Prime95 is alarming enough. I still highly suspect a hardware issue, even if brand new.

Seeing you run 2 i7's, a 920, with a GTS240 and a 960, with ATI 5800 series and faults are from the i7 with the GTS240 .

Maybe the video driver, card itself, temps of the GTS240.
Most problems I had myself, came from a 9800GTX+(getting too hot) and a GTX250, also getting too hot.
Since I switched them with a GTX470 & 480, although running in PCI-Ex2 and 1x (!) and improved cooling, no faults.
Now running 4 Einstein (CPU) and 4 SETI CUDA, 2 per card, or 1 GPUgrid, per card.
Maybe try an older video driver and do check your temps.

---

This is a fresh install of win7 from a known good copy. I ran memtest overnite and the puter rebooted at some point. I ran prime95 again, found an error. Tech support updated video drivers. There is a bug in the system........just trying to track and squash. Case is open..breathing good...GTS240 runs around 62° while crunching. Puter is running chkdsk /r If this passes I will swap mem from other dell. I have to be discreet in what I say to tech support.......they want to push it off as a software problem....will wait and see what happens.
Thanks to all for your support. The problem is not resolved but you all have steered me in some positive directions.......given me a new perspective. Gotta go crunch some wu's to find ET......whatever she/he/it looks like.

---

May the blessings of the Lord overtake you and run you down

I ran memtest overnite and the puter rebooted at some point.

It rebooted by itself? Not a very good sign I'd say as memtest runs quite barebone (no OS underneath).

---

Metod ...

I ran memtest overnite and the puter rebooted at some point.

It rebooted by itself? Not a very good sign I'd say as memtest runs quite barebone (no OS underneath).

Well, an OS can be anything. By defintion, an Operating System simply allows software to interact with hardware by providing API hooks and other services/features. My USB copy of Memtest86 actually runs Linux to boot and load the testing software.

But I agree that Memtest86 is minimalist and a crash with Memtest usually indicates a hardware failure of some sort. It definately removes the possibility of Windows and Windows device drivers being the issue, so I can't agree that it is a software issue. I think the Dell support team is wrong here.

I ran memtest overnite and the puter rebooted at some point.

It rebooted by itself? Not a very good sign I'd say as memtest runs quite barebone (no OS underneath).

Well, an OS can be anything. By defintion, an Operating System simply allows software to interact with hardware by providing API hooks and other services/features. My USB copy of Memtest86 actually runs Linux to boot and load the testing software.

But I agree that Memtest86 is minimalist and a crash with Memtest usually indicates a hardware failure of some sort. It definately removes the possibility of Windows and Windows device drivers being the issue, so I can't agree that it is a software issue. I think the Dell support team is wrong here.

First off, Dell computers are junk imho. Secondly if it crashes during memtest86 it is a hardware issue. Either heat, timing or something hardware related. Being a brand new computer doesn't mean a thing. I have seen many components fail within 90 days. There are reasons you get a warranty's from companies.

Dell is also not known for telling people when they are selling a computer with known hardware issues. See the lawsuit against them for just the same issue.

---

I ran memtest overnite and the puter rebooted at some point.

It rebooted by itself? Not a very good sign I'd say as memtest runs quite barebone (no OS underneath).

Well, an OS can be anything. By defintion, an Operating System simply allows software to interact with hardware by providing API hooks and other services/features. My USB copy of Memtest86 actually runs Linux to boot and load the testing software.

Not really, Memtest86 doesn't use any OS underneath. It has some legacy ties with linux (mainly startup procedures and some HW detection) - see brief history. Your USB copy might use one or other boot loaders that come with linux, those are not limited to loading linux kernel, they could boot many different OSes.

---

Metod ...

I ran memtest overnite and the puter rebooted at some point.

It rebooted by itself? Not a very good sign I'd say as memtest runs quite barebone (no OS underneath).

Well, an OS can be anything. By defintion, an Operating System simply allows software to interact with hardware by providing API hooks and other services/features. My USB copy of Memtest86 actually runs Linux to boot and load the testing software.

Not really, Memtest86 doesn't use any OS underneath. It has some legacy ties with linux (mainly startup procedures and some HW detection) - see brief history. Your USB copy might use one or other boot loaders that come with linux, those are not limited to loading linux kernel, they could boot many different OSes.

Everything has an OS, whether it uses a traditional one or a custom one. The boot loader and even hardware detection routines would still be classified as it's own OS.

I ran memtest overnite and the puter rebooted at some point.

It rebooted by itself? Not a very good sign I'd say as memtest runs quite barebone (no OS underneath).

Well, an OS can be anything. By defintion, an Operating System simply allows software to interact with hardware by providing API hooks and other services/features. My USB copy of Memtest86 actually runs Linux to boot and load the testing software.

Not really, Memtest86 doesn't use any OS underneath. It has some legacy ties with linux (mainly startup procedures and some HW detection) - see brief history. Your USB copy might use one or other boot loaders that come with linux, those are not limited to loading linux kernel, they could boot many different OSes.

Like he said the definition of an O/S is anything that allows communication and control of hardware by software. So the bootloader from linux is linux allowing memtest to use the hardware. An operating system.

---

Traveling through space at ~67,000mph!

I ran memtest overnite and the puter rebooted at some point.

It rebooted by itself? Not a very good sign I'd say as memtest runs quite barebone (no OS underneath).

Well, an OS can be anything. By defintion, an Operating System simply allows software to interact with hardware by providing API hooks and other services/features. My USB copy of Memtest86 actually runs Linux to boot and load the testing software.

Not really, Memtest86 doesn't use any OS underneath. It has some legacy ties with linux (mainly startup procedures and some HW detection) - see brief history. Your USB copy might use one or other boot loaders that come with linux, those are not limited to loading linux kernel, they could boot many different OSes.

Like he said the definition of an O/S is anything that allows communication and control of hardware by software. So the bootloader from linux is linux allowing memtest to use the hardware. An operating system.

Well, his words about what OS is: OS simply allows SW to interact with HW by providing API hooks... Does something count as OS when all those routines needed to interact with HW are in fact part of SW itself? Not IMHO.
Think of any half-decent DOS application that carried many HW-related routines with them simply because OS did not provide any API for a lot of them.

And bootloader is not an OS, it's just a temporary (as: gets completely discarded off as soon as OS kernel starts to execute) solution to get highly complicated OS kernel (or stand-alone application) loaded in memory. In ideal world this should be done by machine firmware. Most common type of FW these days is BIOS but that one is stone age and not able to do it for any modern OS. EFI should be better at it but then OSes tend to be picky about middleware.

Everything has an OS, whether it uses a traditional one or a custom one. The boot loader and even hardware detection routines would still be classified as it's own OS.

Hmmm ... I think memory management and process switching/scheduling are most prominent entities of a modern OS. Memtest86 lacks both of them.
HW abstraction could also be done within firmware. I guess the latter is not done in PC world as too different HW could be hooked to a modern computer. Not so far in the past many decent computers didn't have that freedom and OS only supported those pieces of hardware that were also recognized by FW.

Heck, some time in the past also PC compatibles were thought of as decent machines (DOS used hardware exclusively through BIOS functions - no driver hell).

---

Metod ...

Well, his words about what OS is: OS simply allows SW to interact with HW by providing API hooks... Does something count as OS when all those routines needed to interact with HW are in fact part of SW itself? Not IMHO.

Well, in fact it does. The original IBM ROM routine comes to mind. Remember that the BIOS is nothing more than software too.

Think of any half-decent DOS application that carried many HW-related routines with them simply because OS did not provide any API for a lot of them.

Those DOS applications often used DOS to access certain things or made BIOS calls directly, and sometimes even accessed the hardware directly. In effect, that makes the software the Operating System.

And bootloader is not an OS, it's just a temporary (as: gets completely discarded off as soon as OS kernel starts to execute) solution to get highly complicated OS kernel (or stand-alone application) loaded in memory.

Agreed. Anyone who remembers the old multi-boot manager System Commander knows that the bootloader isn't the OS.

Hmmm ... I think memory management and process switching/scheduling are most prominent entities of a modern OS.

An OS, regardless if it offers the advanced services of a "modern OS" or even a basic OS needed to run basic things, is still an OS.

DOS had no memory management built in, yet it was an OS. It required a device driver to handle memory management (HIMEM.SYS & EMM386.EXE being the official ones from the OS vendor).

Memtest86 lacks both of them.

Because for it's purposes that kind of support is unnecessary. The makers of Memtest know that they need a lightweight system to simply load the software and let it do its thing.

Heck, some time in the past also PC compatibles were thought of as decent machines (DOS used hardware exclusively through BIOS functions - no driver hell).

Technically, the BIOS is nothing more than a set of extremely basic and rudimentary drivers for the hardware.

Well, his words about what OS is: OS simply allows SW to interact with HW by providing API hooks... Does something count as OS when all those routines needed to interact with HW are in fact part of SW itself? Not IMHO.
Think of any half-decent DOS application that carried many HW-related routines with them simply because OS did not provide any API for a lot of them.

And bootloader is not an OS, it's just a temporary (as: gets completely discarded off as soon as OS kernel starts to execute) solution to get highly complicated OS kernel (or stand-alone application) loaded in memory. In ideal world this should be done by machine firmware. Most common type of FW these days is BIOS but that one is stone age and not able to do it for any modern OS. EFI should be better at it but then OSes tend to be picky about middleware.

Everything has an OS, whether it uses a traditional one or a custom one. The boot loader and even hardware detection routines would still be classified as it's own OS.

Hmmm ... I think memory management and process switching/scheduling are most prominent entities of a modern OS. Memtest86 lacks both of them.
HW abstraction could also be done within firmware. I guess the latter is not done in PC world as too different HW could be hooked to a modern computer. Not so far in the past many decent computers didn't have that freedom and OS only supported those pieces of hardware that were also recognized by FW.

Heck, some time in the past also PC compatibles were thought of as decent machines (DOS used hardware exclusively through BIOS functions - no driver hell).

The short and sweet of all of this is anything that allows control of hardware. Pretty simple to understand I suppose. If you really want to look at it nothing is considered an operating system without all it's parts. The bootloader + memtest = OS. Bill Gates idea of an operating system was simply a program being able to write to disk. He later changed his differentiation to handling all input/output, controls the behavior of peripheral devices using software drivers, and provides methods, events and properties for programmable software objects (ActiveX, Dynamic Link Libraries, The Win-32 API, COM and so forth). But obviously that's a bit self serving as he is describing windows.

When it comes down to the bare minimum I think it depends on who you ask. In my opinion an operating system is a piece of software that lays between the hardware and the user and give the user control of the hardware. And when you load memtest, it's running the linux bootload and kernel. So it's running linux.

As far as DOS using the bios to control hardware functions etc. At a bare minimum technically a bios could be considered an operating system at it's truest form. Basic Input and Output System. Modern day operating systems have given us task scheduling, memory management, and instrumentation it doesn't mean all operating systems have to have that. Would kind of be like expecting every care to have every option available anywhere or it's not a car.

---

Traveling through space at ~67,000mph!