Secure hardware

Does the choice in processor, or harddisk, or board make some difference in the security of a computer-system? It doesn't have to be much difference but would e.g. an Intel-system be slightly more secure than an AMD-system? Considering that modern hardware tends to come with some possible bugs and/or possible exploits, I could imagine that some hardware components would be less vulnerable than other, similar hardware components. So I wonder, which hardware brands are best-known for the least possible vulnerabilities and bugs? (And not just processors, btw!)

Well.. that depends on your definition of "secure". If your talking about hacking/exploiting a processor, mobo, video card, ram, PSU, etc... lol, then no it doesn't matter. If your talking about heat, that's different. What processor runs cooler, etc... But there has already been tons of discussions on that topic...

Back in the day BIOS virii were pretty common, but nowadays you got BIOS write protection, which means nothing can write to the BIOS unless your turn it off...

Not too sure about right now. You might want to do some googling (and a wiki :)) for Microsoft Palladium which was the first name of MS proposed secured system for both hardware and software.

http://www.epic.org/privacy/consumer...palladium.html
http://www.theregister.co.uk/2002/06...ts_it_vendors/

The info in the links are out of date butthey still give some indication about the way MS and partners were thinking. MS has backtracked on a lot since then.

I think it's been renamed as 'Trusted Computing' which the public spin being that only authorised code will run on the machine and spyware/viruses ect will be blocked within the hardware. Those with a more paranoid nature belive the systemn will be used to block the machine to only play DRM software/ hardware and media files.

Some of these ideas are becoming apparent in MS Vista
--or not. You've got me looking now.

http://www.theregister.co.uk/2005/09...g_controversy/

https://www.trustedcomputinggroup.org/home

There are some things you might wish to consider based on physical security.

A CD ROM/Floppy/USB port offers a route to get malware onto the system.

A case that can be opened can have other hardware added (eg ADSL modem to allow ecternal access) or the hard drive to be removed and cloned, allowing the system to analysed/craked at leisure

A computer that can be physically carried out of the room allows the system to be opened and hence the above applies

Etc, etc

Hi

Let me look at it from a slightly different angle. The question I ask
myself is, how do you evaluate the security a "system" provides?
"For this", criterions are specificied (ISO 15408). The EAL's are
described in an overview at IACS[1].

At a certain (high) level of assurance, see DoD 5200.28-STD[2] it holds that
"no computer system can be considered truly secure if the basic hardware
and software mechanisms that enforce the security policy are themselves
subject to unauthorized modification or subversion."

Unfortunately, I would like to know much more about the certification processes than I
do, hence I am careful with what I am saying. Anyway, the highest (I think)
assurance level certified operating system, STOP[3a,3b], only has achieved
this certification by the use of particular hardware!

In this context, hardware is very relevant to achieve high-assurance level
certification. I would be happy to listen to more qualified voices.


Cheers.


[1] http://www.cesg.gov.uk/site/iacs/ind...displayPage=13
[2] http://www.fas.org/irp/nsa/rainbow/std001.htm
[3a] http://niap.nist.gov/cc-scheme/st/ST_VID3012a.html
[3b] http://www.cygnacom.com/labs/pfSEL0181T3b.htm

I believe this would fall under assurance and good architecture, etc .......... how can you design good code following the criteria and then piss away all your potential assurance by not holding your hardware to the yardsticks of the industry?



(Orange book)
3.3.3 Assurance
3.3.3.1 Operational Assurance
3.3.3.1.1 System Architecture
3.3.3.1.2 System Integrity

Quote:

---

The TCB shall maintain a domain for its own execution
that protects it from external interference or tampering
(e.g., by modification of its code or data structures).
The TCB shall maintain process isolation through the
provision of distinct address spaces under its control.
The TCB shall be internally structured into well-defined
largely independent modules. It shall make effective use
of available hardware to separate those elements that are
protection-critical from those that are not. The TCB
modules shall be designed such that the principle of
least privilege is enforced. Features in hardware, such
as segmentation, shall be used to support logically
distinct storage objects with separate attributes
(namely: readable, writeable). The user interface to the
TCB shall be completely defined and all elements of the
TCB identified. The TCB shall be designed and structured
to use a complete, conceptually simple protection
mechanism with precisely defined semantics. This
mechanism shall play a central role in enforcing the
internal structuring of the TCB and the system. The
TCB shall incorporate significant use of layering,
abstraction and data hiding. Significant system
engineering shall be directed toward minimizing the
complexity of the TCB and excluding from
the TCB modules that are not protection-critical.
3.3.3.1.2 System Integrity
Hardware and/or software features shall be provided that
can be used to periodically validate the correct
operation of the on-site hardware and firmware elements
of the TCB

---

NCSC-TG-021(Purple Book)

Quote:

---

entails a global perspective of the entire system

---

TC-2 REFERENCE MONITOR PERSPECTIVE


Attached below is a simple yet comprehensive pdf

Once again, I have forgotten my glasses at school lol... but what I made out it's sounding like, I'm wrong and you guys are saying that there is a difference security wise between AMD and Intel for example... How is this?????

More features = More problems

Recently we had the fiasco over Intel's Hyperthreading and cache timings. This feature had the side effect of allowing you to explore what is in the cache as it is being used. That could make it somewhat less secure. But it isn't a major issue.


Another thing you may consider secure/insecure is the ability to perform un-intended hardware modifications to your processors.

Such as overclocking.

Intel tries their best to prevent overclocking, and recently one of their chipsets and processor combinations somehow enforced a tight FSB speed. But it was fairly quickly reverse-engineered by other chipset makers with their own compatible chipsets and such security measures were no longer possible. There are also some software modifications to many of their smaller processors that allow control over their CPU speeds.

AMD generally allowed many modifications but are now trying to prevent them...it used to be possible to cut your own speed modifications into the PCB, but now that is pretty difficult... There are some software modifications that allow you to change the speed the hardware operates at.

There are other processors that you can perform modifications to to overclock...including that in a TI-83 calculator, a Palm Pilot / PDA, etc. You can even overclock a clock, but that's not useful since it wouldn't tell the time anymore... You could also make a 12 hour clock with hour and minute hands out of a 5400RPM hard drive, but that's another story...


You pretty much can reprogram any hardware. Sometimes it isn't financially possible because you may need a fab or electron microscope to reverse engineer it, but it is theoretically do-able.

It probably doesn't help show how "insecure" it is as much as the more common things people here have mentioned... In some systems there are multiple redundant processors that check for errors on critical stuff, but that's very specialized and exorbantly expensive stuff...something you wouldn't see in your standard x86 processor workstation.

Quote:

---

Those with a more paranoid nature belive the systemn will be used to block the machine to only play DRM software/ hardware and media files.

---

And I am one of them. Today, hardware is irrelevant to computer security.
A processor will execute the instructions fed to it. A hard drive will save the file
you send to it. The monitor will display what it's told to display.

Some evil wannabe fascist monster<rcgreen starts foaming at mouth> idiots
in the business want to insert "security" into the firmware on devices like hard drives
and vid cards<rcgreen puts on tinfoil hat>, and it would be a baaaaaaaaaaad thing.

But the answer is no, hardware has(today) no effect on security AFAIK.
:cool:

Hardware has no effect on security? Okay, then how about network printers? And there's a lot more hardware that are almost computers all by themselves. Phones these days are also not that safe anymore as they used to be because some wireless virus that seems to be running around wildly.
But my main concern is about internal hardware though. One of the first Intel processors had a floating-point bug in them that would mess up some calculations. I can imagine that modern processors might have even more bugs. And maybe even bugs that would allow someone to break into the system or just break the system. And of course, not only the processor could be abused for this but perhaps also network cards, IDE cards, graphic-cards, etc.
It doesn't have to be about hackers getting access. It could also just be some nasty instruction to some piece of hardware that just kills the system until a reset will reincarnate it again.

I've read a book about computer-virusses that was printed in 1989. ("Computervirusses" By Data-Becker) It mentioned a virus that could actually damage a diskdrive by moving the drive head to a non-existing track. The hardware didn't check for illegal values thus the head started to bump against the case and thus could get damaged. These are the vulnerabilities that I am thinking about.