We have just released a long-awaited update to one of our flagship products, Elcomsoft Distributed Password Recovery. While you can learn more about what’s been added and changed from our official announcement, in this post we’d like to share some insight about the path we took to design this update. (more…)
Posts Tagged ‘GPU acceleration’
Elcomsoft Phone Breaker Update: Improved iCloud Acquisition, Two-Factor Authentication and Stronger Brute ForceWednesday, December 17th, 2014
We are excited to announce an update to one of our oldest mobile forensic tools, Elcomsoft Phone Breaker. In this release we mostly targeted iCloud acquisition, although we’ve made some changes to the password recovery algorithm targeting iOS offline backups. All in all, the new tool can be used under a wider range of circumstances, squeezes more juice of your existing acceleration hardware and adds support for newest and greatest AMD and NVIDIA boards.
Do you think you know everything about creating and using backups of Apple iOS devices? Probably not. Our colleague and friend Vladimir Bezmaly (MVP Consumer security, Microsoft Security Trusted Advisor) shares some thoughts, tips and tricks on iTunes and iCloud backups.
Mobile phones are everywhere. They are getting increasingly more complex and increasingly more powerful, producing, consuming and storing more information than ever. Today’s smart mobile devices are much more than just phones intended to make and receive calls. Let’s take Apple iPhone. The iPhone handles our mail, plans our appointments, connects us to other people via social networks, takes and shares pictures, and serves as a gaming console, eBook reader, barcode scanner, Web browser, flashlight, pedometer and whatnot. As a result, your typical iPhone handles tons of essential information, keeping the data somewhere in the device. But what if something happens to the iPhone? Or what if nothing happens, but you simply want a newer-and-better model? Restoring the data from a backup would be the simplest way of initializing a new device. But wait… what backup?
Users in general are reluctant to make any sort of backup. They could make a backup copy once after reading an article urging them to back up their data… but that would be it. Apple knows its users, and decided to explore the path yet unbeaten, making backups completely automatic and requiring no user intervention. Two options are available: local backups via iTunes and cloud backups via Apple iCloud.
In short, standard key-derivation function, PBKDF2, is used in a very strange way, to say the least. Where Apple has used 2’000 iterations in iOS 3.x, and 10’000 iterations in iOS 4.x, BlackBerry uses only one.
So password verification is (was) so fast/simple that we did not care about implementing it on the GPU — modern CPU is able to crack almost 8 million passwords per second (thanks to multi-threading and AES-NI). We would not call that the vulnerability, but still the weak link.
But new versions of BlackBerry Desktop Software have been released reсently (6.0 for Windows and 2.0 for Mac). And as always, there are bad news and there are good news.
Most modern CPUs are multi-core – it is not easy to find even a laptop with less than two cores these days. And for desktops, 4 cores are usual now.
Password recovery is one of most CPU-intensive tasks, and it fits best into multi-processor architecture. Every CPU (or CPU core) get its own portion of passwords to try (i.e. to check their validness), and they all work in parallel. As simple as that.
So what we’re doing in our software is running multiple threads – as many as the number of CPUs (or cores) available. And the rest is being done by the operating system, that assigns the threads to cores (well, in most cases we don’t care what particular core is going to execute a particular thread, because they are all equal; the only exception is when one or more of the cores is doing something already, I mean something CPU-intensive as well).
Some time ago we wrote about the smallest password cracking device. Not suitable for you? No problem, here is another one: not as small, but definitely more powerfull: Audi. Yes, it's a car. No, we're not kidding. Just read NVIDIA and Audi Marry Silicon Valley Technology with German Engineering press release from NVIDIA. Or if you need more information, The New MMI Generation from Audi might be also helpful. In brief: Audi A8 luxury sedan is equipped with an entertainment system that uses two GPUs from NVIDIA. We have no idea what are these chips (may be Fermi?) and is it technically possible to load our own code to them, but still funny, isn't it?
ATI is going to release Radeon HD 5000 cards (5850, 5870, 5870 X2) in October — well, hopefully. The top one (HD 5870X2: single-PCB, dual-GPU) will retail for $599.
As for NVIDIA’s new GT300, the specifications were revealed in April. In brief, it groups processing cores in sets of 32 (up from 24 in GT200) — up to 512 cores total for the high-end part. If the clocks remain the same as on GT200, that will double the overall performance. And there are other improvements as well: e.g. GT300 cores rely on MIMD-similar functions. Some fresh information about GT300 availability:
- Where is the nVidia GT300?
- nVidia plans GT300 demos for late September
- NVIDIA GeForce Drivers Include Details on GT300 GPU Series
You may ask — what about Intel? Well, new Core i5 and i7 (codename Lynnfield) now available. Nothing revolutionary new, just Intel P55 Express Chipset support: integrating both a 16-lane PCI Express 2 graphics port and two-channel memory controller on a single chip (previous chipsets required separate northbridge and southbridge), as well as several minor improvements. More information and some benchmarks at Intel Lynnfield; Core i5 750 and Core i7 870 Evaluation and New Intel Core i5, i7 Processors Product Matrix.
And still [almost] noting about Intel Larrabee, mostly just rumors:
Finally, funny article: NVIDIA to Intel: Your Days Are Numbered
Just about two weeks ago, ATI has introduced the fastest GPU yet: FirePro V8750. 800 shader engines, 115.2 GB/s memory bandwidth, 2 GB frame buffer memory (GDDR5), two DisplayPort outputs, one DVI output. Thinking about purchasing it? The cost is as high as $1,800. More details at Tom’s Hardware.
Want to compare ATI with NVIDIA? Then read ATI Stream vs. NVIDIA CUDA – GPGPU computing battle royale. Or you can use our Wireless Security Auditor (which supports cards from both manufacturers) for your own tests.
Wow, Adobe rethinks PDF security. Curious why? Because of vulnerabilities in Abobe Reader (and so zero-day exploits), of course. From the article:
According to Finnish security company F-Secure, patching 48.9% of all targeted attacks conducted this year involved a malicious PDF file attached to a legitimate-looking e-mail, a huge change from 2008, when PDFs made up just 28.6% of targeted attacks.
But security model of PDF encryption/protection is not going to change, [un]fortunately. (more…)