christianhahn

I recommend starting with DirectX 9. DX10 changed things up which can get in the way of learning the basics. The SDK can be downloaded from the MS website. Best of luck -CH

Please excuse the sequential post, but I forgot to mention something. It is extremely important that you research the position & department you are applying for. This is the question I always lead with and sets the mood for the entire interview. You would be surprised how many people don't do their due diligence and prepare. (Especially considering that I take the time and read their resume before the interview.) Thanks

I recommend starting low. Modern PCs have far too many layers of abstraction to be able to comprehend off the bat. I recommend a C-based embedded platform. Personally, I started off with assembly on an 8051. They are cheap and C compilers are readily available. Best of luck -CH

As a system/firmware programmer for a mixed-signal IC start-up, I can definitely say that the expectations for a new-hire at a start-up are higher than that of a larger, established company. Start-ups have much slimmer margins and as such cannot afford to hire unqualified people. However, as the above poster said, working for a start-up does provide near transparency of the entire design, qualification and release process. (You will be exposed to many different facets of the industry that would otherwise be obscured in a large company.) Although, I am fairly young (mid-twenties), I have interviewed people that are 20+ years older than me. From this, I have picked up several notes. Experience is not everything. A candidate with 10+ years in the same position should immediately be scrutinized. Why has he not been promoted (or made senior)? Some candidates are good at saying what they think you want to hear. It is best to say as little as possible and let them lead (e.g. “Impress me...â€) When in doubt…take them to the whiteboard. 90% of all meeting rooms have whiteboards. (Not only are they decorative, but also highly functional.) If there was ever the right time to ask a question: now is the time, not after you hire them. Problem solving and debugging are paramount. Some interviewees just don’t have the basic analytical skills to debug complex systems. Solutions must be time-efficient and plausible. To this date, I have only had one person leave an interview prematurely (and he was courteous enough to excuse himself and save me the remaining 40 minutes.) I also recommend you immerse yourself in a few side projects. Since you enjoy computer hardware, I recommend you work your way down to at least C-programming. (C is considered a must have.) You can expand your passion for digital logic by picking up Verilog (alternatively VHDL). Don’t stress about syntax too much. If an interview hinges on syntax then you probably don’t want the job anyways. (As long as the interviewee can navigate the right resources and resolve syntax issues.) Even if you don’t get hired, don’t be afraid to ask for feedback! Best of luck in the New Year -CH

Umeshwarnath, The day I upgrade my 3930k I will definitely give it a proper overclock. I have simply never been in the habit of overclocking my main workstation. (Especially considering it is a nearly fanless build with 24/7 up-time) "would a 72 hour p95 make you happy?" >> It would definitely go a long way... At the end it comes down to how one prioritizes stability/performance/cost/life-time. Thanks for you comment! Unfortunately, I couldn't quite figure out how to make this a reply to your comment. :)

Waiting on a HSPC Top-Deck Tech Station (HPTX form factor) :)

Currently, I am a System & Firmware Engineering working out of Sunnyvale, CA. (at a fabless, RF mixed-signal IC design company) I would definitely encourage others to pursue a career in EE field, and don't limit yourself to digital (analog & system engineers are in much higher demand, more compensated and versatile). At the end of the day, do whatever makes you happy. -CH P.s. Definitely agree with Mattress regarding proper attire :) Although, I do favor a pair of black jeans...

Although a responsible overclock can extend the performance envelope in certain applications, it has several drawbacks: Reduces MTBF Voids device warranty >> Operating an Intel/AMD/etc. part beyond specified operating conditions burns an OTP fuse within device which voids warranty Reduces system stability >> Increasing clock frequencies without also driving-up voltages will increase timing & logic-level error rates. Increasing voltages beyond recommended operating conditions causes premature degradation For non-mission critical applications, overclocking remains a novelty means of increasing immediate performance with little weight on long-term up-time. Although it is easy to say that an overclock appears “stable†(and seemingly free), the simple fact of the matter is that over a large enough sample size, the above statements are apparent. (Needless to say, overclocks are not stable over PVT.) Nonetheless, overclocking is becoming increasingly popular and backed by an ever-growing online/offline community. I am surprised how much traction overclocking has gotten. (i.e. the amount of overclocking targeted computer hardware available today). -CH (And yes, I am that guy that bought a 3930K (a Xeon that failed one to many scan chains) without the intention of overclocking it)