Surface Pro 2 Gets Teardown Treatment
Before you buy, check out what's inside the Surface Pro 2!
It's been just days since Microsoft's second generation Surface hit the market, its release timed with Windows 8.1. We went hands on with the device at Microsoft's event in September. But until we've had time to finish our review testing, we rely on the likes of iFixit.
The iFixit team has torn into the Surface Pro 2 and notes that Microsoft has switched from a Micron/Marvell SSD to SK Hynic HFS128G3AMNB 128 GB mSATA 6.0 Gbps SS. There’s also a Core i5-4200U CPU, 2 GB of Mobile DDR2 RAM (four ICs for 4 GB in total) an Atmel MXT154E Touchscreen Controller, an Atmel UC256L3U 32-bit AVR Microcontroller, a Marvell Avastar 88W8797 Integrated 2x2 WLAN/Bluetooth/FM Single-Chip SoC, an MXIC MX25L4006EZNI 4Mbit SPI Flash, a Texas Instruments TPS5162 (ACTIVE) 2-Phase DCAP+ Step-Down Controller, a Texas Instruments TPS51367 Integrated FET Converter, a Winbond 25Q128FVSQ Serial Flash, Winbond 25X05CL Serial Flash, and a Winbond 25X40CL1G 4 Mb Flash.
In total, the guys at iFixit found more than 90 screws along with a battery that is stuck to the rear case and a glued down display. This is combined with an LCD and digitiser that are fused, all of which combined scores the Surface Pro 2 a 1 out of 10 on iFixit’s repairability scale. Still, if you do want to repair one of these at home, you can check out iFixit’s guide here.
Follow Jane McEntegart @JaneMcEntegart. Follow us @tomshardware, on Facebook and on Google+.

I want to see the performance of these two????
Screws ALWAYS > glue.
I am pretty sure its LPDDR3 as well since the i5-4200U only supports DDR3.
Probably a typo.
Either way the Pro 2 looks nice. I would like to see battery life.
Sean1357, This is a full blown x86 CPU so the Apple A7 wont beat it in raw performance. Battery life it may but not in performance.
DDR2 on Haswell certainly sounds highly unlikely since Intel has not made DDR2-based mainstream CPUs in two or three generations already.
DDR2 on Haswell certainly sounds highly unlikely since Intel has not made DDR2-based mainstream CPUs in two or three generations already.
The last generation to support DDR2 was the 45nm Core 2 series as it had a motherboard based MC. Once they went to a IMC (with first gen Core I) they moved to DDR3 only.
Reason 1 - They are notoriously late at reporting anything, so an indepth review is out of the question
Reason 2 - Other sites get one delivered early and review in advance, but Microsoft doesn't give one to Tom's because it is almost as bias against MS as Cnet
Is it 1, 2 or both?
You can't compare x86 and ARM processors. Even with benchmarks, it's not possible to get an accurate comparison. The architecture is completely different and as a result, the instruction set is completely different as well. Since neither can run software written specifically for the other (excluding web-based apps), you simply can't compare them.
Of course you can compare them: what matters at the end of the day is how quickly they can get a given set of tasks done within a given cost, space and power budgets. It does not really matter what the architecture or instruction set is. If your architecture does not have libraries as tightly optimized as another, that simply becomes part of your platform's handicaps and one more item to add to the list of things to look into if you want your platform to perform better in that particular type of applications.
As long as a given platform can run applications that matter to normal people or even specific crowds, benchmarking that application across all eligible platforms is as valid a benchmark as any other for their respective crowds.
Reason 1 - They are notoriously late at reporting anything, so an indepth review is out of the question
Reason 2 - Other sites get one delivered early and review in advance, but Microsoft doesn't give one to Tom's because it is almost as bias against MS as Cnet
Is it 1, 2 or both?
Demand, do you have any notable evidence of this anti-MS bias from the reporting side? Serious question - I just don't think I've seen much - and what I have seen, I've seen mirrored in articles about the other big players in the industry. They make some snarky "biased" comments about MS, Google, Apple from time to time, but I don't see a systematic singling out of one to be targeted. I actually thought their coverage of the initial Surface events was pretty good - and most of the other stuff you can just chalk up to Tom's being slow.
Reason 1 - They are notoriously late at reporting anything, so an indepth review is out of the question
Reason 2 - Other sites get one delivered early and review in advance, but Microsoft doesn't give one to Tom's because it is almost as bias against MS as Cnet
Is it 1, 2 or both?
Demand, do you have any notable evidence of this anti-MS bias from the reporting side? Serious question - I just don't think I've seen much - and what I have seen, I've seen mirrored in articles about the other big players in the industry. They make some snarky "biased" comments about MS, Google, Apple from time to time, but I don't see a systematic singling out of one to be targeted. I actually thought their coverage of the initial Surface events was pretty good - and most of the other stuff you can just chalk up to Tom's being slow.
To me the bias is not as much in the content in the articles but rather the amount of positive vs the amount of negative articles for every company.