DUAL CORE Haswell i5-4200m Review using Sager NP7330 / Clevo W230ST

Introduction
Setup
Benchmark Results
. Artificial Benchmarks
. Game Benchmarks
. CPU Benchmarks
Temperatures
and Power Consumption
Battery Life
Conclusion


intro INTRODUCTION

Intel has finally released their dual core Haswell mobile CPU's and Sager is offering them with their latest laptop offerings including the W230ST. Larry at LPC Digital was kind enough to send me a sample CPU to evaluate in my NP7330, so I put it to the test.

Prema also has released his latest modded BIOS which allows for users to adjust number of cores being utilized, so I also tested the i7-4800MQ in my laptop limiting it to two cores.

setup SETUP

First let's look at the significant differences between the two CPU's being evaluated:

CPU	i7-4800MQ	i5-4200m
TDP	47W	37W
Physical Cores	4	2
Hyperthreading	Yes	Yes
Rated MHz	2700	2500
Max Boost 1 Core	3700	3100
Max Boost 2 Core	3600	3000
Max Boost 3 Core	3500	N/A
Max Boost 4 Core	3500	N/A
Max XTU Overclock	+400 MHz	Not Adjustable*
iGPU	HD 4600 1.3GHz	HD 4600 1.15GHz

* By not adjustable, the i5-4200m cannot adjust the clock frequency at all, even downclock. But voltage is adjustable just like with the quad core variants.





The system used as noted is the Sager NP7330 (Clevo W230ST) with the following configuration:

System	Sager NP7300 (Cleveo W230ST)
CPU	i7-4800MQ i5-4200m
GPU	nVidia GeForce GTX 765m
GPU Driver	nVidia 331.40 Beta
RAM	2x8GB DDR3 1866 Corsair
Drive 1	128GB Plextor mSATA III SSD
Drive 2	960GB Crucial M500 SATA III SSD
Wi-Fi	Intel 7260 802.11AC
Mouse	Logitech Anywhere MX
BIOS	Prema's Modded

Unfortunately the nVidia R331 or 331.58 WHQL drivers came out just after I had finished testing this with the 331.40 beta drivers, and noticed a significant performance improvement with these newer WHQL drivers (i.e. 3DMark 11 from 4300 to 4800 score). But the relative performance difference should be the same.

The i7-4800MQ was modified with Intel XTU by dropping stock voltage -80mV just to keep temps under control. This was not done with the dual core CPU just to evaluate it from a stock perspective.

Three CPU/System configurations were tested:

(1) i7-4800MQ stock speeds with -80mv adjustement
(2) i7-4800MQ with 2 cores disabled at stock speeds with -80mv adjustment
(3) i5-4200M stock speeds and voltage

Each of the configurations were evaluated for general benchmark/performance, CPU and GPU heat, power consumption, CPU utilization (between both dual core configs), and light browsing battery life.


bench RESULTS

artbench ARTIFICIAL BENCHMARKS

3DMark 11






3DMark 13








Allbenchmark Catzilla 1.0






Unigine Heaven






Unigine Valley






gamebench GAME BENCHMARKS

Battlefield 3






Bioshock Infinite






Civilization 5
For Civilization 5, I used the built in FireTuner SDK that allows for auto turns for the human player.
I started a 10 civ large map and ran it for 200 turns to let the unit count increase.
Then I measured the amount of time it took to run for 30 turns.






Crysis 3






Dirt 3






Grid 2






Metro Last Light






Total War Shogun 2






Sleeping Dogs






Tomb Raider






cpubench CPU BENCHMARKS

7-Zip Compress Open Source Game FlightGear



Cinebench 11.5





wPrime





x264





heatpower HEAT and POWER CONSUMPTION

CPU Temperatures








GPU Temperatures






power POWER CONSUMPTION







battery BATTERY LIFE

The three configurations were tested for battery life. The system is using Prema's Modded BIOS ( biosmods.wordpress.com and was configured for the battery test using the following:

- Power Saver mode
- LCD at 20% brighness
- Backlit keyboard OFF
- Drives set to shut off after 30 minutes
- Fore Firefox Websites refreshed with 5-15 minute intervals
- Battery drained from 100% to 5% with automatic shutdown at 5%

The resulting battery life:





conclusion CONCLUSION

I used the i5-4200m in my Sager NP7330 for the last couple of weeks and was actually impressed with the overall performance. For genaral use and gaming it seemed to perform remarkably well, even for BF3, which is known to love quad cores over dual cores. Obviously multi-threaded CPU tasks like video rendering, encoding, and compiling will perform much better using a quad core.

Power and temperatures were down slightly even from the quad core running two cores, although not significantly so. 3-5C on average. The i5-4200m also stayed pegged at 3.0GHz regardless of the load thrown at it. There is still opportunity to drop temps a bit by lowering voltage like was done with the quad core.

Battery life didn't make much of a difference with dual or quad or quad with just two cores. So unfortunately you can't expect that to improve.

Overall if you're on a budget, it can't hurt to save the $75-100 difference and put that towards an SSD. You won't see much difference in gaming and general use, just if you need more intense CPU work completed then you will sacrifice that performance.

 

Reserved for future use.

 

Great job as usual! From this review, I can certainly agree with your below statement...

"... if you're on a budget, it can't hurt to save the $75-100 difference and put that towards an SSD. You won't see much difference in gaming and general use..."


---