Home

News

Forums

Hardware

CPUs

Mainboards

Video

Guides

CPU Prices

Memory Prices

Shop



Sharky Extreme : February 9, 2012





Regular Sections

- Weekly CPU Prices
- Weekly Memory Prices
- PC Buyer's Guides
- Private Eye
- Forums Spotlight
- The Rear View
- The Silicon Money Pit
- SharkyForums
- Site Info
- Links
- About Us

However, in Q3Test Phase 4 v1.07 the impact of SSE is virtually nil, and the performance gain registered amounts to one factor between the Celeron and P3 CPUs: MHz speed. id Software representatives have stated recently that Q3Test is not indicative of the final Quake3: Arena's performance level as several optimizations along with transformation and lighting have yet to be completed in its code.

It will be interesting to see how SSE support matures in Quake3: Arena between today's Q3Test software to the final retail product later this year.

Another interesting result of the testing we performed for the P3-600 lies in the L2 cache dependent scores gained from using Ziff-Davis' CPUMark99.

CPUMark99 is a test that places a large amount of strain on a system's CPU, including its L1 and L2 cache. Since the Celeron CPUs have 128KB of L2 cache running at the core's full speed versus the ½ core speed 512KB of L2 cache that all P3 CPUs are equipped with the test seems to adversely penalize the limited amount of cache on the Celerons versus the P3 CPUs.

In real world applications the amount that a CPU's L2 cache is hammered doesn't seem to match the degree of stress placed on it by the synthetic benchmarks, which is why the Celeron CPUs tend to more closely match the P3's results when set to the same MHz level in real world tests.




Copyright © 2001 INT Media Group, Incorporated. All Rights Reserved. Legal Notices | Licensing , Reprints , & Permissions | Privacy Policy