Navigate:

Advice
Beginners
BIOS Guide
CPUs
Links
Mainboards
Memory
Network
Storage
Video/Sound Cards

Contact
Forum
SiteMap
Sponsors
WebNews
Home
. .

Prices:

Mainboards

ABIT
ASUS
Chaintech
Shuttle
Soyo
Tyan

CPU
Intel
P4 2.4C-800
P4 2.6C-800
P4 2.8C-800
P4 3.0-800
P4 3.2-800

AMD
AthlonXP
XP 1700+
XP 2000+
XP 2400+
XP 2500+
XP 2700+
XP 3000+
XP 3200+

Athlon64
Athlon64 3200+
Athlon64 FX-51

Opteron
Opteron 240
Opteron 242
Opteron 244
Opteron 246

Memory

Corsair
Crucial
Kingston
Mushkin
OCZ

Search Prices:








































































































































What are you
shopping for?



































































































































































LOSTCIRCUITS

SHORTCUTS:
The Bug that Wasn't
-50 Series / B3 Silicon
Test Configurations
Memory Subsystem
Power Consumption
TrueSpace and Power Efficiency
Cinebench
DVD-Shrink, MainConcept
VirtualDub/DivX
3DMark'06
World In Conflict
Crysis
F.E.A.R.
UnrealTournament3
Overclocking - The Final Analysis

Give Us Some Feedback on this Review

 AMD's Phenom X4 9850 - Silicon Revision B3
(Author: Michael Schuette, April 6, 2008)

B3 Silicon Memory Subsystem

In the beginning of this article we hypothesized about the effect of a hierarchy of patches and work-arounds and mentioned our suspicion that enacting an increased workload on the BIOS level and then negating it on the level of the OS cannot be the same as just having a pure, unadulterated performance configuration.

Since we have to accept at face value that there weren't really any other changes in the B3 silicon revision, we assume that the differences between a Phenom X4 9850 and a Phenom 9900 in memory performance can solely be attributed to the patch (or lack thereof). For starters, we look at the latest version of Sisoft Sandra to see what happens to the memory bandwidth.

SiSoft Sandra Memory Bandwidth Benchmark

It is pretty obvious that there is a significant performance increase of the 9850 compared to the 9900 or the 9600. The numbers we see in the case of the 9850 are further roughly equivalent to those we got with a pre-patch BIOS on the ASUS M3A32 MVP-Deluxe.

Cachemem Memory Latency

Bandwidth is one thing, latencies are another important performance measure of the memory subsystem. In order to make sure that we are not comparing apples and oranges, we ran cachemem with the 9850 overclocked to a 13x multiplier, that is 2.6 GHz to have identical runtime conditions for both 9900 and 9850.

Memory subsystem access latenciesw, lower is better. The 9850 (solid blocks) shaves off about 5-10% of memory access latencies compared to the 9900 (transparent blocks) in "red" overdrive mode.

next page: => Power Consumption =>

All advice and educational articles on LostCircuits are free, but if you feel you can, please make a small donation to us!
Thank you!

General disclaimer: This page only reflects the author's personal opinion and assumes no responsibility whatsoever regarding any of the contents or any damages that may occur explicitly or implicitly from reading the contents of this site. All names and trademarks mentioned in this review are the exclusive property of the respective parent companies.
All contents of this site are protected by international copyright laws. Reproduction of the contents even in parts is not allowed except after written permission by the author and referral to this site.
Copyright 2002 - 2008 LostCircuits