facebook rss twitter

Review: Intel 32nm Core i5 661 Westmere CPU+GPU performance numbers

by David Ross on 4 January 2010, 09:00 3.45

Tags: Intel (NASDAQ:INTC)

Quick Link: HEXUS.net/qavfu

Add to My Vault: x

System setup and notes

System Intel Core i5 661 LGA1156  Intel Core i5 750 LGA1156  Intel Core 2 LGA775  AMD Phenom II X4 AM3 
Processors
Intel Core i5 661 (3.33GHz, 512KB L2, 4MB L3, dual-core, £145 estimated
Intel Core i5 750 (2.66GHz, 1MB L2, 8MB L3, quad-core, £149
Intel Core 2 Quad Q9400 (2.66GHz, 6MB L2 cache, quad-core, £143)

Intel Core 2 Duo E8500 (3.16GHz, 6MB L2 cache, dual-core £144)
AMD Phenom II X4 965 BE (3.40GHz, 2MB L2, 6MB L3, AM3, quad-core, £139)

AMD Phenom II X4 905e (2.50GHz, 2MB L2, 6MB L3, AM3, quad-core, £126)
Motherboard
Intel DH55TC (Intel H55 + ICH10R)
Intel DH55TC (Intel H55 + ICH10R)
Intel DG45ID (Intel G45 + ICH10R)
Gigabyte MA785GPMT-UD2H (785G + SB710)
BIOS revision
0027 (19/11/09)
0027 (19/11/09)
0107 (24/06/09)
F4c beta (12/11/09)
Memory
4GB (2 x 2GB) Crucial DDR3-1333
4GB (2 x 2GB) Crucial DDR3-1333 
4GB (2 x 2GB) Corsair DOMINATOR CM2X2048-8500C5D
4GB (2 x 2GB) Kingston HyperX KHX1600C8D3K2/4GX
Memory timings and speed
9-9-9-24-1T @ DDR3-1,330MHz 
9-9-9-24-1T @ DDR3-1,330MHz 
5-5-5-15-2T @ DDR2-800MHz 
9-9-9-24-2T @ DDR3-1,607MHz 
Hard disk drive
Seagate 1000GB SATAII (ST31000528AS)
Optical drive
Samsung SH-D163 SATA DVD-ROM
Motherboard software
Intel Inf 9.1.1.1020 + Matrix Storage Manager 9.5.0.1037
Intel Inf 9.1.1.1020 + Matrix Storage Manager 9.5.0.1037
Intel Inf 9.1.1.1020 + Matrix Storage Manager 9.5.0.1037
Standard Vista SP1 drivers + AHCI 3.1.1540.127
Graphics
Intel HD Graphics IGP
BFG GeForce GT 220 512MB discrete
Intel X4500 HD IGP
Radeon HD 4200 IGP
Graphics driver
15.16.4.64.2008_PV (8.15.10.2008)
ForceWare 195.62
15.16.2.64.1986_PV (8.15.10.1986)
Catalyst 9.11
Operating system
Windows 7 RTM Ultimate, 64-bit
Power supply
Corsair CX400W
Monitor
Dell 2405FPW/Dell 3007WFP-HC

Software

Benchmarks ScienceMark 2.0 memory latency 
SiSoft Sandra 2010 (16.11), float buffered
HEXUS.PiFast calculation to 10m places
StaxRip X.264 + AAC encoder - two-pass (both passes recorded)
CINEBENCH R10 multi-CPU render 64-bit
POV-Ray 3.7.0 beta 34 64-bit

Company of Heroes: Opposing Forces v2.5 - DX9 and DX10 
Call of Duty 4: MW v1.7.568 - DX9
Enemy Territory: Quake Wars v1.5 - OpenGL

StaxRip X.264 + AAC encoder - two-pass and QuickTime 7.6.2 (multitasking)
StaxRip X.264 + AAC encoder - two-pass and ET: QW, v1.5 (multitasking)

Power-draw at idle and 2D load, as a platform

Notes

What we're trying to do is to compare the Core i5 661 against a range of other processors in the same pricing bracket. The Core i5 750 (Lynnfield) is run on the same Intel DH55TC motherboard but requires a discrete graphics card as it doesn't support integrated graphics. We've chosen a GeForce GT 220 - a step up from modern IGPs - that costs from £45

We've also chosen a couple of chips from Intel's Core 2 range, encompassing both dual-core and quad-core models. Again, pricing is similar to the '661, and both Core 2s are sat on top of an Intel G45 motherboard with previous-generation X4500 HD graphics. What the comparison will tell us is whether the Clarkdale CPU+GPU is a better bet than having a CPU-only chip and motherboard-mounted, integrated graphics.

AMD's chip pricing is such that we can justify including the top-of-the-range Phenom II X4 965 Black Edition (125W), as well as a 65W 905e. Both are run on a Gigabyte 785G motherboard that has the latest Radeon HD 4200 IGP.

So it's Intel's newest dual-core against one other dual-core (E8500) and four quad-core CPUs. Then there's the comparison between Intel HD Graphics, X4500 HD, and Radeon HD 4200. Gaming tests are conducted at 1,024x768 low-quality settings.

Apart from the discrete card-toting Core i5 750, platform costs should be similar because the price of the CPUs is very similar, as is the cost of the underlying motherboards. The Intel Core 2 system is the only one to use cheaper DDR2, though.

We'll examine system-wide power-draw and overclocking, too.