Author: Vincent Diepeveen
Date: 09:51:42 07/09/03
Go up one level in this thread
On July 08, 2003 at 23:58:58, Michael Neish wrote: Here are the details for the freaks: Name: Singh, S.R. (swaranNOrajSPAMsinghPLEASE@hotmail.com) 6/23/03 Direct from WWDC, Steve Jobs has just announced the G5 for Apple systems. High-lights include: - Runs up to 2.0GHz - 1GHz FSB - 12 unit core - 215 instructions in-flight - 2 double precision FPUs - 2 fully-symmetric integer units - 2 load-store units - 130nm - SOI - 58M transistors - manufactured on 300mm wafers at Fishkill Chipset has been disclosed as well: - G5 System Controller - Point-to-Point - Designed by Apple, apparantly - Fabbed by IBM - 8GBPS of BW - Independent busses, so 16GBPS of BW for dual systems (?) - Using 400MHz 128bit DDR memory == 6.4GBPS of BW => Slowdown is now the RAM - AGP 8x I/O: - PCI-X (133MHz) - 2GBPS of BW - Hypertransport Interconnects (yay!) - 1x 133MHz 64-bit slot - 2x 100MHz 64-bit slot - Serial ATA interface - Firewire 400 & 800 - USB 2.0 - Gigabit Ethernet - Bluetooth support - Optical audio inputs/outputs (S/PDIF, I would assume, with TOSlink formfactor) - Up to 8GB of RAM Other: - 4x Superdrives in all models (DVD-RW writer) - GeForce FX5200 in the low-end model - Radeon 9700 Pro in the high-end models - New cases (Aluminum instead of plastic) - 9 variable-speed fans, yet only 35dB, amazingly (I have trouble believing this) Prices: $1999 for 1x 1.6GHz, 80GB HD $2399 for 1x 1.8GHz $2999 for 2x 2.0GHz (!!>) Shipping starts in August, amazingly enough. This is even harder to believe. ------------ Name: Tvar' (unununium@hotmail.com) 6/23/03 Spec comparison grabbed from feed at arstechnica.com 3 system comparison 3.0 GHz P4, 2.0 GHz dual G5, 3.06 dual Xeon Using gcc 3.3 Single processor tests: SpecINT2000 scores: 880 (1164 for Intel compiler) - P4, 836 - Xeon, 800 - G5 SpecFP2000: 693 (1213 for Intel compiler), 646, 890 Dual Xeon vs. G5 Spec rate comparison: Single P4 reference, Dual Xeon, Dual G5 SPECint rate: 10.3, 16.7, 17.2 SPECfp rate: 8.1, 11.1, 15.7 Amazing what the Intel compiler achieves in comparison to GCC 3.3 on Spec. Most compiler benchmarking I've seen has shown later versions of GCC to be pretty close to Intel's compiler for most real world applications, I wonder how much of Intel's advantage when using ICC is applicable outside of SPEC, for which they have tuned the compiler. > >Hello, > >With the release of Apple's G5 series last week, I came here -- after some time >-- hoping to read people's comments and opinions about it. To my surprise, the >topic doesn't seem to have been discussed. I thought that Apple's claims to >have released the first 64-bit personal computer would have at least caused some >minor ripples. > >I'm not at all knowledgeable on computer hardware and architecture. I would >have thought that the G5 would be a very good processor for bitboard-based >programs. Is this really so, or is it -- like Apple's speed benchmark results >posted on their Web site -- illusory? Also, regarding Chess programming, does >the G5 have any advantages over the G4, apart from raw speed? > >Thanks to anyone who answers. > >Cheers, > >Mike.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.