- #31
Smurf
- 441
- 3
Well, In steve's defence I'm hardly representative of this forum. I'm not even a science major.
Comparing AMD and Intel Processors - Steve's Experience
- Thread starter Stevedye56
- Start date
-
- Tags
- Experience
Click For Summary
Discussion Overview
The discussion revolves around preferences between AMD and Intel processors, exploring personal experiences, performance comparisons, and broader architectural considerations. Participants share their opinions on the current state of both brands, as well as alternative architectures like SPARC and PowerPC.
Discussion Character
- Debate/contested
- Exploratory
- Technical explanation
Main Points Raised
- Some participants express a preference for AMD, citing personal satisfaction with AMD processors, while others argue that Intel's Core 2 models currently outperform AMD offerings.
- One participant mentions a bias due to using a Mac, questioning the relevance of AMD in that context.
- Another participant highlights the GHz performance of the Pentium D, suggesting that clock speed is a more important factor than brand loyalty.
- Concerns are raised about the proprietary nature of x86 architecture, with some participants advocating for the openness of SPARC and PowerPC architectures.
- There is a discussion about the implications of Intel's legal actions against AMD and the challenges for other manufacturers in the x86 space.
- Some participants express confusion about the relevance of SPARC and PowerPC, asking for more information about these architectures.
- Several participants engage in a back-and-forth regarding the definitions and implications of proprietary versus open architectures.
Areas of Agreement / Disagreement
Participants do not reach a consensus on which processor brand is superior, with multiple competing views on the performance and relevance of AMD and Intel. The discussion also remains unresolved regarding the merits of alternative architectures like SPARC and PowerPC.
Contextual Notes
Participants express varying levels of familiarity with different processor architectures, and there are unresolved questions about the implications of proprietary technology in computing.
- #32
Stevedye56
- 402
- 0
Ok cool. No hard feelings although you must admit it was getting heated. And I do agree with you that it was a good opportunity and that many people were unaware.
/rant
lol.
/rant
lol.
- #33
Smurf
- 441
- 3
So.. where would I get a RISC chipset, or a RISC-compatible computer?
- #34
Stevedye56
- 402
- 0
Smurf said:
I found a few after doing some Google Product searches and Ebay searches.
- #35
Stevedye56
- 402
- 0
Or in a Nintendo 64 
- #36
graphic7
Gold Member
- 450
- 2
Genesi has some very affordable (http://www.genesippc.com/) PowerPC systems. Most new SPARC systems aren't exactly affordable (the new Sun Ultra 25 is the low-end SPARC workstation sold by Sun and it starts at ~ $2,500); however, many that are in the 7- to -10-year-old range can be had for pennies:
Ultra 2 ~ $50 - $200
Ultra 60 ~ $100-$300
Ultra 80 ~ $200-$400
Blade 1000 ~ $300 - $1,500
These prices vary per configuration. For example, my Blade 1000 /w 2x750MHz UltraSPARC-III processors (8MB of L2 cache/proc), 2GB of memory, and a single 36.6GB internal fibre-channel disk would run around $400 now, but a config with 2x1.2GHz UltraSPARC-III CU procs would run close to $1,500.
The fact that SPARC systems are generally more expensive doesn't mean that SPARC is better than PowerPC (in some regard, it *may* be, but this is irrelevant). All this implies is that the current set of SPARC manufacturers (Fujitsu and Sun) aren't producing economical SPARCv9-compliant systems, whereas, vendors like Genesi are producing such PowerPC-based systems. IBM, on the hand, while a PowerPC vendor uses PowerPC-based processors in their high-end graphics workstations and servers, but these are far from economical (similar to SPARC in terms of cost).
The market for SPARC workstations has gotten much better. One may shudder at the starting price of $2,500 for the Ultra 25; however, my Blade 1000 in the year 2000 would've run around $30,000.
Ultra 2 ~ $50 - $200
Ultra 60 ~ $100-$300
Ultra 80 ~ $200-$400
Blade 1000 ~ $300 - $1,500
These prices vary per configuration. For example, my Blade 1000 /w 2x750MHz UltraSPARC-III processors (8MB of L2 cache/proc), 2GB of memory, and a single 36.6GB internal fibre-channel disk would run around $400 now, but a config with 2x1.2GHz UltraSPARC-III CU procs would run close to $1,500.
The fact that SPARC systems are generally more expensive doesn't mean that SPARC is better than PowerPC (in some regard, it *may* be, but this is irrelevant). All this implies is that the current set of SPARC manufacturers (Fujitsu and Sun) aren't producing economical SPARCv9-compliant systems, whereas, vendors like Genesi are producing such PowerPC-based systems. IBM, on the hand, while a PowerPC vendor uses PowerPC-based processors in their high-end graphics workstations and servers, but these are far from economical (similar to SPARC in terms of cost).
The market for SPARC workstations has gotten much better. One may shudder at the starting price of $2,500 for the Ultra 25; however, my Blade 1000 in the year 2000 would've run around $30,000.
Last edited:
- #37
Anttech
- 233
- 0
And what's even more interesting (paradox) is what is inside the Xbox 360 (built by M$) and it aint x86
http://en.wikipedia.org/wiki/Xbox_360#Central_processing_unit
http://en.wikipedia.org/wiki/Xbox_360#Central_processing_unit
- #38
graphic7
Gold Member
- 450
- 2
Anttech said:
I'm sure Intel wasn't happy about this, but then again, not many vendors have multi-cored CPUs that are built for floating-point calculations. At the time Microsoft made their decision to go with the PowerPC, there was the UltraSPARC-T1, which has a shared FPU unit across all of its 8 cores (this implies floating-point performance is not stellar; however, the UltraSPARC-T1 is excellent for integer workloads), and the IBM POWER4/POWER5, which certainly aren't consumer chips.
Its far more economical for these multi-cored CPU vendors to share the FPU across all the cores and/or share the external cache across all the cores, and many of them do. The UltraSPARC-T2 (the upcoming replacement for the -T1) will give each core its own FPU.
- #39
Smurf
- 441
- 3
how do you pronounce "x86"? Do you say the 'x' like
"ex-eighty six" or what?
"ex-eighty six" or what?