Click Here
home features news forums classifieds faqs links search
6074 members 
Amiga Q&A /  Free for All /  Emulation /  Gaming / (Latest Posts)
Login

Nickname

Password

Lost Password?

Don't have an account yet?
Register now!

Support Amigaworld.net
Your support is needed and is appreciated as Amigaworld.net is primarily dependent upon the support of its users.
Donate

Menu
Main sections
Home
Features
News
Forums
Classifieds
Links
Downloads
Extras
OS4 Zone
IRC Network
AmigaWorld Radio
Newsfeed
Top Members
Amiga Dealers
Information
About Us
FAQs
Advertise
Polls
Terms of Service
Search

IRC Channel
Server: irc.amigaworld.net
Ports: 1024,5555, 6665-6669
SSL port: 6697
Channel: #Amigaworld
Channel Policy and Guidelines

Who's Online
38 crawler(s) on-line.
 19 guest(s) on-line.
 0 member(s) on-line.



You are an anonymous user.
Register Now!
 invent:  42 mins ago
 clebin:  1 hr ago
 michalsc:  1 hr 11 mins ago
 NutsAboutAmiga:  1 hr 15 mins ago
 Mobileconnect:  1 hr 29 mins ago
 matthey:  1 hr 33 mins ago
 dan.hutch:  1 hr 39 mins ago
 amigang:  1 hr 42 mins ago
 rtek1193:  1 hr 44 mins ago
 Comi:  2 hrs 4 mins ago

/  Forum Index
   /  Amiga OS4 Hardware
      /  Trevor's Amiga Blog
Register To Post

Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 | 60 | 61 | 62 | 63 | 64 | 65 | 66 | 67 | 68 | 69 | 70 | 71 | 72 | 73 | 74 | 75 | 76 )
PosterThread
pavlor 
Re: Trevor's Amiga Blog
Posted on 28-Jun-2021 17:15:24
#1501 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9255
From: Unknown

@Hammer

Quote:
A4000/030 wasn't good enough for Doom.


486 and slow ISA GFX could reach a similar result. My A1200 with somewhat faster CPU (50 MHz) has the very same speed as my 486SX 25 MHz notebook with ISA VGA GFX (cca 10 FPS in Doom benchmark) - it could be somewhat lower with only 25 MHz 68030, but Doom depends as much on fast GFX as on fast CPU.

Quote:
PC Mag 11 October 1993, Page 419


Are you sure, you didnt mean 11 October 1994? (it was 12 October in 1993 and there is no such advert on that page).

 Status: Offline
Profile     Report this post  
BigD 
Re: Trevor's Amiga Blog
Posted on 28-Jun-2021 17:25:24
#1502 ]
Elite Member
Joined: 11-Aug-2005
Posts: 5969
From: UK

@pavlor

In my experience 030/50 plus AGA on ADoom was fine for Doom 2s early levels and only stuttered a bit on later ones (still completed it). Yes, Final Doom needed something with a bit more grunt but should C= have shipped a souped up A1200/Falcon Plus spec machine and if iD Software had been paid handsomely Im sure it would have sold Amigas but probably not quick enough to save C= unless AAA had been ready with chunky pixel support in 1991!

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

 Status: Offline
Profile     Report this post  
matthey 
Re: Trevor's Amiga Blog
Posted on 28-Jun-2021 20:23:15
#1503 ]
Super Member
Joined: 14-Mar-2007
Posts: 1233
From: Kansas

BigD Quote:

In my experience 030/50 plus AGA on ADoom was fine for Doom 2s early levels and only stuttered a bit on later ones (still completed it). Yes, Final Doom needed something with a bit more grunt but should C= have shipped a souped up A1200/Falcon Plus spec machine and if iD Software had been paid handsomely Im sure it would have sold Amigas but probably not quick enough to save C= unless AAA had been ready with chunky pixel support in 1991!


The 68030@50MHz only had a 256 byte ICache and 256 byte DCache so it was sensitive to memory performance. Unfortunately, CBM usually used the cheapest memory possible including chip memory which is part of the reason why ECS and AGA are slow and the reason why the Jay Miner designed Ranger chipset was rejected despite being ready in 1989 before ECS (Ranger would have had 128 colors as well as higher resolutions with better performance). Some 3rd party accelerators used high performance memory with a high clocked 68030 to give integer performance close to that of a 68040@25Mhz on a CBM accelerator with poor memory performance (and likely that of a cheap 486SX too).

I don't think AAA would have saved the Amiga even if it was available in 1991. AAA custom chips would have been expensive (many more transistors than AGA) so the chipset would have only been used for high end Amigas which were not selling in high enough volumes to lower the cost with mass production. The Ranger chipset, also using VRAM which gives twice the memory bandwidth, would have had a better chance at success if first introduced in high end Amigas as VRAM was being used in enough other computers to allow shared mass production of the VRAM by the manufacturers (shared mass production is what propelled the PC clone market as well). VRAM prices soon dropped to only 20% more expensive than DRAM which was the primary expense of the Ranger chipset and could have then been used for all Amigas. AAA benefited from the lower cost of VRAM but the custom chips would have remained too expensive for low end Amigas. AAA had poor compatibility with OCS/ECS as well. Just as AGA was determined to be a cheaper and more practical alternative to AAA, the AA+ chipset was more likely to be the next Amiga chipset with logic and specs taken from the permanently shelved AAA. SAGA is more along the lines of the AA+ chipset, minus some blitter functionality currently. It keeps good OCS/ECS/AGA compatibility but adds features like chunky gfx and 8 voice 16 bit audio planned for AAA.

Last edited by matthey on 28-Jun-2021 at 08:31 PM.

 Status: Offline
Profile     Report this post  
BigD 
Re: Trevor's Amiga Blog
Posted on 28-Jun-2021 22:00:47
#1504 ]
Elite Member
Joined: 11-Aug-2005
Posts: 5969
From: UK

@matthey

Quote:
Some 3rd party accelerators used high performance memory with a high clocked 68030 to give integer performance close to that of a 68040@25Mhz on a CBM accelerator with poor memory performance (and likely that of a cheap 486SX too).


That'll be it then as mine used EDO Simms. Nice!

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor's Amiga Blog
Posted on 29-Jun-2021 5:08:21
#1505 ]
Elite Member
Joined: 9-Mar-2003
Posts: 4270
From: Australia

@pavlor

Quote:

pavlor wrote:
@Hammer

Quote:
A4000/030 wasn't good enough for Doom.


486 and slow ISA GFX could reach a similar result. My A1200 with somewhat faster CPU (50 MHz) has the very same speed as my 486SX 25 MHz notebook with ISA VGA GFX (cca 10 FPS in Doom benchmark) - it could be somewhat lower with only 25 MHz 68030, but Doom depends as much on fast GFX as on fast CPU.

Quote:
PC Mag 11 October 1993, Page 419


Are you sure, you didnt mean 11 October 1994? (it was 12 October in 1993 and there is no such advert on that page).


PC Mag 28 Sep 1993, Page 419
From MS Engineering Inc.
Barebones 486SX-25 VLB (128KB L2 cache, 4MB RAM, 125 MB HDD, FDD, Keyboard) = $750

Add the following items
ET-4000W.32 VLB 1MB video RAM = $129
Pro Audio 16: $159
Total: $1,038

--------------------
PC Mag 26 Oct 1993, page 397

From AST Computers
Bravo LC with 123W bundle
486SX 25Mhz (120Mb HDD, 4 MB RAM, 1 MB video RAM) = $999


In PC Mag 26 Oct 1993, page 389
From AST Computers
Bravo LC with 123W bundle
486SX 25Mhz (120Mb HDD, 4 MB RAM, 1 MB video RAM) = $999


From NEC Ready Desktops
486SX25 (4MB RAM, 170 MB HDD, FDD, Keyboard) = $999


From Leading Edge, WinPro 486 E
486SX25 (4MB RAM, 120 MB HDD, FDD, Keyboard, Windows Accelerator) = $970


From Compaq ProLinea Desktops
486SX25 (4MB RAM, 120 MB HDD, FDD, Keyboard, Mouse, VLB graphics) = $999



----------------------------

From Amiga World Magazine (November 1993), page 70 of 100,
Amiga 4000/030 (68030/68882 @ 25 Mhz) = $1599. LOL, C= plans to fail.

Amiga 4000/040 (68040 @ 25 Mhz) = $2299. LOL, C= plans to fail.



From USA's Amiga World Magazine (November 1993), page 58 of 100,
Price listed in USD in November 1993

A1200/020, 2MB, price $379
A3000/030 at 25Mhz, 5MB, 105HD, price $899
A3000T/030 at 25Mhz, 5MB, 200MB HDD, price $1199
A3000T/040 at 25Mhz, 5MB, 200MB HDD, price $1599
A3000s are missing the AGA chipset.


Cost estimate for 68040 card, $1599 - $1199, cost for 68040 A3640 card = $400

A1200's $379 + 68040 A3640 card's $400 = $779.

A3640 includes extra glue chips for supporting the Zorro III bus is not needed for the A1200 version.


Commodore could have out-of-the-box configured A1200 with 68040 at 25Mhz for slightly above $779 (i.e. add 4MB fast ram, small HDD) which could compete against $1000 out-of-the-box 486 33Mhz based PC. Amiga 1200 with 68LC040 + 4MB fast ram at $779 cost would be targeting Doom-type PC games.

Both Intel 80486 and Motorola 68040 are pipelined-based CPU designs.

Last edited by Hammer on 29-Jun-2021 at 05:49 AM.
Last edited by Hammer on 29-Jun-2021 at 05:44 AM.
Last edited by Hammer on 29-Jun-2021 at 05:43 AM.
Last edited by Hammer on 29-Jun-2021 at 05:41 AM.
Last edited by Hammer on 29-Jun-2021 at 05:35 AM.
Last edited by Hammer on 29-Jun-2021 at 05:33 AM.
Last edited by Hammer on 29-Jun-2021 at 05:21 AM.
Last edited by Hammer on 29-Jun-2021 at 05:21 AM.

_________________
Core i9-9900K, DDR4-3800 32 GB RAM, GeForce RTX 3080 Ti
Ryzen 9 3900X, DDR4-3200 32 GB RAM, GeForce RTX 2080 Ti
Amiga 1200 (rev 1D1, KS 3.2, TF1260, 68060 @ 63 Mhz, 128 MB)
Amiga 500 (rev 6A, KS 3.2, 68K 50Mhz, 12 MB RAM)

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor's Amiga Blog
Posted on 29-Jun-2021 5:47:07
#1506 ]
Elite Member
Joined: 9-Mar-2003
Posts: 4270
From: Australia

@matthey

Quote:

matthey wrote:
BigD Quote:

In my experience 030/50 plus AGA on ADoom was fine for Doom 2s early levels and only stuttered a bit on later ones (still completed it). Yes, Final Doom needed something with a bit more grunt but should C= have shipped a souped up A1200/Falcon Plus spec machine and if iD Software had been paid handsomely Im sure it would have sold Amigas but probably not quick enough to save C= unless AAA had been ready with chunky pixel support in 1991!


The 68030@50MHz only had a 256 byte ICache and 256 byte DCache so it was sensitive to memory performance. Unfortunately, CBM usually used the cheapest memory possible including chip memory which is part of the reason why ECS and AGA are slow and the reason why the Jay Miner designed Ranger chipset was rejected despite being ready in 1989 before ECS (Ranger would have had 128 colors as well as higher resolutions with better performance). Some 3rd party accelerators used high performance memory with a high clocked 68030 to give integer performance close to that of a 68040@25Mhz on a CBM accelerator with poor memory performance (and likely that of a cheap 486SX too).

I don't think AAA would have saved the Amiga even if it was available in 1991. AAA custom chips would have been expensive (many more transistors than AGA) so the chipset would have only been used for high end Amigas which were not selling in high enough volumes to lower the cost with mass production. The Ranger chipset, also using VRAM which gives twice the memory bandwidth, would have had a better chance at success if first introduced in high end Amigas as VRAM was being used in enough other computers to allow shared mass production of the VRAM by the manufacturers (shared mass production is what propelled the PC clone market as well). VRAM prices soon dropped to only 20% more expensive than DRAM which was the primary expense of the Ranger chipset and could have then been used for all Amigas. AAA benefited from the lower cost of VRAM but the custom chips would have remained too expensive for low end Amigas. AAA had poor compatibility with OCS/ECS as well. Just as AGA was determined to be a cheaper and more practical alternative to AAA, the AA+ chipset was more likely to be the next Amiga chipset with logic and specs taken from the permanently shelved AAA. SAGA is more along the lines of the AA+ chipset, minus some blitter functionality currently. It keeps good OCS/ECS/AGA compatibility but adds features like chunky gfx and 8 voice 16 bit audio planned for AAA.

Amiga Ranger with 2MB VRAM as Chip Ram was completed in 1987.

Amiga Ranger targeted 10241024 pixels with 128 colors (7-bit color depth) which is an overkill for 320x256/320x200 pixels with 256 color games.

A1200 has 32-bit bus DRAM instead of A500's 16-bit bus DRAM setup. AGA wasn't completely a 32-bit graphics chipset.

Sharp X68000 with VRAM was also a sales failure.

Unlike 68030, 486SX includes an 8 KB L1 cache unified code and data cache.

Last edited by Hammer on 29-Jun-2021 at 06:52 AM.
Last edited by Hammer on 29-Jun-2021 at 06:11 AM.
Last edited by Hammer on 29-Jun-2021 at 06:07 AM.
Last edited by Hammer on 29-Jun-2021 at 06:01 AM.
Last edited by Hammer on 29-Jun-2021 at 05:59 AM.

_________________
Core i9-9900K, DDR4-3800 32 GB RAM, GeForce RTX 3080 Ti
Ryzen 9 3900X, DDR4-3200 32 GB RAM, GeForce RTX 2080 Ti
Amiga 1200 (rev 1D1, KS 3.2, TF1260, 68060 @ 63 Mhz, 128 MB)
Amiga 500 (rev 6A, KS 3.2, 68K 50Mhz, 12 MB RAM)

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor's Amiga Blog
Posted on 29-Jun-2021 5:54:28
#1507 ]
Elite Member
Joined: 9-Mar-2003
Posts: 4270
From: Australia

@BigD

Quote:

BigD wrote:
@pavlor

In my experience 030/50 plus AGA on ADoom was fine for Doom 2s early levels and only stuttered a bit on later ones (still completed it). Yes, Final Doom needed something with a bit more grunt but should C= have shipped a souped up A1200/Falcon Plus spec machine and if iD Software had been paid handsomely Im sure it would have sold Amigas but probably not quick enough to save C= unless AAA had been ready with chunky pixel support in 1991!

A1200 with 68LC040 + 4 MB fast ram targeting Doom-type games SKU's purpose is to hold the line instead of being pincered by both 486SX PCs and SNES.

_________________
Core i9-9900K, DDR4-3800 32 GB RAM, GeForce RTX 3080 Ti
Ryzen 9 3900X, DDR4-3200 32 GB RAM, GeForce RTX 2080 Ti
Amiga 1200 (rev 1D1, KS 3.2, TF1260, 68060 @ 63 Mhz, 128 MB)
Amiga 500 (rev 6A, KS 3.2, 68K 50Mhz, 12 MB RAM)

 Status: Offline
Profile     Report this post  
pavlor 
Re: Trevor's Amiga Blog
Posted on 29-Jun-2021 15:47:43
#1508 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9255
From: Unknown

@Hammer

Quote:
Barebones 486SX-25 VLB (128KB L2 cache, 4MB RAM, 125 MB HDD, FDD, Keyboard) = $750


Too bad there is nothing more about motherboard configuration (eg. how many additional ISA slots are there?).

I dont think it is fair to compare mostly ISA based PCs with full 32 bit architecture like A4000. As you wrote, Commodore had no product for this market segment between A1200 and A4000.

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor's Amiga Blog
Posted on 1-Jul-2021 5:07:33
#1509 ]
Elite Member
Joined: 9-Mar-2003
Posts: 4270
From: Australia

@pavlor

Quote:

pavlor wrote:
@Hammer

Quote:
Barebones 486SX-25 VLB (128KB L2 cache, 4MB RAM, 125 MB HDD, FDD, Keyboard) = $750


Too bad there is nothing more about motherboard configuration (eg. how many additional ISA slots are there?).

I dont think it is fair to compare mostly ISA based PCs with full 32 bit architecture like A4000. As you wrote, Commodore had no product for this market segment between A1200 and A4000.

A4000 AGA wasn't a full 32-bit graphics chipset, but C='s AGA is faster than IBM VGA.

A3000/A4000 Buster can NOT sustain four Zorro III at the same i.e. are you claiming four 32 bit Zorro III slots have a bus backbone equivalent to 128 bits?


My cited 1993 era has a VLB slot which is 32-bit running from 25 Mhz or higher.

"VLB" in the "Barebones 486SX-25 VLB" statement is important since it has a VLB slot.

"Barebones 486SX-25 VLB" also has a 128 KB L2 cache on the motherboard which is missing on A4000/030 and A4000/040. A single VLB slot is needed for a fast VGA clone card.

Modern AMD X570 motherboard such as ASUS ROG X570 Strix can sustain a single PCI-E slot with 16 lanes at PCI-E version 4.0 performance. Using two PCI-E slots for 3D cards results in 8 lanes allocation for each slot. Workstation PC such as AMD ThreadRipper has expanded PCI-E beyond 24 lanes.

Legacy PCs with AGP have a single fast graphics slot.

The desktop PC market is used to a fast single graphics slot .


Last edited by Hammer on 01-Jul-2021 at 05:37 AM.
Last edited by Hammer on 01-Jul-2021 at 05:34 AM.
Last edited by Hammer on 01-Jul-2021 at 05:28 AM.
Last edited by Hammer on 01-Jul-2021 at 05:13 AM.
Last edited by Hammer on 01-Jul-2021 at 05:12 AM.
Last edited by Hammer on 01-Jul-2021 at 05:08 AM.

_________________
Core i9-9900K, DDR4-3800 32 GB RAM, GeForce RTX 3080 Ti
Ryzen 9 3900X, DDR4-3200 32 GB RAM, GeForce RTX 2080 Ti
Amiga 1200 (rev 1D1, KS 3.2, TF1260, 68060 @ 63 Mhz, 128 MB)
Amiga 500 (rev 6A, KS 3.2, 68K 50Mhz, 12 MB RAM)

 Status: Offline
Profile     Report this post  
MEGA_RJ_MICAL 
Re: Trevor's Amiga Blog
Posted on 1-Jul-2021 8:51:22
#1510 ]
Cult Member
Joined: 13-Dec-2019
Posts: 507
From: AMIGAWORLD.NET WAS ORIGINALLY FOUNDED BY DAVID DOYLE

@matthey
@hammer

Dear friends,

Is there any other thread you might want to turn into endless banter about half clock cycles and the coating of the 7th capacitor from the left?

I can open a new one for you if you please!
The topic doesn't matter, does it?




Re: Cinnamon buns recipe
by matthey-hammer on 1-Jul-2021 5:30:00

No!! The oven cooks better @ 3.22MHz if the RISC reg-mem architecture has a 32 bit ALU, 16 bit data bus, 24 bit address bus, 32 bit C0-A9, F0-B7 (10 GP registers) 16*16, 32/16 just look at this benchmark from 1979

ALU-Instructions 16MB 64KB
-----------------------------------------------------------
nop 2929.4 3009.9
add-reg 884.9 925.7
add-im16 903.0 897.2
shift-reg 531.5 542.5
shift-im8 537.3 536.1
and-reg 916.5 925.8
and-im16. 542.4 540.6
mul-reg 294.4 294.2
mul-im16 198.5 202.8
div-reg 32.2 32.6
-----------------------------------------------------------
ALU-Workloads 16MB 64KB
-----------------------------------------------------------
Workload AAAA 1010.7 999.8
Workload LA 666.3 674.2
Workload LAA 790.6 792.7
Workload LAAA 803.4 802.4
Workload LAAAA 862.2 862.6
Workload LLA 610.9 619.4
Workload LLAA 723.0 721.7
Workload LLAAA 673.3 686.9
Workload LLAAAA 797.9 795.2
Workload LAALA 688.5 673.7
WORKLOAD addq 1019.9 1027.6
WORKLOAD add.w 1029.9 1017.5
WORKLOAD add.l 302.7 306.2
WORKLOAD lea 567.5 576.0
Work 26 388.7 399.3
Work 2266 455.4 472.1
Work 4x2 4x6 474.6 475.8
Work 8x2 8x6 471.0 471.1
WORK memadd 301.9 307.8
WORK memadd++ 302.1 305.3
-----------------------------------------------------------
ALU-Addressmodes 16MB 64KB
-----------------------------------------------------------
LOAD ea-im(r) 527.8 523.2
LOAD ea-(r)++ 377.7 375.2
LOAD ea-im(r)+A 329.7 328.1
LOAD ea-(r,r) 331.0 325.0
write ea-im(r) 521.4 531.1
write ea-(r)++ 296.1 296.2
write ea-im(r)A 291.5 293.6
write ea-(r,r) 290.9 294.3
-----------------------------------------------------------
ALU-Branches 16MB 64KB
-----------------------------------------------------------
gosub chain-1 8.9 8.9
gosub chain-4 12.4 12.4
gosub chain-8 13.2 13.2
gosub chain-16 13.7 13.7
gosub chain-32 14.2 14.2
BCC IF(FALSE) 140.2 140.9
BCC IF(TRUE) 69.1 69.0
BCC IF(50%) 82.5 82.1
BCC IF(25%) 75.7 75.3
goto-bra 377.1 377.0
goto-bra + 2A 83.6 83.7
goto-bra + 4A 74.0 74.1
-----------------------------------------------------------
ALU-Latencies 16MB 64KB
-----------------------------------------------------------
add-reg-2loop 165.0 164.0
add-reg-4loop 292.6 292.6
add-reg-6loop 394.0 390.2
add-reg-8loop 477.4 477.3
add-reg-16loop 696.4 695.9
add-reg-32loop 924.3 925.7
add-reg-64loop 1078.0 1062.7
add-reg-128loop 1185.1 1200.0
alu_latency+0 529.8 518.9
alu_latency+1 840.0 822.1
alu_latency+2 830.5 837.8
alu_latency+3 0.1 836.1
alu_latency+4 0.1 795.3
cache_latency+0 269.6 269.1
cache_latency+1 262.7 263.4
cache_latency+2 210.6 209.3
alu->ea_lat+0 294.1 293.6
alu->ea_lat+1 225.2 224.6
alu->ea_lat+2 187.3 187.5
-----------------------------------------------------------
Measuring memory latency:
Result is million random accesses per sec.
Higher value is faster.
Memory Latency 16MB
-----------------------------------------------------------
rand-read 2MB 5.2
rand-read 4MB 5.2
rand-read 8MB 5.1
rand-read 16MB 3.4
-----------------------------------------------------------
Measuring memory throughput:
Results are in MB/sec. Higher value is faster.
Memory 2 Memory
Alignment 0-0 16MB 64KB 4KB 1KB
-----------------------------------------------------------
libc memcpy 128.6 127.3 125.7 115.9
read 8 44.9 45.0 45.0 45.0
read 8x4 69.8 70.0 69.9 70.0
read 32 128.7 128.3 129.0 128.8
read 32x4 178.1 177.8 177.6 177.6
write 8 57.9 57.7 57.8 58.1
write 8x4 36.3 36.1 36.2 36.3
write 32 117.9 117.9 117.2 117.8
write 32x4 123.4 123.6 123.5 123.5
copy 8 69.5 69.7 69.2 69.8
copy 8x4 98.9 98.8 99.0 98.1
copy 32 123.9 123.5 123.3 122.3
copy 32x4 129.0 128.8 129.0 127.4
-----------------------------------------------------------
Cache 2 Cache
Alignment 0-0 16MB 64KB 4KB 1KB
-----------------------------------------------------------
libc memcpy 128.9 900.6 1765.2 875.8
read 8 45.1 53.6 54.7 54.5
read 8x4 70.0 94.0 97.4 97.3
read 32 128.9 244.5 263.0 264.0
read 32x4 177.2 559.9 660.4 653.4
write 8 57.8 71.8 73.0 73.0
write 8x4 36.2 50.7 51.5 51.3
write 32 117.9 288.6 291.7 292.4
write 32x4 123.7 699.2 703.6 705.7
copy 8 69.7 115.2 119.1 119.3
copy 8x4 98.7 207.7 218.4 218.7
copy 32 123.7 433.8 476.7 475.8
copy 32x4 129.5 790.5 948.7 937.0
-----------------------------------------------------------

_________________
I HAVE ABS OF STEEL
--
CAN YOU SEE ME? CAN YOU HEAR ME? OK FOR WORK

 Status: Offline
Profile     Report this post  
NutsAboutAmiga 
Re: Trevor's Amiga Blog
Posted on 1-Jul-2021 11:16:15
#1511 ]
Elite Member
Joined: 9-Jun-2004
Posts: 11927
From: Norway

@MEGA_RJ_MICAL

When the spammer complains about spamming you know it gone too far.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

 Status: Offline
Profile     Report this post  
pavlor 
Re: Trevor's Amiga Blog
Posted on 1-Jul-2021 15:25:10
#1512 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9255
From: Unknown

@Hammer

Quote:
are you claiming four 32 bit Zorro III slots have a bus backbone equivalent to 128 bits?


Of course no. I mean comparing a computer with 32 bit bus with 4 slots and another computer with one 32 bit slot and other 16 bit slots is misleading. 486DX based computers with several 32 bit slots (VLB or even EISA) did cost a lot more back then (just see product comparison in one of the magazines you linked).

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor's Amiga Blog
Posted on 2-Jul-2021 7:16:20
#1513 ]
Elite Member
Joined: 9-Mar-2003
Posts: 4270
From: Australia

@pavlor

Quote:

pavlor wrote:
@Hammer

Quote:
are you claiming four 32 bit Zorro III slots have a bus backbone equivalent to 128 bits?


Of course no. I mean comparing a computer with 32 bit bus with 4 slots and another computer with one 32 bit slot and other 16 bit slots is misleading. 486DX based computers with several 32 bit slots (VLB or even EISA) did cost a lot more back then (just see product comparison in one of the magazines you linked).

Your argument doesn't help with mass-producing Amiga gaming system with 68LC040 targeting Doom games. A4000 wasn't designed for mass production at the same level as A1200s.

Certain add-on cards do NOT need 32-bit slots e.g. my 1996 era Pentium 150 PC was bundled with ISA PnP Yamaha 16 bit sound card.

My argument was about an A1200 with 68LC040 + 4MB fast ram as an official C= SKU in the alternate timeline similar to the real-life October 2013's Macintosh Quadra 605 with 68LC040 and introductory price of $1000.

For the 1993 time period, Apple responded against barebone $790-to-$1000 486SX VLB based PCs. There's NO response from C=.

Your argument is repeating the same C= stupid management direction that exposed a price gap between baseline A1200 and expensive A4000/030. To this day, classic Amigas have cost problems exceeding f_cking 68030. I hated Amiga Technologies GmBH's Walker 68030 specs for the 1995 time period.





Last edited by Hammer on 02-Jul-2021 at 07:29 AM.
Last edited by Hammer on 02-Jul-2021 at 07:27 AM.

_________________
Core i9-9900K, DDR4-3800 32 GB RAM, GeForce RTX 3080 Ti
Ryzen 9 3900X, DDR4-3200 32 GB RAM, GeForce RTX 2080 Ti
Amiga 1200 (rev 1D1, KS 3.2, TF1260, 68060 @ 63 Mhz, 128 MB)
Amiga 500 (rev 6A, KS 3.2, 68K 50Mhz, 12 MB RAM)

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor's Amiga Blog
Posted on 2-Jul-2021 7:37:04
#1514 ]
Elite Member
Joined: 9-Mar-2003
Posts: 4270
From: Australia

@NutsAboutAmiga


Quote:

NutsAboutAmiga wrote:
@MEGA_RJ_MICAL

When the spammer complains about spamming you know it gone too far.



My post about A1200 with 68LC040 as an official C= in the alternative (What IF) timeline is related to A1222.

I don't want to see yet another baseline Amiga with 68020/68030 level CPU! I hated Amiga Technologies gmbh's Walker 68030 specs for the 1995 time period. These 68030 level machine specs do NOT progress the Amiga platform.

Trevor's attempt to redefine baseline Amiga as A1222 level specs at the Vampire price range is commendable.

Last edited by Hammer on 02-Jul-2021 at 07:40 AM.

_________________
Core i9-9900K, DDR4-3800 32 GB RAM, GeForce RTX 3080 Ti
Ryzen 9 3900X, DDR4-3200 32 GB RAM, GeForce RTX 2080 Ti
Amiga 1200 (rev 1D1, KS 3.2, TF1260, 68060 @ 63 Mhz, 128 MB)
Amiga 500 (rev 6A, KS 3.2, 68K 50Mhz, 12 MB RAM)

 Status: Offline
Profile     Report this post  
pavlor 
Re: Trevor's Amiga Blog
Posted on 2-Jul-2021 14:04:30
#1515 ]
Elite Member
Joined: 10-Jul-2005
Posts: 9255
From: Unknown

@Hammer

Quote:
Your argument


That was not my argument.

I wrote there was no product from Commodore for a market between A1200 and A4000 and that comparison of a ISA PC with A4000 is not fair. Nothing less, nothing more.

Commodore had nearly exactly the same strategy since the great success of C64:

1) Cheap low margin computers as a basis of company growth
2) Underfunded research/development

Commodore lost higher margin PC market with the failure of CBM II and its later attempts to return were unsuccessful (A2500, A3000 etc.). There was also no will to invest revenues into R/D (to the astonishment of then day press). Deadly outcome for Commodore could be foreseen nearly decade before its bancruptcy: its stagnant R/D strategy combined with low margins was not sustainable in a market, where thechnology progress accelerated.

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor's Amiga Blog
Posted on 3-Jul-2021 10:22:12
#1516 ]
Elite Member
Joined: 9-Mar-2003
Posts: 4270
From: Australia

@pavlor

Quote:

pavlor wrote:
@Hammer

Quote:
Your argument


That was not my argument.

I wrote there was no product from Commodore for a market between A1200 and A4000 and that comparison of a ISA PC with A4000 is not fair. Nothing less, nothing more.

A4000's four 32 bit ZIII slots advantage means little when my cited 486SX PC examples have at least a fast VLB slot and ISA slots are enough for 16 bit sound cards.

Cost reduced A4000 can be sold without the secondary bus board, but A1200 base hardware is cheaper for mass production.

PC Mag 26 Oct 1993, Page 138
From Micron Computers Inc.

425VL ValueLine CD = $1299 includes
CPU: 486SX 25Mhz
RAM: 4MB
Slots, two VLB , 6 ISA
CD-ROM, FDD
Graphics Accelerator 1MB
14" Monitor
Mouse, Keyboard
MS-DOS, Windows 3.1



Quote:

Commodore had nearly exactly the same strategy since the great success of C64:

1) Cheap low margin computers as a basis of company growth
2) Underfunded research/development

Commodore lost higher margin PC market with the failure of CBM II and its later attempts to return were unsuccessful (A2500, A3000 etc.). There was also no will to invest revenues into R/D (to the astonishment of then day press). Deadly outcome for Commodore could be foreseen nearly decade before its bancruptcy: its stagnant R/D strategy combined with low margins was not sustainable in a market, where thechnology progress accelerated.


1. At the low entry price segment, C64's success didn't factor in Nintendo's exclusive games business model competition. C64 business model can not compete against Nintendo's tax subsidizing hardware business model.

Nintendo's business model was the basis for Sony's Playstation and Microsoft's Xbox business model copycats.

C64's graphics specs were competitive from 1982-to-1985.

C= has no short-term plans to counter SNES SuperFX2's "can also run Doom" PR tick box.

C= has no short-term plans to counter Jack Carmack's anti-Amiga "can't run Doom" mindshare statement.

C= is aware of the certain chunky graphics format game from the X86 PC, hence CD32's Akiko chip, but the 68EC020 CPU is hamstrung without fast memory.

Apple has Macintosh Quadra 605 with 68LC040 @25Mhz and an entry $1000 price in October 2013. Doom was officially released in 1994 for the Macintosh.

2. C64's success was built on taking over MOS and run it down. C64 was largely driven by "less crap than VIC-20" Jack Tramiel's administration.

C64's follow-on C128 wasn't successful since it's just aging C64 gaming hardware with low color high-resolution business mode (notice the pattern). Amiga ECS was aging gaming hardware with low color high-resolution business mode.

To replace C64 65xx, C= needs technology injection from the Amiga Corporation.
To replace Amiga 68K, C= needs technology injection from HP's PA-RISC.

C='s in-house 16-bit computer was crap i.e. Commodore 900 (Zilog Z8000 CPU, high-resolution monochrome). C128 has a Z80 CPU with its high-resolution monochrome mode. IBM PC EGA is already superior to Commodore 900's high-resolution monochrome.

Jack Tramiel's Atari ST is superior when compared to Commodore 900.

Prior to C='s takeover of MOS, 65xx CPU was competitive against Motorola's 6800. MOS 65xx didn't evolve like the X86.

Last edited by Hammer on 03-Jul-2021 at 11:08 AM.
Last edited by Hammer on 03-Jul-2021 at 10:50 AM.
Last edited by Hammer on 03-Jul-2021 at 10:46 AM.
Last edited by Hammer on 03-Jul-2021 at 10:39 AM.
Last edited by Hammer on 03-Jul-2021 at 10:37 AM.
Last edited by Hammer on 03-Jul-2021 at 10:26 AM.
Last edited by Hammer on 03-Jul-2021 at 10:23 AM.

_________________
Core i9-9900K, DDR4-3800 32 GB RAM, GeForce RTX 3080 Ti
Ryzen 9 3900X, DDR4-3200 32 GB RAM, GeForce RTX 2080 Ti
Amiga 1200 (rev 1D1, KS 3.2, TF1260, 68060 @ 63 Mhz, 128 MB)
Amiga 500 (rev 6A, KS 3.2, 68K 50Mhz, 12 MB RAM)

 Status: Offline
Profile     Report this post  
Hammer 
Re: Trevor's Amiga Blog
Posted on 3-Jul-2021 11:34:16
#1517 ]
Elite Member
Joined: 9-Mar-2003
Posts: 4270
From: Australia

@pavlor

Quote:

pavlor wrote:
@Hammer

Quote:
A4000/030 wasn't good enough for Doom.


486 and slow ISA GFX could reach a similar result. My A1200 with somewhat faster CPU (50 MHz) has the very same speed as my 486SX 25 MHz notebook with ISA VGA GFX (cca 10 FPS in Doom benchmark) - it could be somewhat lower with only 25 MHz 68030, but Doom depends as much on fast GFX as on fast CPU.


https://thandor.net/benchmark/32
For Doom benchmark,
A fast SVGA clone with ISA slot can reach 27.10 fps e.g. SuperVGA ET4000 1MB ISA or ET4000AX 1MB ISA with Pentium 100.

IBM VGA (roll-down frame rates) is slow regardless of Athlon XP at 1800 Mhz.

IBM VGA bottleneck is worst than the ISA slot.

I know IBM PS/2 Model 55SX (386SX16, 5MB RAM) with IBM VGA has roll-down frame rates with Pinball Fantasies i.e. A500 is superior.

I cited the ET4000 1MB card for $129 for a reason.


https://www.youtube.com/watch?v=KQDEKoRcXZc&t=259s
386DX33 with Tseng Labs ET4000 SVGA running Doom.

Last edited by Hammer on 04-Jul-2021 at 04:58 PM.
Last edited by Hammer on 03-Jul-2021 at 11:42 AM.
Last edited by Hammer on 03-Jul-2021 at 11:38 AM.
Last edited by Hammer on 03-Jul-2021 at 11:37 AM.

_________________
Core i9-9900K, DDR4-3800 32 GB RAM, GeForce RTX 3080 Ti
Ryzen 9 3900X, DDR4-3200 32 GB RAM, GeForce RTX 2080 Ti
Amiga 1200 (rev 1D1, KS 3.2, TF1260, 68060 @ 63 Mhz, 128 MB)
Amiga 500 (rev 6A, KS 3.2, 68K 50Mhz, 12 MB RAM)

 Status: Offline
Profile     Report this post  
Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 | 60 | 61 | 62 | 63 | 64 | 65 | 66 | 67 | 68 | 69 | 70 | 71 | 72 | 73 | 74 | 75 | 76 )

[ home ][ about us ][ privacy ] [ forums ][ classifieds ] [ links ][ news archive ] [ link to us ][ user account ]
Copyright (C) 2000 - 2019 Amigaworld.net.
Amigaworld.net was originally founded by David Doyle