Sujet : Re: Is Parallel Programming Hard, And, If So, What Can You Do About It?
De : cr88192 (at) *nospam* gmail.com (BGB)
Groupes : comp.archDate : 23. May 2025, 17:29:03
Autres entêtes
Organisation : A noiseless patient Spider
Message-ID : <100q848$5fc7$2@dont-email.me>
References : 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
User-Agent : Mozilla Thunderbird
On 5/23/2025 7:36 AM, MitchAlsup1 wrote:
On Fri, 23 May 2025 6:09:37 +0000, BGB wrote:
On 5/23/2025 12:35 AM, Lawrence D'Oliveiro wrote:
On Fri, 23 May 2025 00:18:53 -0000 (UTC), Waldek Hebisch wrote:
>
It is pretty clear that due to drive mechanics track cache/buffer is
useful.
>
Only if you don’t take the statistics of real-world cache behaviour into
account.
>
However, the real question is about size: how big should it be.
>
It can never be big enough to make a difference.
>
Sometimes, it is not the size of the cache that matters.
>
>
Say, for example, a typical cache configuration in my core:
L1 D$: 32K, direct mapped
L1 I$: 16K, direct mapped
L2: 256K, direct mapped.
>
OK, so say I stick a 4K cache between the L1 and L2 caches.
Seems kinda useless just based on sizes.
It is called a "victim buffer"
I had usually understood it that a "victim buffer" was typically glued directly to the L1 cache. This was outside the L1 cache, and after the main TLB.
So, the L1 ring consisted of:
Join/VC -> L1D$ -> L1I$ -> TLB -> Join/VC
The Join/VC stage was either a joiner between the L1 ring and L2 ring, or the 4K victim cache (optional, but does help).
The L2 ring generally had the L2 cache, ROMs, a bridge to the MMIO bus, and the display and rasterizer modules (which had native interfaces on the ringbus).
Except: This small cache is 4-way set associative and so can absorb a
bunch of conflict misses, and notably reducing the number of cache
misses in the L2 cache. It can bring a performance benefit, despite
being small.
And generally fully associative
On the FPGA, the cost difference between a 4-way 4K cache and 4-entry (full assoc) cache was small, but the using LUTRAMs gave a better hit-rate.
The cost difference between 4-way and 8-way was a lot more steep (and seemingly doesn't improve hit rate enough to compensate for the fairly significant cost increase).
I suspect it might be different tradeoffs on an ASIC though.
Why?
Because it helps.
Yes.
Date | Sujet | # | | Auteur |
10 May 25 | Is Parallel Programming Hard, And, If So, What Can You Do About It? | 164 | | Thomas Koenig |
11 May 25 |  Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 133 | | MitchAlsup1 |
11 May 25 |   Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 132 | | Al Kossow |
12 May 25 |    Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 131 | | Lawrence D'Oliveiro |
12 May 25 |     Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 130 | | MitchAlsup1 |
12 May 25 |      Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 129 | | Lawrence D'Oliveiro |
12 May 25 |       Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | MitchAlsup1 |
12 May 25 |        Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Lawrence D'Oliveiro |
12 May 25 |       Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 126 | | Terje Mathisen |
12 May 25 |        Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 3 | | Anton Ertl |
12 May 25 |         Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Lawrence D'Oliveiro |
13 May 25 |         Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Stephen Fuld |
12 May 25 |        Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 122 | | Lawrence D'Oliveiro |
13 May 25 |         Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 4 | | Stefan Monnier |
13 May 25 |          Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 3 | | Lawrence D'Oliveiro |
13 May 25 |           Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Stefan Monnier |
13 May 25 |           Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Stephen Fuld |
13 May 25 |         Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 116 | | Anton Ertl |
13 May 25 |          Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 115 | | Lawrence D'Oliveiro |
13 May 25 |           Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 114 | | Stephen Fuld |
14 May 25 |            Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 61 | | Lawrence D'Oliveiro |
14 May 25 |             Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 58 | | Stephen Fuld |
18 May 25 |              Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 57 | | Lawrence D'Oliveiro |
19 May 25 |               Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 6 | | Lynn Wheeler |
19 May 25 |                Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 5 | | Vir Campestris |
19 May 25 |                 Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 4 | | Stephen Fuld |
20 May 25 |                  Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Vir Campestris |
21 May03:38 |                  Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | Lynn Wheeler |
21 May04:49 |                   Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Stephen Fuld |
19 May 25 |               Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 50 | | MitchAlsup1 |
20 May 25 |                Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 49 | | MitchAlsup1 |
20 May 25 |                 Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 48 | | Lawrence D'Oliveiro |
20 May 25 |                  Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 47 | | BGB |
20 May 25 |                   Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 45 | | Stefan Monnier |
20 May 25 |                    Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | BGB |
20 May19:35 |                     Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Stephen Fuld |
21 May01:29 |                    Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 42 | | Lawrence D'Oliveiro |
21 May02:08 |                     Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 37 | | BGB |
21 May04:46 |                      Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 36 | | Lawrence D'Oliveiro |
21 May04:58 |                       Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 35 | | Stephen Fuld |
21 May18:19 |                        Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 34 | | Anton Ertl |
22 May03:19 |                         Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 32 | | George Neuner |
22 May07:51 |                          Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 31 | | BGB |
22 May13:12 |                           Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 30 | | Torbjorn Lindgren |
22 May18:39 |                            Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 29 | | BGB |
22 May23:41 |                             Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 28 | | Lawrence D'Oliveiro |
23 May00:36 |                              Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 27 | | BGB |
23 May15:21 |                               Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 24 | | John Levine |
23 May16:17 |                                Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 17 | | MitchAlsup1 |
23 May17:57 |                                 Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 16 | | Stephen Fuld |
23 May18:43 |                                  Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 15 | | MitchAlsup1 |
24 May01:26 |                                   Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 12 | | BGB |
24 May18:32 |                                    Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 11 | | BGB |
24 May21:36 |                                     Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 4 | | John Levine |
24 May22:45 |                                      Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 3 | | Michael S |
24 May22:54 |                                       Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | BGB |
26 May21:09 |                                        Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Terje Mathisen |
24 May22:07 |                                     Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | MitchAlsup1 |
24 May23:26 |                                      Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | BGB |
25 May21:24 |                                     Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 4 | | Lars Poulsen |
25 May21:47 |                                      Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | John Levine |
25 May21:51 |                                      Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | BGB |
26 May00:23 |                                      Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | MitchAlsup1 |
24 May23:17 |                                   Re: recycling | 2 | | Brian G. Lucas |
25 May07:24 |                                    Re: recycling | 1 | | George Neuner |
23 May16:53 |                                Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 6 | | BGB |
23 May18:34 |                                 Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 5 | | BGB |
24 May06:38 |                                  Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 4 | | BGB |
24 May17:57 |                                   Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | MitchAlsup1 |
24 May20:24 |                                    Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | BGB |
24 May22:45 |                                   Re: the power of junk, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | BGB |
23 May17:34 |                               Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | jseigh |
23 May17:39 |                                Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | BGB |
22 May12:32 |                         Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Dan Cross |
21 May04:54 |                     Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 4 | | Stephen Fuld |
21 May06:39 |                      Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 3 | | Lawrence D'Oliveiro |
21 May07:42 |                       Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Stephen Fuld |
21 May08:08 |                       Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | BGB |
21 May01:57 |                   Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Lawrence D'Oliveiro |
14 May 25 |             Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | John Levine |
14 May 25 |              Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Thomas Koenig |
23 May01:18 |            Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 52 | | Waldek Hebisch |
23 May06:35 |             Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 5 | | Lawrence D'Oliveiro |
23 May07:09 |              Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 4 | | BGB |
23 May13:36 |               Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 3 | | MitchAlsup1 |
23 May17:29 |                Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | BGB |
24 May04:17 |                 Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Lawrence D'Oliveiro |
23 May16:28 |             Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 46 | | Stephen Fuld |
23 May22:03 |              Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 44 | | Stefan Monnier |
24 May17:23 |               Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 39 | | Stephen Fuld |
25 May19:05 |                Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 38 | | Waldek Hebisch |
25 May20:13 |                 Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 23 | | Stephen Fuld |
25 May20:36 |                  Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 12 | | MitchAlsup1 |
26 May05:23 |                   Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 11 | | Chris M. Thomasson |
26 May18:42 |                    Re: fuzzy disks, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 10 | | John Levine |
26 May19:16 |                     Re: fuzzy disks, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 9 | | Stephen Fuld |
26 May19:58 |                      Re: fuzzy disks, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 7 | | Stefan Monnier |
26 May20:19 |                       Re: fuzzy disks, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 5 | | John Levine |
27 May05:52 |                        Re: fuzzy disks, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 4 | | Stephen Fuld |
27 May09:34 |                         Re: fuzzy disks, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 3 | | Anton Ertl |
27 May15:34 |                          Re: fuzzy disks, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | Stephen Fuld |
26 May21:36 |                       Re: fuzzy disks, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Stephen Fuld |
26 May21:45 |                      Re: fuzzy disks, Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Stephen Fuld |
26 May00:16 |                  Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 6 | | Stefan Monnier |
26 May08:13 |                  Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | Lawrence D'Oliveiro |
26 May20:20 |                  Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 2 | | Waldek Hebisch |
25 May21:21 |                 Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | BGB |
26 May07:46 |                 Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 13 | | Anton Ertl |
25 May21:55 |               Drive Caches (Re: Is Parallel Programming Hard, ...) | 4 | | Lars Poulsen |
23 May23:19 |              Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Lawrence D'Oliveiro |
18 May 25 |         Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Lawrence D'Oliveiro |
19 May 25 |  Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 29 | | quadibloc |
21 May02:31 |  Re: Is Parallel Programming Hard, And, If So, What Can You Do About It? | 1 | | Chris M. Thomasson |