diff options
| author | Bryan Newbold <bnewbold@archive.org> | 2017-08-02 13:02:55 -0700 |
|---|---|---|
| committer | Bryan Newbold <bnewbold@archive.org> | 2017-11-21 16:06:28 -0800 |
| commit | a32854cc0297a87e266142b5a6a2f995263e571f (patch) | |
| tree | db1a74500912e33ada938cdaa6087a91b6f036a4 /networking | |
| parent | 331d9f78dbfd8bcdebfb14a4d87ba1b91adb0110 (diff) | |
| download | knowledge-a32854cc0297a87e266142b5a6a2f995263e571f.tar.gz knowledge-a32854cc0297a87e266142b5a6a2f995263e571f.zip | |
networking/bandwidth: latency and more granularity
Diffstat (limited to 'networking')
| -rw-r--r-- | networking/bandwidth.page | 27 |
1 files changed, 19 insertions, 8 deletions
diff --git a/networking/bandwidth.page b/networking/bandwidth.page index c9b2bee..f516087 100644 --- a/networking/bandwidth.page +++ b/networking/bandwidth.page @@ -1,12 +1,13 @@ -Throughput data/hour data/day data/month data/year -------------- ---------- ------------ ------------ ------------ -128 Kbit/sec 57 MByte 1.4 GByte 41.5 GByte 505.6 GByte -1 Mbit/sec 450 MByte 10.8 GByte 324 GByte 3.8 TByte -20 Mbit/sec 9 Gbyte 100.8 GByte 6.48 TByte 78.8 TByte -1 Gbit/sec 450 GByte 10.8 TByte 324 TByte 3.94 PByte -40 Gbit/sec 18 TByte 432 TByte 13 PByte 158 PByte -1 Tbit/sec 360 TByte 10.8 PByte 324 PByte 3.94 EByte +Throughput data/hour data/day data/month data/year +--------------- ------------ ------------ ------------ ------------ +128 Kbit/sec 58.3 MByte 1.4 GByte 41.5 GByte 505.6 GByte +1 Mbit/sec 450 MByte 10.8 GByte 324 GByte 3.8 TByte +20 Mbit/sec 4.2 GByte 100.8 GByte 6.48 TByte 78.8 TByte +100 Mbit/sec 21 GByte 504.0 GByte 32.4 TByte 394.0 TByte +1 Gbit/sec 450 GByte 10.8 TByte 324 TByte 3.94 PByte +40 Gbit/sec 18 TByte 432.0 TByte 12.9 PByte 157.6 PByte +1 Tbit/sec 360 TByte 10.8 PByte 324 PByte 3.94 EByte 200 GByte a month is about 610 Kbit/sec ("kbps") continuous. 24 TByte a month is about 74 MBit/sec continuous. @@ -14,3 +15,13 @@ Throughput data/hour data/day data/month data/year Networks must be provisioned for "peak use" periods, which likely correlate for users in the same region, so enforcing or charging for total throughput doesn't seem to make much sense? + + +Rate Latency Ops/hour 1 mil ops 100 mil ops +--------- --------- --------- ----------- ------------ +1/sec 1 sec 3.6k 11.6 days 3.17 years +20/sec 50 ms 72k 13.8 hours 57.5 days +100/sec 10 ms 360k 2.7 hours 11.25 days +250/sec 4 ms 0.9mil 1.1 hours 4.58 days +1k/sec 1 ms 3.6mil 16.6 min 27.6 hours + |