Re:Tesla's TTPoE at Hot Chips 2024: Replacing TCP for Low La...
by throw0101d on Wednesday August 28, 2024 @11:48AM
(#41380825)
> One that's been optimized for better performance on cheaper hardware than you can get with TCP h/w offload.
How many ≥10 Gbps chipsets that you'd find in a typical server do not have offload nowadays?
Further, once you're in the ≥50 Gbps card range you can often get ROCE, which helps with things like latency.
Re:Tesla's TTPoE at Hot Chips 2024: Replacing TCP for Low La...
by lallysingh on Wednesday August 28, 2024 @12:08PM
(#41381070)
And you're still paying the other performance and efficiency costs of TCP. ROCE also isn't a magic bullet.
Every system has a cost and tradeoffs. Just because someone took an unusual path doesn't mean that they were wrong. And the larger and more specialized their use case is, the less likely that a generic solution is the best match.
Re:Tesla's TTPoE at Hot Chips 2024: Replacing TCP for Low La...
by chipdart on Saturday August 31, 2024 @05:25PM
(#41412123)
> And you're still paying the other performance and efficiency costs of TCP. ROCE also isn't a magic bullet.
Tesla's own charts show ROCE achieving also one-way write latencies in the single-digit microsecond range. If that doesn't qualify as magic bullet, what does that say about TTPoE?
(Disclaimer: I work at Tesla, not related to this group, opinions on public info only)