A decade ago, when traditional machine learning techniques were first being commercialized, training was incredibly hard and expensive, but because models were relatively small, inference – running new data through a model to cause an application to act or react – was easy. …
For Financial Services Firms, AI Inference Is As Challenging As Training was written by Timothy Prickett Morgan at The Next Platform.
Kubernetes networking is deceptively simple on the surface, until it breaks, silently leaks data, or opens the door to a full-cluster compromise. As modern workloads become more distributed and ephemeral, traditional logging and metrics just can’t keep up with the complexity of cloud-native traffic flows.
That’s where Calico Whisker comes in. Whisker is a lightweight Kubernetes-native observability tool created by Tigera. It offers deep insights into real-time traffic flow patterns, without requiring you to deploy heavyweight service meshes or packet sniffer. And here’s something you won’t get anywhere else: Whisker is data plane-agnostic. Whether you run Calico eBPF data plane, nftables, or iptables, you’ll get the same high-fidelity flow logs with consistent fields, format, and visibility. You don’t have to change your data plane, Whisker fits right in and shows you the truth, everywhere.
Let’s walk through 5 network issues Whisker helps you catch early, before they turn into outages or security incidents.
Traditional observability tools often show whether a packet was forwarded, accepted or dropped, but not why. They lack visibility into which Kubernetes network policy was responsible or if one was even applied.
With Whisker, each network flow is paired with:
Generative AI is arguably the most complex application that humankind has ever created, and the math behind it is incredibly complex even if the results are simple enough to understand. …
Skimpy HBM Memory Opens Up The Way For AI Inference Memory Godbox was written by Timothy Prickett Morgan at The Next Platform.
SPONSORED FEATURE The hyperscalers and big cloud builders have their own technical and political reasons for designing proprietary Arm server chips and having them built by Taiwan Semiconductor Manufacturing Co. …
AMD EPYC Is A More Universal Hybrid Cloud Substrate Than Arm was written by Timothy Prickett Morgan at The Next Platform.
AI agents bring with them the promise of being able to autonomously solve complex tasks put before them, from finding and analyzing the necessary data, choosing tools, and making decisions without human intervention to learning from their mistakes and adapting to changes. …
Cisco’s Outshift Incubator Sends Agentic AI Protocol To The Linux Foundation was written by Jeffrey Burt at The Next Platform.
Both the global economy and spending on information technology are so vast that it is hard to really grasp the numbers sometimes. …
How To Cash In On Massive Datacenter Spending was written by Timothy Prickett Morgan at The Next Platform.
Linux 6.11+ kernels provide TCX attachment points for eBPF programs to efficiently examine packets as they ingress and egress the host. The latest version of the open source Host sFlow agent includes support for TCX packet sampling to stream industry standard sFlow telemetry to a central collector for network wide visibility, e.g. Deploy real-time network dashboards using Docker compose describes how to quickly set up a Prometheus database and use Grafana to build network dashboards.
static __always_inline void sample_packet(struct __sk_buff *skb, __u8 direction) {
__u32 key = skb->ifindex;
__u32 *rate = bpf_map_lookup_elem(&sampling, &key);
if (!rate || (*rate > 0 && bpf_get_prandom_u32() % *rate != 0))
return;
struct packet_event_t pkt = {};
pkt.timestamp = bpf_ktime_get_ns();
pkt.ifindex = skb->ifindex;
pkt.sampling_rate = *rate;
pkt.ingress_ifindex = skb->ingress_ifindex;
pkt.routed_ifindex = direction ? 0 : get_route(skb);
pkt.pkt_len = skb->len;
pkt.direction = direction;
__u32 hdr_len = skb->len < MAX_PKT_HDR_LEN ? skb->len : MAX_PKT_HDR_LEN;
if (hdr_len > 0 && bpf_skb_load_bytes(skb, 0, pkt.hdr, hdr_len) < 0)
return;
bpf_perf_event_output(skb, &events, BPF_F_CURRENT_CPU, &pkt, sizeof(pkt));
}
SEC("tcx/ingress")
int tcx_ingress(struct __sk_buff *skb) {
sample_packet(skb, 0);
return TCX_NEXT;
}
SEC("tcx/egress")
int tcx_egress(struct __sk_buff *skb) {
sample_packet(skb, 1);
return TCX_NEXT;
}
The sample.bpf.c file Continue reading
Is an LLM a stubborn donkey, a genie, or a slot machine (and why)? Find out in the Who is LLM? article by Martin Fowler.
It is beginning to look like Intel plans to milk the impending 18A manufacturing process for a long time. …
Intel Puts The Process Horse Back In Front Of The Foundry Cart was written by Timothy Prickett Morgan at The Next Platform.